Hugging Face's logo

mkhalifa
/
instrucode

Model card Files Community
Muhammad Khalifa commited on
Commit
f633d8a
·
1 Parent(s): e81d122

Add code llama finetuned

Browse files
Files changed (7) hide show
  1. with_input/decomp_code_with_intermediates/code-llama-7b/best_model/adapter_config.json +21 -0
  2. with_input/decomp_code_with_intermediates/code-llama-7b/best_model/adapter_model.bin +3 -0
  3. with_input/decomp_code_with_intermediates/code-llama-7b/best_model/optimizer.pt +3 -0
  4. with_input/decomp_code_with_intermediates/code-llama-7b/best_model/rng_state.pth +3 -0
  5. with_input/decomp_code_with_intermediates/code-llama-7b/best_model/scheduler.pt +3 -0
  6. with_input/decomp_code_with_intermediates/code-llama-7b/best_model/trainer_state.json +3987 -0
  7. with_input/decomp_code_with_intermediates/code-llama-7b/best_model/training_args.bin +3 -0
with_input/decomp_code_with_intermediates/code-llama-7b/best_model/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model_name_or_path": "codellama/CodeLlama-7b-hf",
3
+ "bias": "none",
4
+ "enable_lora": null,
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "lora_alpha": 16,
9
+ "lora_dropout": 0.05,
10
+ "merge_weights": false,
11
+ "modules_to_save": null,
12
+ "peft_type": "LORA",
13
+ "r": 16,
14
+ "target_modules": [
15
+ "q_proj",
16
+ "k_proj",
17
+ "v_proj",
18
+ "o_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
with_input/decomp_code_with_intermediates/code-llama-7b/best_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a5f67f006e2c30a5287a1ace2446c0547cd4db9ee5898ba02f6d6c7c6f323cd8
3
+ size 67201357
with_input/decomp_code_with_intermediates/code-llama-7b/best_model/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c1108187daeae37960d2975963b3e6b02fcc8450f6d4285a6283e1aee290884
3
+ size 134433093
with_input/decomp_code_with_intermediates/code-llama-7b/best_model/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dbea311e5e3eb932b68869ced9dd85239e1909b99c710339eba9db6cc5be30b4
3
+ size 14575
with_input/decomp_code_with_intermediates/code-llama-7b/best_model/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b8a552baf6c150e604da119ed0d02582dd401ec37cc5af2c72ca7701792ca0b7
3
+ size 627
with_input/decomp_code_with_intermediates/code-llama-7b/best_model/trainer_state.json ADDED
@@ -0,0 +1,3987 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.3255977928638458,
3
+ "best_model_checkpoint": "checkpoints/instrucode/with_input/decomp_code_with_intermediates/code-llama-7b/checkpoint-6200",
4
+ "epoch": 3.987138263665595,
5
+ "eval_steps": 200,
6
+ "global_step": 6200,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 2.1e-05,
14
+ "loss": 0.9764,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.01,
19
+ "learning_rate": 5.1e-05,
20
+ "loss": 0.9894,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.02,
25
+ "learning_rate": 7.8e-05,
26
+ "loss": 0.927,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.03,
31
+ "learning_rate": 0.00010799999999999998,
32
+ "loss": 0.7652,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.03,
37
+ "learning_rate": 0.000138,
38
+ "loss": 0.5665,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.04,
43
+ "learning_rate": 0.000165,
44
+ "loss": 0.519,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.05,
49
+ "learning_rate": 0.00019199999999999998,
50
+ "loss": 0.4727,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.05,
55
+ "learning_rate": 0.00022199999999999998,
56
+ "loss": 0.4446,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.06,
61
+ "learning_rate": 0.00025199999999999995,
62
+ "loss": 0.4186,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.06,
67
+ "learning_rate": 0.00028199999999999997,
68
+ "loss": 0.4141,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.07,
73
+ "learning_rate": 0.00029984364820846905,
74
+ "loss": 0.4202,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.08,
79
+ "learning_rate": 0.00029945276872964165,
80
+ "loss": 0.3988,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.08,
85
+ "learning_rate": 0.0002990618892508143,
86
+ "loss": 0.4003,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.09,
91
+ "learning_rate": 0.00029867100977198696,
92
+ "loss": 0.3957,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.1,
97
+ "learning_rate": 0.0002982801302931596,
98
+ "loss": 0.3905,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.1,
103
+ "learning_rate": 0.0002978892508143322,
104
+ "loss": 0.3984,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.11,
109
+ "learning_rate": 0.00029749837133550487,
110
+ "loss": 0.3776,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.12,
115
+ "learning_rate": 0.00029710749185667747,
116
+ "loss": 0.3971,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.12,
121
+ "learning_rate": 0.0002967166123778501,
122
+ "loss": 0.382,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.13,
127
+ "learning_rate": 0.0002963257328990228,
128
+ "loss": 0.3754,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.13,
133
+ "eval_loss": 0.3773495554924011,
134
+ "eval_runtime": 502.3677,
135
+ "eval_samples_per_second": 3.981,
136
+ "eval_steps_per_second": 0.498,
137
+ "step": 200
138
+ },
139
+ {
140
+ "epoch": 0.14,
141
+ "learning_rate": 0.00029593485342019544,
142
+ "loss": 0.3761,
143
+ "step": 210
144
+ },
145
+ {
146
+ "epoch": 0.14,
147
+ "learning_rate": 0.00029554397394136804,
148
+ "loss": 0.372,
149
+ "step": 220
150
+ },
151
+ {
152
+ "epoch": 0.15,
153
+ "learning_rate": 0.0002951530944625407,
154
+ "loss": 0.3697,
155
+ "step": 230
156
+ },
157
+ {
158
+ "epoch": 0.15,
159
+ "learning_rate": 0.00029476221498371335,
160
+ "loss": 0.3651,
161
+ "step": 240
162
+ },
163
+ {
164
+ "epoch": 0.16,
165
+ "learning_rate": 0.00029437133550488595,
166
+ "loss": 0.3758,
167
+ "step": 250
168
+ },
169
+ {
170
+ "epoch": 0.17,
171
+ "learning_rate": 0.0002939804560260586,
172
+ "loss": 0.3722,
173
+ "step": 260
174
+ },
175
+ {
176
+ "epoch": 0.17,
177
+ "learning_rate": 0.0002935895765472312,
178
+ "loss": 0.3841,
179
+ "step": 270
180
+ },
181
+ {
182
+ "epoch": 0.18,
183
+ "learning_rate": 0.00029319869706840386,
184
+ "loss": 0.388,
185
+ "step": 280
186
+ },
187
+ {
188
+ "epoch": 0.19,
189
+ "learning_rate": 0.0002928078175895765,
190
+ "loss": 0.3775,
191
+ "step": 290
192
+ },
193
+ {
194
+ "epoch": 0.19,
195
+ "learning_rate": 0.00029241693811074917,
196
+ "loss": 0.3711,
197
+ "step": 300
198
+ },
199
+ {
200
+ "epoch": 0.2,
201
+ "learning_rate": 0.0002920260586319218,
202
+ "loss": 0.361,
203
+ "step": 310
204
+ },
205
+ {
206
+ "epoch": 0.21,
207
+ "learning_rate": 0.00029163517915309443,
208
+ "loss": 0.3709,
209
+ "step": 320
210
+ },
211
+ {
212
+ "epoch": 0.21,
213
+ "learning_rate": 0.0002912442996742671,
214
+ "loss": 0.3679,
215
+ "step": 330
216
+ },
217
+ {
218
+ "epoch": 0.22,
219
+ "learning_rate": 0.00029085342019543974,
220
+ "loss": 0.3775,
221
+ "step": 340
222
+ },
223
+ {
224
+ "epoch": 0.23,
225
+ "learning_rate": 0.00029046254071661234,
226
+ "loss": 0.3759,
227
+ "step": 350
228
+ },
229
+ {
230
+ "epoch": 0.23,
231
+ "learning_rate": 0.000290071661237785,
232
+ "loss": 0.3616,
233
+ "step": 360
234
+ },
235
+ {
236
+ "epoch": 0.24,
237
+ "learning_rate": 0.0002896807817589576,
238
+ "loss": 0.3606,
239
+ "step": 370
240
+ },
241
+ {
242
+ "epoch": 0.24,
243
+ "learning_rate": 0.00028928990228013025,
244
+ "loss": 0.3653,
245
+ "step": 380
246
+ },
247
+ {
248
+ "epoch": 0.25,
249
+ "learning_rate": 0.0002888990228013029,
250
+ "loss": 0.3586,
251
+ "step": 390
252
+ },
253
+ {
254
+ "epoch": 0.26,
255
+ "learning_rate": 0.00028850814332247556,
256
+ "loss": 0.3548,
257
+ "step": 400
258
+ },
259
+ {
260
+ "epoch": 0.26,
261
+ "eval_loss": 0.36366358399391174,
262
+ "eval_runtime": 503.4894,
263
+ "eval_samples_per_second": 3.972,
264
+ "eval_steps_per_second": 0.497,
265
+ "step": 400
266
+ },
267
+ {
268
+ "epoch": 0.26,
269
+ "learning_rate": 0.0002881172638436482,
270
+ "loss": 0.3755,
271
+ "step": 410
272
+ },
273
+ {
274
+ "epoch": 0.27,
275
+ "learning_rate": 0.0002877263843648208,
276
+ "loss": 0.3778,
277
+ "step": 420
278
+ },
279
+ {
280
+ "epoch": 0.28,
281
+ "learning_rate": 0.00028733550488599347,
282
+ "loss": 0.3673,
283
+ "step": 430
284
+ },
285
+ {
286
+ "epoch": 0.28,
287
+ "learning_rate": 0.0002869446254071661,
288
+ "loss": 0.3544,
289
+ "step": 440
290
+ },
291
+ {
292
+ "epoch": 0.29,
293
+ "learning_rate": 0.00028655374592833873,
294
+ "loss": 0.3651,
295
+ "step": 450
296
+ },
297
+ {
298
+ "epoch": 0.3,
299
+ "learning_rate": 0.0002861628664495114,
300
+ "loss": 0.3536,
301
+ "step": 460
302
+ },
303
+ {
304
+ "epoch": 0.3,
305
+ "learning_rate": 0.00028577198697068404,
306
+ "loss": 0.3494,
307
+ "step": 470
308
+ },
309
+ {
310
+ "epoch": 0.31,
311
+ "learning_rate": 0.00028538110749185664,
312
+ "loss": 0.3648,
313
+ "step": 480
314
+ },
315
+ {
316
+ "epoch": 0.32,
317
+ "learning_rate": 0.0002849902280130293,
318
+ "loss": 0.3489,
319
+ "step": 490
320
+ },
321
+ {
322
+ "epoch": 0.32,
323
+ "learning_rate": 0.00028459934853420195,
324
+ "loss": 0.3574,
325
+ "step": 500
326
+ },
327
+ {
328
+ "epoch": 0.33,
329
+ "learning_rate": 0.00028420846905537455,
330
+ "loss": 0.353,
331
+ "step": 510
332
+ },
333
+ {
334
+ "epoch": 0.33,
335
+ "learning_rate": 0.0002838175895765472,
336
+ "loss": 0.3545,
337
+ "step": 520
338
+ },
339
+ {
340
+ "epoch": 0.34,
341
+ "learning_rate": 0.0002834267100977198,
342
+ "loss": 0.3598,
343
+ "step": 530
344
+ },
345
+ {
346
+ "epoch": 0.35,
347
+ "learning_rate": 0.00028303583061889246,
348
+ "loss": 0.3582,
349
+ "step": 540
350
+ },
351
+ {
352
+ "epoch": 0.35,
353
+ "learning_rate": 0.0002826449511400651,
354
+ "loss": 0.3777,
355
+ "step": 550
356
+ },
357
+ {
358
+ "epoch": 0.36,
359
+ "learning_rate": 0.0002822540716612378,
360
+ "loss": 0.3441,
361
+ "step": 560
362
+ },
363
+ {
364
+ "epoch": 0.37,
365
+ "learning_rate": 0.00028186319218241043,
366
+ "loss": 0.3637,
367
+ "step": 570
368
+ },
369
+ {
370
+ "epoch": 0.37,
371
+ "learning_rate": 0.00028147231270358303,
372
+ "loss": 0.3576,
373
+ "step": 580
374
+ },
375
+ {
376
+ "epoch": 0.38,
377
+ "learning_rate": 0.0002810814332247557,
378
+ "loss": 0.3576,
379
+ "step": 590
380
+ },
381
+ {
382
+ "epoch": 0.39,
383
+ "learning_rate": 0.00028069055374592834,
384
+ "loss": 0.3706,
385
+ "step": 600
386
+ },
387
+ {
388
+ "epoch": 0.39,
389
+ "eval_loss": 0.35524383187294006,
390
+ "eval_runtime": 504.6501,
391
+ "eval_samples_per_second": 3.963,
392
+ "eval_steps_per_second": 0.495,
393
+ "step": 600
394
+ },
395
+ {
396
+ "epoch": 0.39,
397
+ "learning_rate": 0.00028029967426710094,
398
+ "loss": 0.3604,
399
+ "step": 610
400
+ },
401
+ {
402
+ "epoch": 0.4,
403
+ "learning_rate": 0.0002799087947882736,
404
+ "loss": 0.3666,
405
+ "step": 620
406
+ },
407
+ {
408
+ "epoch": 0.41,
409
+ "learning_rate": 0.0002795179153094462,
410
+ "loss": 0.3618,
411
+ "step": 630
412
+ },
413
+ {
414
+ "epoch": 0.41,
415
+ "learning_rate": 0.00027912703583061885,
416
+ "loss": 0.3594,
417
+ "step": 640
418
+ },
419
+ {
420
+ "epoch": 0.42,
421
+ "learning_rate": 0.0002787361563517915,
422
+ "loss": 0.3614,
423
+ "step": 650
424
+ },
425
+ {
426
+ "epoch": 0.42,
427
+ "learning_rate": 0.00027834527687296416,
428
+ "loss": 0.3552,
429
+ "step": 660
430
+ },
431
+ {
432
+ "epoch": 0.43,
433
+ "learning_rate": 0.0002779543973941368,
434
+ "loss": 0.362,
435
+ "step": 670
436
+ },
437
+ {
438
+ "epoch": 0.44,
439
+ "learning_rate": 0.0002775635179153094,
440
+ "loss": 0.3562,
441
+ "step": 680
442
+ },
443
+ {
444
+ "epoch": 0.44,
445
+ "learning_rate": 0.0002771726384364821,
446
+ "loss": 0.354,
447
+ "step": 690
448
+ },
449
+ {
450
+ "epoch": 0.45,
451
+ "learning_rate": 0.0002767817589576547,
452
+ "loss": 0.3425,
453
+ "step": 700
454
+ },
455
+ {
456
+ "epoch": 0.46,
457
+ "learning_rate": 0.00027639087947882733,
458
+ "loss": 0.3523,
459
+ "step": 710
460
+ },
461
+ {
462
+ "epoch": 0.46,
463
+ "learning_rate": 0.000276,
464
+ "loss": 0.3557,
465
+ "step": 720
466
+ },
467
+ {
468
+ "epoch": 0.47,
469
+ "learning_rate": 0.0002756091205211726,
470
+ "loss": 0.3501,
471
+ "step": 730
472
+ },
473
+ {
474
+ "epoch": 0.48,
475
+ "learning_rate": 0.00027521824104234524,
476
+ "loss": 0.3547,
477
+ "step": 740
478
+ },
479
+ {
480
+ "epoch": 0.48,
481
+ "learning_rate": 0.0002748273615635179,
482
+ "loss": 0.352,
483
+ "step": 750
484
+ },
485
+ {
486
+ "epoch": 0.49,
487
+ "learning_rate": 0.00027443648208469055,
488
+ "loss": 0.3625,
489
+ "step": 760
490
+ },
491
+ {
492
+ "epoch": 0.5,
493
+ "learning_rate": 0.00027404560260586315,
494
+ "loss": 0.35,
495
+ "step": 770
496
+ },
497
+ {
498
+ "epoch": 0.5,
499
+ "learning_rate": 0.0002736547231270358,
500
+ "loss": 0.3647,
501
+ "step": 780
502
+ },
503
+ {
504
+ "epoch": 0.51,
505
+ "learning_rate": 0.0002732638436482084,
506
+ "loss": 0.3364,
507
+ "step": 790
508
+ },
509
+ {
510
+ "epoch": 0.51,
511
+ "learning_rate": 0.00027287296416938107,
512
+ "loss": 0.3603,
513
+ "step": 800
514
+ },
515
+ {
516
+ "epoch": 0.51,
517
+ "eval_loss": 0.3504890501499176,
518
+ "eval_runtime": 503.7269,
519
+ "eval_samples_per_second": 3.97,
520
+ "eval_steps_per_second": 0.496,
521
+ "step": 800
522
+ },
523
+ {
524
+ "epoch": 0.52,
525
+ "learning_rate": 0.0002724820846905537,
526
+ "loss": 0.3633,
527
+ "step": 810
528
+ },
529
+ {
530
+ "epoch": 0.53,
531
+ "learning_rate": 0.0002720912052117264,
532
+ "loss": 0.3473,
533
+ "step": 820
534
+ },
535
+ {
536
+ "epoch": 0.53,
537
+ "learning_rate": 0.000271700325732899,
538
+ "loss": 0.355,
539
+ "step": 830
540
+ },
541
+ {
542
+ "epoch": 0.54,
543
+ "learning_rate": 0.00027130944625407163,
544
+ "loss": 0.3604,
545
+ "step": 840
546
+ },
547
+ {
548
+ "epoch": 0.55,
549
+ "learning_rate": 0.0002709185667752443,
550
+ "loss": 0.3579,
551
+ "step": 850
552
+ },
553
+ {
554
+ "epoch": 0.55,
555
+ "learning_rate": 0.00027052768729641694,
556
+ "loss": 0.3412,
557
+ "step": 860
558
+ },
559
+ {
560
+ "epoch": 0.56,
561
+ "learning_rate": 0.00027013680781758954,
562
+ "loss": 0.3539,
563
+ "step": 870
564
+ },
565
+ {
566
+ "epoch": 0.57,
567
+ "learning_rate": 0.0002697459283387622,
568
+ "loss": 0.3687,
569
+ "step": 880
570
+ },
571
+ {
572
+ "epoch": 0.57,
573
+ "learning_rate": 0.0002693550488599348,
574
+ "loss": 0.3486,
575
+ "step": 890
576
+ },
577
+ {
578
+ "epoch": 0.58,
579
+ "learning_rate": 0.00026896416938110746,
580
+ "loss": 0.3415,
581
+ "step": 900
582
+ },
583
+ {
584
+ "epoch": 0.59,
585
+ "learning_rate": 0.0002685732899022801,
586
+ "loss": 0.346,
587
+ "step": 910
588
+ },
589
+ {
590
+ "epoch": 0.59,
591
+ "learning_rate": 0.00026818241042345277,
592
+ "loss": 0.3568,
593
+ "step": 920
594
+ },
595
+ {
596
+ "epoch": 0.6,
597
+ "learning_rate": 0.0002677915309446254,
598
+ "loss": 0.3482,
599
+ "step": 930
600
+ },
601
+ {
602
+ "epoch": 0.6,
603
+ "learning_rate": 0.000267400651465798,
604
+ "loss": 0.3555,
605
+ "step": 940
606
+ },
607
+ {
608
+ "epoch": 0.61,
609
+ "learning_rate": 0.0002670097719869707,
610
+ "loss": 0.3424,
611
+ "step": 950
612
+ },
613
+ {
614
+ "epoch": 0.62,
615
+ "learning_rate": 0.0002666188925081433,
616
+ "loss": 0.3568,
617
+ "step": 960
618
+ },
619
+ {
620
+ "epoch": 0.62,
621
+ "learning_rate": 0.00026622801302931593,
622
+ "loss": 0.3461,
623
+ "step": 970
624
+ },
625
+ {
626
+ "epoch": 0.63,
627
+ "learning_rate": 0.0002658371335504886,
628
+ "loss": 0.3459,
629
+ "step": 980
630
+ },
631
+ {
632
+ "epoch": 0.64,
633
+ "learning_rate": 0.0002654462540716612,
634
+ "loss": 0.3625,
635
+ "step": 990
636
+ },
637
+ {
638
+ "epoch": 0.64,
639
+ "learning_rate": 0.00026505537459283385,
640
+ "loss": 0.3507,
641
+ "step": 1000
642
+ },
643
+ {
644
+ "epoch": 0.64,
645
+ "eval_loss": 0.346477210521698,
646
+ "eval_runtime": 508.3509,
647
+ "eval_samples_per_second": 3.934,
648
+ "eval_steps_per_second": 0.492,
649
+ "step": 1000
650
+ },
651
+ {
652
+ "epoch": 0.65,
653
+ "learning_rate": 0.0002646644951140065,
654
+ "loss": 0.3519,
655
+ "step": 1010
656
+ },
657
+ {
658
+ "epoch": 0.66,
659
+ "learning_rate": 0.00026427361563517916,
660
+ "loss": 0.3535,
661
+ "step": 1020
662
+ },
663
+ {
664
+ "epoch": 0.66,
665
+ "learning_rate": 0.00026388273615635176,
666
+ "loss": 0.3486,
667
+ "step": 1030
668
+ },
669
+ {
670
+ "epoch": 0.67,
671
+ "learning_rate": 0.0002634918566775244,
672
+ "loss": 0.3503,
673
+ "step": 1040
674
+ },
675
+ {
676
+ "epoch": 0.68,
677
+ "learning_rate": 0.000263100977198697,
678
+ "loss": 0.3564,
679
+ "step": 1050
680
+ },
681
+ {
682
+ "epoch": 0.68,
683
+ "learning_rate": 0.00026271009771986967,
684
+ "loss": 0.3426,
685
+ "step": 1060
686
+ },
687
+ {
688
+ "epoch": 0.69,
689
+ "learning_rate": 0.0002623192182410423,
690
+ "loss": 0.3383,
691
+ "step": 1070
692
+ },
693
+ {
694
+ "epoch": 0.69,
695
+ "learning_rate": 0.000261928338762215,
696
+ "loss": 0.3595,
697
+ "step": 1080
698
+ },
699
+ {
700
+ "epoch": 0.7,
701
+ "learning_rate": 0.0002615374592833876,
702
+ "loss": 0.3508,
703
+ "step": 1090
704
+ },
705
+ {
706
+ "epoch": 0.71,
707
+ "learning_rate": 0.00026114657980456024,
708
+ "loss": 0.3587,
709
+ "step": 1100
710
+ },
711
+ {
712
+ "epoch": 0.71,
713
+ "learning_rate": 0.0002607557003257329,
714
+ "loss": 0.3488,
715
+ "step": 1110
716
+ },
717
+ {
718
+ "epoch": 0.72,
719
+ "learning_rate": 0.00026036482084690555,
720
+ "loss": 0.3437,
721
+ "step": 1120
722
+ },
723
+ {
724
+ "epoch": 0.73,
725
+ "learning_rate": 0.00025997394136807815,
726
+ "loss": 0.3526,
727
+ "step": 1130
728
+ },
729
+ {
730
+ "epoch": 0.73,
731
+ "learning_rate": 0.0002595830618892508,
732
+ "loss": 0.3535,
733
+ "step": 1140
734
+ },
735
+ {
736
+ "epoch": 0.74,
737
+ "learning_rate": 0.0002591921824104234,
738
+ "loss": 0.3445,
739
+ "step": 1150
740
+ },
741
+ {
742
+ "epoch": 0.75,
743
+ "learning_rate": 0.00025880130293159606,
744
+ "loss": 0.3503,
745
+ "step": 1160
746
+ },
747
+ {
748
+ "epoch": 0.75,
749
+ "learning_rate": 0.0002584104234527687,
750
+ "loss": 0.353,
751
+ "step": 1170
752
+ },
753
+ {
754
+ "epoch": 0.76,
755
+ "learning_rate": 0.00025801954397394137,
756
+ "loss": 0.3503,
757
+ "step": 1180
758
+ },
759
+ {
760
+ "epoch": 0.77,
761
+ "learning_rate": 0.00025762866449511397,
762
+ "loss": 0.3416,
763
+ "step": 1190
764
+ },
765
+ {
766
+ "epoch": 0.77,
767
+ "learning_rate": 0.0002572377850162866,
768
+ "loss": 0.35,
769
+ "step": 1200
770
+ },
771
+ {
772
+ "epoch": 0.77,
773
+ "eval_loss": 0.3429148495197296,
774
+ "eval_runtime": 511.8802,
775
+ "eval_samples_per_second": 3.907,
776
+ "eval_steps_per_second": 0.488,
777
+ "step": 1200
778
+ },
779
+ {
780
+ "epoch": 0.78,
781
+ "learning_rate": 0.0002568469055374593,
782
+ "loss": 0.3561,
783
+ "step": 1210
784
+ },
785
+ {
786
+ "epoch": 0.78,
787
+ "learning_rate": 0.0002564560260586319,
788
+ "loss": 0.3492,
789
+ "step": 1220
790
+ },
791
+ {
792
+ "epoch": 0.79,
793
+ "learning_rate": 0.00025606514657980454,
794
+ "loss": 0.3416,
795
+ "step": 1230
796
+ },
797
+ {
798
+ "epoch": 0.8,
799
+ "learning_rate": 0.0002556742671009772,
800
+ "loss": 0.3506,
801
+ "step": 1240
802
+ },
803
+ {
804
+ "epoch": 0.8,
805
+ "learning_rate": 0.0002552833876221498,
806
+ "loss": 0.3524,
807
+ "step": 1250
808
+ },
809
+ {
810
+ "epoch": 0.81,
811
+ "learning_rate": 0.00025489250814332245,
812
+ "loss": 0.3593,
813
+ "step": 1260
814
+ },
815
+ {
816
+ "epoch": 0.82,
817
+ "learning_rate": 0.0002545016286644951,
818
+ "loss": 0.3501,
819
+ "step": 1270
820
+ },
821
+ {
822
+ "epoch": 0.82,
823
+ "learning_rate": 0.00025411074918566776,
824
+ "loss": 0.3497,
825
+ "step": 1280
826
+ },
827
+ {
828
+ "epoch": 0.83,
829
+ "learning_rate": 0.00025371986970684036,
830
+ "loss": 0.3401,
831
+ "step": 1290
832
+ },
833
+ {
834
+ "epoch": 0.84,
835
+ "learning_rate": 0.000253328990228013,
836
+ "loss": 0.347,
837
+ "step": 1300
838
+ },
839
+ {
840
+ "epoch": 0.84,
841
+ "learning_rate": 0.0002529381107491856,
842
+ "loss": 0.3392,
843
+ "step": 1310
844
+ },
845
+ {
846
+ "epoch": 0.85,
847
+ "learning_rate": 0.00025254723127035827,
848
+ "loss": 0.3457,
849
+ "step": 1320
850
+ },
851
+ {
852
+ "epoch": 0.86,
853
+ "learning_rate": 0.0002521563517915309,
854
+ "loss": 0.3511,
855
+ "step": 1330
856
+ },
857
+ {
858
+ "epoch": 0.86,
859
+ "learning_rate": 0.0002517654723127036,
860
+ "loss": 0.3493,
861
+ "step": 1340
862
+ },
863
+ {
864
+ "epoch": 0.87,
865
+ "learning_rate": 0.0002513745928338762,
866
+ "loss": 0.3256,
867
+ "step": 1350
868
+ },
869
+ {
870
+ "epoch": 0.87,
871
+ "learning_rate": 0.00025098371335504884,
872
+ "loss": 0.3566,
873
+ "step": 1360
874
+ },
875
+ {
876
+ "epoch": 0.88,
877
+ "learning_rate": 0.0002505928338762215,
878
+ "loss": 0.3446,
879
+ "step": 1370
880
+ },
881
+ {
882
+ "epoch": 0.89,
883
+ "learning_rate": 0.00025020195439739415,
884
+ "loss": 0.341,
885
+ "step": 1380
886
+ },
887
+ {
888
+ "epoch": 0.89,
889
+ "learning_rate": 0.00024981107491856675,
890
+ "loss": 0.3446,
891
+ "step": 1390
892
+ },
893
+ {
894
+ "epoch": 0.9,
895
+ "learning_rate": 0.0002494201954397394,
896
+ "loss": 0.3388,
897
+ "step": 1400
898
+ },
899
+ {
900
+ "epoch": 0.9,
901
+ "eval_loss": 0.3407489061355591,
902
+ "eval_runtime": 513.6559,
903
+ "eval_samples_per_second": 3.894,
904
+ "eval_steps_per_second": 0.487,
905
+ "step": 1400
906
+ },
907
+ {
908
+ "epoch": 0.91,
909
+ "learning_rate": 0.000249029315960912,
910
+ "loss": 0.342,
911
+ "step": 1410
912
+ },
913
+ {
914
+ "epoch": 0.91,
915
+ "learning_rate": 0.00024863843648208466,
916
+ "loss": 0.3335,
917
+ "step": 1420
918
+ },
919
+ {
920
+ "epoch": 0.92,
921
+ "learning_rate": 0.0002482475570032573,
922
+ "loss": 0.3335,
923
+ "step": 1430
924
+ },
925
+ {
926
+ "epoch": 0.93,
927
+ "learning_rate": 0.00024785667752442997,
928
+ "loss": 0.3453,
929
+ "step": 1440
930
+ },
931
+ {
932
+ "epoch": 0.93,
933
+ "learning_rate": 0.00024746579804560257,
934
+ "loss": 0.3445,
935
+ "step": 1450
936
+ },
937
+ {
938
+ "epoch": 0.94,
939
+ "learning_rate": 0.00024707491856677523,
940
+ "loss": 0.3369,
941
+ "step": 1460
942
+ },
943
+ {
944
+ "epoch": 0.95,
945
+ "learning_rate": 0.0002466840390879479,
946
+ "loss": 0.3386,
947
+ "step": 1470
948
+ },
949
+ {
950
+ "epoch": 0.95,
951
+ "learning_rate": 0.0002462931596091205,
952
+ "loss": 0.3322,
953
+ "step": 1480
954
+ },
955
+ {
956
+ "epoch": 0.96,
957
+ "learning_rate": 0.00024590228013029314,
958
+ "loss": 0.3374,
959
+ "step": 1490
960
+ },
961
+ {
962
+ "epoch": 0.96,
963
+ "learning_rate": 0.00024551140065146574,
964
+ "loss": 0.3372,
965
+ "step": 1500
966
+ },
967
+ {
968
+ "epoch": 0.97,
969
+ "learning_rate": 0.0002451205211726384,
970
+ "loss": 0.3421,
971
+ "step": 1510
972
+ },
973
+ {
974
+ "epoch": 0.98,
975
+ "learning_rate": 0.00024472964169381105,
976
+ "loss": 0.3316,
977
+ "step": 1520
978
+ },
979
+ {
980
+ "epoch": 0.98,
981
+ "learning_rate": 0.0002443387622149837,
982
+ "loss": 0.3264,
983
+ "step": 1530
984
+ },
985
+ {
986
+ "epoch": 0.99,
987
+ "learning_rate": 0.00024394788273615633,
988
+ "loss": 0.3443,
989
+ "step": 1540
990
+ },
991
+ {
992
+ "epoch": 1.0,
993
+ "learning_rate": 0.00024355700325732896,
994
+ "loss": 0.3359,
995
+ "step": 1550
996
+ },
997
+ {
998
+ "epoch": 1.0,
999
+ "learning_rate": 0.0002431661237785016,
1000
+ "loss": 0.3305,
1001
+ "step": 1560
1002
+ },
1003
+ {
1004
+ "epoch": 1.01,
1005
+ "learning_rate": 0.00024277524429967425,
1006
+ "loss": 0.3444,
1007
+ "step": 1570
1008
+ },
1009
+ {
1010
+ "epoch": 1.02,
1011
+ "learning_rate": 0.0002423843648208469,
1012
+ "loss": 0.3377,
1013
+ "step": 1580
1014
+ },
1015
+ {
1016
+ "epoch": 1.02,
1017
+ "learning_rate": 0.00024199348534201953,
1018
+ "loss": 0.338,
1019
+ "step": 1590
1020
+ },
1021
+ {
1022
+ "epoch": 1.03,
1023
+ "learning_rate": 0.00024160260586319216,
1024
+ "loss": 0.326,
1025
+ "step": 1600
1026
+ },
1027
+ {
1028
+ "epoch": 1.03,
1029
+ "eval_loss": 0.33855700492858887,
1030
+ "eval_runtime": 507.6074,
1031
+ "eval_samples_per_second": 3.94,
1032
+ "eval_steps_per_second": 0.493,
1033
+ "step": 1600
1034
+ },
1035
+ {
1036
+ "epoch": 1.04,
1037
+ "learning_rate": 0.00024121172638436479,
1038
+ "loss": 0.3296,
1039
+ "step": 1610
1040
+ },
1041
+ {
1042
+ "epoch": 1.04,
1043
+ "learning_rate": 0.00024082084690553744,
1044
+ "loss": 0.3276,
1045
+ "step": 1620
1046
+ },
1047
+ {
1048
+ "epoch": 1.05,
1049
+ "learning_rate": 0.00024042996742671007,
1050
+ "loss": 0.3373,
1051
+ "step": 1630
1052
+ },
1053
+ {
1054
+ "epoch": 1.05,
1055
+ "learning_rate": 0.00024003908794788272,
1056
+ "loss": 0.3362,
1057
+ "step": 1640
1058
+ },
1059
+ {
1060
+ "epoch": 1.06,
1061
+ "learning_rate": 0.00023964820846905538,
1062
+ "loss": 0.3314,
1063
+ "step": 1650
1064
+ },
1065
+ {
1066
+ "epoch": 1.07,
1067
+ "learning_rate": 0.00023925732899022798,
1068
+ "loss": 0.3407,
1069
+ "step": 1660
1070
+ },
1071
+ {
1072
+ "epoch": 1.07,
1073
+ "learning_rate": 0.00023886644951140064,
1074
+ "loss": 0.3264,
1075
+ "step": 1670
1076
+ },
1077
+ {
1078
+ "epoch": 1.08,
1079
+ "learning_rate": 0.00023847557003257326,
1080
+ "loss": 0.3458,
1081
+ "step": 1680
1082
+ },
1083
+ {
1084
+ "epoch": 1.09,
1085
+ "learning_rate": 0.00023808469055374592,
1086
+ "loss": 0.3234,
1087
+ "step": 1690
1088
+ },
1089
+ {
1090
+ "epoch": 1.09,
1091
+ "learning_rate": 0.00023769381107491857,
1092
+ "loss": 0.3166,
1093
+ "step": 1700
1094
+ },
1095
+ {
1096
+ "epoch": 1.1,
1097
+ "learning_rate": 0.00023730293159609118,
1098
+ "loss": 0.3302,
1099
+ "step": 1710
1100
+ },
1101
+ {
1102
+ "epoch": 1.11,
1103
+ "learning_rate": 0.00023691205211726383,
1104
+ "loss": 0.3345,
1105
+ "step": 1720
1106
+ },
1107
+ {
1108
+ "epoch": 1.11,
1109
+ "learning_rate": 0.00023652117263843646,
1110
+ "loss": 0.3428,
1111
+ "step": 1730
1112
+ },
1113
+ {
1114
+ "epoch": 1.12,
1115
+ "learning_rate": 0.00023613029315960911,
1116
+ "loss": 0.3279,
1117
+ "step": 1740
1118
+ },
1119
+ {
1120
+ "epoch": 1.13,
1121
+ "learning_rate": 0.00023573941368078174,
1122
+ "loss": 0.3369,
1123
+ "step": 1750
1124
+ },
1125
+ {
1126
+ "epoch": 1.13,
1127
+ "learning_rate": 0.00023534853420195437,
1128
+ "loss": 0.3342,
1129
+ "step": 1760
1130
+ },
1131
+ {
1132
+ "epoch": 1.14,
1133
+ "learning_rate": 0.000234957654723127,
1134
+ "loss": 0.3384,
1135
+ "step": 1770
1136
+ },
1137
+ {
1138
+ "epoch": 1.14,
1139
+ "learning_rate": 0.00023456677524429965,
1140
+ "loss": 0.3276,
1141
+ "step": 1780
1142
+ },
1143
+ {
1144
+ "epoch": 1.15,
1145
+ "learning_rate": 0.0002341758957654723,
1146
+ "loss": 0.3337,
1147
+ "step": 1790
1148
+ },
1149
+ {
1150
+ "epoch": 1.16,
1151
+ "learning_rate": 0.00023378501628664494,
1152
+ "loss": 0.3257,
1153
+ "step": 1800
1154
+ },
1155
+ {
1156
+ "epoch": 1.16,
1157
+ "eval_loss": 0.3377059996128082,
1158
+ "eval_runtime": 515.0038,
1159
+ "eval_samples_per_second": 3.883,
1160
+ "eval_steps_per_second": 0.485,
1161
+ "step": 1800
1162
+ },
1163
+ {
1164
+ "epoch": 1.16,
1165
+ "learning_rate": 0.00023339413680781757,
1166
+ "loss": 0.3293,
1167
+ "step": 1810
1168
+ },
1169
+ {
1170
+ "epoch": 1.17,
1171
+ "learning_rate": 0.0002330032573289902,
1172
+ "loss": 0.3354,
1173
+ "step": 1820
1174
+ },
1175
+ {
1176
+ "epoch": 1.18,
1177
+ "learning_rate": 0.00023261237785016285,
1178
+ "loss": 0.351,
1179
+ "step": 1830
1180
+ },
1181
+ {
1182
+ "epoch": 1.18,
1183
+ "learning_rate": 0.0002322214983713355,
1184
+ "loss": 0.332,
1185
+ "step": 1840
1186
+ },
1187
+ {
1188
+ "epoch": 1.19,
1189
+ "learning_rate": 0.00023183061889250813,
1190
+ "loss": 0.3231,
1191
+ "step": 1850
1192
+ },
1193
+ {
1194
+ "epoch": 1.2,
1195
+ "learning_rate": 0.00023143973941368076,
1196
+ "loss": 0.3384,
1197
+ "step": 1860
1198
+ },
1199
+ {
1200
+ "epoch": 1.2,
1201
+ "learning_rate": 0.0002310488599348534,
1202
+ "loss": 0.3279,
1203
+ "step": 1870
1204
+ },
1205
+ {
1206
+ "epoch": 1.21,
1207
+ "learning_rate": 0.00023065798045602604,
1208
+ "loss": 0.3205,
1209
+ "step": 1880
1210
+ },
1211
+ {
1212
+ "epoch": 1.22,
1213
+ "learning_rate": 0.00023026710097719867,
1214
+ "loss": 0.3311,
1215
+ "step": 1890
1216
+ },
1217
+ {
1218
+ "epoch": 1.22,
1219
+ "learning_rate": 0.00022987622149837133,
1220
+ "loss": 0.3301,
1221
+ "step": 1900
1222
+ },
1223
+ {
1224
+ "epoch": 1.23,
1225
+ "learning_rate": 0.00022948534201954393,
1226
+ "loss": 0.3112,
1227
+ "step": 1910
1228
+ },
1229
+ {
1230
+ "epoch": 1.23,
1231
+ "learning_rate": 0.00022909446254071658,
1232
+ "loss": 0.3371,
1233
+ "step": 1920
1234
+ },
1235
+ {
1236
+ "epoch": 1.24,
1237
+ "learning_rate": 0.00022870358306188924,
1238
+ "loss": 0.3454,
1239
+ "step": 1930
1240
+ },
1241
+ {
1242
+ "epoch": 1.25,
1243
+ "learning_rate": 0.00022831270358306187,
1244
+ "loss": 0.3285,
1245
+ "step": 1940
1246
+ },
1247
+ {
1248
+ "epoch": 1.25,
1249
+ "learning_rate": 0.00022792182410423452,
1250
+ "loss": 0.3356,
1251
+ "step": 1950
1252
+ },
1253
+ {
1254
+ "epoch": 1.26,
1255
+ "learning_rate": 0.00022753094462540712,
1256
+ "loss": 0.3209,
1257
+ "step": 1960
1258
+ },
1259
+ {
1260
+ "epoch": 1.27,
1261
+ "learning_rate": 0.00022714006514657978,
1262
+ "loss": 0.3182,
1263
+ "step": 1970
1264
+ },
1265
+ {
1266
+ "epoch": 1.27,
1267
+ "learning_rate": 0.00022674918566775243,
1268
+ "loss": 0.3303,
1269
+ "step": 1980
1270
+ },
1271
+ {
1272
+ "epoch": 1.28,
1273
+ "learning_rate": 0.00022635830618892506,
1274
+ "loss": 0.3318,
1275
+ "step": 1990
1276
+ },
1277
+ {
1278
+ "epoch": 1.29,
1279
+ "learning_rate": 0.00022596742671009772,
1280
+ "loss": 0.3367,
1281
+ "step": 2000
1282
+ },
1283
+ {
1284
+ "epoch": 1.29,
1285
+ "eval_loss": 0.3357446789741516,
1286
+ "eval_runtime": 516.1777,
1287
+ "eval_samples_per_second": 3.875,
1288
+ "eval_steps_per_second": 0.484,
1289
+ "step": 2000
1290
+ },
1291
+ {
1292
+ "epoch": 1.29,
1293
+ "learning_rate": 0.00022557654723127034,
1294
+ "loss": 0.3204,
1295
+ "step": 2010
1296
+ },
1297
+ {
1298
+ "epoch": 1.3,
1299
+ "learning_rate": 0.00022518566775244297,
1300
+ "loss": 0.3311,
1301
+ "step": 2020
1302
+ },
1303
+ {
1304
+ "epoch": 1.31,
1305
+ "learning_rate": 0.0002247947882736156,
1306
+ "loss": 0.3288,
1307
+ "step": 2030
1308
+ },
1309
+ {
1310
+ "epoch": 1.31,
1311
+ "learning_rate": 0.00022440390879478826,
1312
+ "loss": 0.319,
1313
+ "step": 2040
1314
+ },
1315
+ {
1316
+ "epoch": 1.32,
1317
+ "learning_rate": 0.0002240130293159609,
1318
+ "loss": 0.3254,
1319
+ "step": 2050
1320
+ },
1321
+ {
1322
+ "epoch": 1.32,
1323
+ "learning_rate": 0.00022362214983713354,
1324
+ "loss": 0.3281,
1325
+ "step": 2060
1326
+ },
1327
+ {
1328
+ "epoch": 1.33,
1329
+ "learning_rate": 0.00022323127035830617,
1330
+ "loss": 0.3254,
1331
+ "step": 2070
1332
+ },
1333
+ {
1334
+ "epoch": 1.34,
1335
+ "learning_rate": 0.0002228403908794788,
1336
+ "loss": 0.3404,
1337
+ "step": 2080
1338
+ },
1339
+ {
1340
+ "epoch": 1.34,
1341
+ "learning_rate": 0.00022244951140065145,
1342
+ "loss": 0.3234,
1343
+ "step": 2090
1344
+ },
1345
+ {
1346
+ "epoch": 1.35,
1347
+ "learning_rate": 0.00022205863192182408,
1348
+ "loss": 0.3244,
1349
+ "step": 2100
1350
+ },
1351
+ {
1352
+ "epoch": 1.36,
1353
+ "learning_rate": 0.00022166775244299673,
1354
+ "loss": 0.3157,
1355
+ "step": 2110
1356
+ },
1357
+ {
1358
+ "epoch": 1.36,
1359
+ "learning_rate": 0.00022127687296416936,
1360
+ "loss": 0.3358,
1361
+ "step": 2120
1362
+ },
1363
+ {
1364
+ "epoch": 1.37,
1365
+ "learning_rate": 0.000220885993485342,
1366
+ "loss": 0.3268,
1367
+ "step": 2130
1368
+ },
1369
+ {
1370
+ "epoch": 1.38,
1371
+ "learning_rate": 0.00022049511400651465,
1372
+ "loss": 0.3371,
1373
+ "step": 2140
1374
+ },
1375
+ {
1376
+ "epoch": 1.38,
1377
+ "learning_rate": 0.00022010423452768727,
1378
+ "loss": 0.3397,
1379
+ "step": 2150
1380
+ },
1381
+ {
1382
+ "epoch": 1.39,
1383
+ "learning_rate": 0.00021971335504885993,
1384
+ "loss": 0.3309,
1385
+ "step": 2160
1386
+ },
1387
+ {
1388
+ "epoch": 1.4,
1389
+ "learning_rate": 0.00021932247557003253,
1390
+ "loss": 0.3165,
1391
+ "step": 2170
1392
+ },
1393
+ {
1394
+ "epoch": 1.4,
1395
+ "learning_rate": 0.00021893159609120519,
1396
+ "loss": 0.315,
1397
+ "step": 2180
1398
+ },
1399
+ {
1400
+ "epoch": 1.41,
1401
+ "learning_rate": 0.00021854071661237784,
1402
+ "loss": 0.3335,
1403
+ "step": 2190
1404
+ },
1405
+ {
1406
+ "epoch": 1.41,
1407
+ "learning_rate": 0.00021814983713355047,
1408
+ "loss": 0.3258,
1409
+ "step": 2200
1410
+ },
1411
+ {
1412
+ "epoch": 1.41,
1413
+ "eval_loss": 0.33434125781059265,
1414
+ "eval_runtime": 507.9975,
1415
+ "eval_samples_per_second": 3.937,
1416
+ "eval_steps_per_second": 0.492,
1417
+ "step": 2200
1418
+ },
1419
+ {
1420
+ "epoch": 1.42,
1421
+ "learning_rate": 0.00021775895765472312,
1422
+ "loss": 0.3262,
1423
+ "step": 2210
1424
+ },
1425
+ {
1426
+ "epoch": 1.43,
1427
+ "learning_rate": 0.00021736807817589573,
1428
+ "loss": 0.3319,
1429
+ "step": 2220
1430
+ },
1431
+ {
1432
+ "epoch": 1.43,
1433
+ "learning_rate": 0.00021697719869706838,
1434
+ "loss": 0.3422,
1435
+ "step": 2230
1436
+ },
1437
+ {
1438
+ "epoch": 1.44,
1439
+ "learning_rate": 0.000216586319218241,
1440
+ "loss": 0.3329,
1441
+ "step": 2240
1442
+ },
1443
+ {
1444
+ "epoch": 1.45,
1445
+ "learning_rate": 0.00021619543973941366,
1446
+ "loss": 0.3387,
1447
+ "step": 2250
1448
+ },
1449
+ {
1450
+ "epoch": 1.45,
1451
+ "learning_rate": 0.00021580456026058632,
1452
+ "loss": 0.3275,
1453
+ "step": 2260
1454
+ },
1455
+ {
1456
+ "epoch": 1.46,
1457
+ "learning_rate": 0.00021541368078175892,
1458
+ "loss": 0.3292,
1459
+ "step": 2270
1460
+ },
1461
+ {
1462
+ "epoch": 1.47,
1463
+ "learning_rate": 0.00021502280130293158,
1464
+ "loss": 0.3302,
1465
+ "step": 2280
1466
+ },
1467
+ {
1468
+ "epoch": 1.47,
1469
+ "learning_rate": 0.0002146319218241042,
1470
+ "loss": 0.328,
1471
+ "step": 2290
1472
+ },
1473
+ {
1474
+ "epoch": 1.48,
1475
+ "learning_rate": 0.00021424104234527686,
1476
+ "loss": 0.3239,
1477
+ "step": 2300
1478
+ },
1479
+ {
1480
+ "epoch": 1.49,
1481
+ "learning_rate": 0.00021385016286644951,
1482
+ "loss": 0.3276,
1483
+ "step": 2310
1484
+ },
1485
+ {
1486
+ "epoch": 1.49,
1487
+ "learning_rate": 0.00021345928338762212,
1488
+ "loss": 0.3307,
1489
+ "step": 2320
1490
+ },
1491
+ {
1492
+ "epoch": 1.5,
1493
+ "learning_rate": 0.00021306840390879477,
1494
+ "loss": 0.3236,
1495
+ "step": 2330
1496
+ },
1497
+ {
1498
+ "epoch": 1.5,
1499
+ "learning_rate": 0.0002126775244299674,
1500
+ "loss": 0.3362,
1501
+ "step": 2340
1502
+ },
1503
+ {
1504
+ "epoch": 1.51,
1505
+ "learning_rate": 0.00021228664495114005,
1506
+ "loss": 0.3347,
1507
+ "step": 2350
1508
+ },
1509
+ {
1510
+ "epoch": 1.52,
1511
+ "learning_rate": 0.00021189576547231268,
1512
+ "loss": 0.3261,
1513
+ "step": 2360
1514
+ },
1515
+ {
1516
+ "epoch": 1.52,
1517
+ "learning_rate": 0.0002115048859934853,
1518
+ "loss": 0.3277,
1519
+ "step": 2370
1520
+ },
1521
+ {
1522
+ "epoch": 1.53,
1523
+ "learning_rate": 0.00021111400651465794,
1524
+ "loss": 0.3235,
1525
+ "step": 2380
1526
+ },
1527
+ {
1528
+ "epoch": 1.54,
1529
+ "learning_rate": 0.0002107231270358306,
1530
+ "loss": 0.326,
1531
+ "step": 2390
1532
+ },
1533
+ {
1534
+ "epoch": 1.54,
1535
+ "learning_rate": 0.00021033224755700325,
1536
+ "loss": 0.3191,
1537
+ "step": 2400
1538
+ },
1539
+ {
1540
+ "epoch": 1.54,
1541
+ "eval_loss": 0.33296582102775574,
1542
+ "eval_runtime": 505.569,
1543
+ "eval_samples_per_second": 3.956,
1544
+ "eval_steps_per_second": 0.494,
1545
+ "step": 2400
1546
+ },
1547
+ {
1548
+ "epoch": 1.55,
1549
+ "learning_rate": 0.00020994136807817588,
1550
+ "loss": 0.3337,
1551
+ "step": 2410
1552
+ },
1553
+ {
1554
+ "epoch": 1.56,
1555
+ "learning_rate": 0.00020955048859934853,
1556
+ "loss": 0.3225,
1557
+ "step": 2420
1558
+ },
1559
+ {
1560
+ "epoch": 1.56,
1561
+ "learning_rate": 0.00020915960912052113,
1562
+ "loss": 0.3257,
1563
+ "step": 2430
1564
+ },
1565
+ {
1566
+ "epoch": 1.57,
1567
+ "learning_rate": 0.0002087687296416938,
1568
+ "loss": 0.3175,
1569
+ "step": 2440
1570
+ },
1571
+ {
1572
+ "epoch": 1.58,
1573
+ "learning_rate": 0.00020837785016286644,
1574
+ "loss": 0.3273,
1575
+ "step": 2450
1576
+ },
1577
+ {
1578
+ "epoch": 1.58,
1579
+ "learning_rate": 0.00020798697068403907,
1580
+ "loss": 0.338,
1581
+ "step": 2460
1582
+ },
1583
+ {
1584
+ "epoch": 1.59,
1585
+ "learning_rate": 0.00020759609120521173,
1586
+ "loss": 0.3197,
1587
+ "step": 2470
1588
+ },
1589
+ {
1590
+ "epoch": 1.59,
1591
+ "learning_rate": 0.00020720521172638433,
1592
+ "loss": 0.3403,
1593
+ "step": 2480
1594
+ },
1595
+ {
1596
+ "epoch": 1.6,
1597
+ "learning_rate": 0.00020681433224755698,
1598
+ "loss": 0.3224,
1599
+ "step": 2490
1600
+ },
1601
+ {
1602
+ "epoch": 1.61,
1603
+ "learning_rate": 0.0002064234527687296,
1604
+ "loss": 0.3481,
1605
+ "step": 2500
1606
+ },
1607
+ {
1608
+ "epoch": 1.61,
1609
+ "learning_rate": 0.00020603257328990227,
1610
+ "loss": 0.329,
1611
+ "step": 2510
1612
+ },
1613
+ {
1614
+ "epoch": 1.62,
1615
+ "learning_rate": 0.00020564169381107492,
1616
+ "loss": 0.3248,
1617
+ "step": 2520
1618
+ },
1619
+ {
1620
+ "epoch": 1.63,
1621
+ "learning_rate": 0.00020525081433224752,
1622
+ "loss": 0.3473,
1623
+ "step": 2530
1624
+ },
1625
+ {
1626
+ "epoch": 1.63,
1627
+ "learning_rate": 0.00020485993485342018,
1628
+ "loss": 0.3251,
1629
+ "step": 2540
1630
+ },
1631
+ {
1632
+ "epoch": 1.64,
1633
+ "learning_rate": 0.0002044690553745928,
1634
+ "loss": 0.3348,
1635
+ "step": 2550
1636
+ },
1637
+ {
1638
+ "epoch": 1.65,
1639
+ "learning_rate": 0.00020407817589576546,
1640
+ "loss": 0.3278,
1641
+ "step": 2560
1642
+ },
1643
+ {
1644
+ "epoch": 1.65,
1645
+ "learning_rate": 0.00020368729641693812,
1646
+ "loss": 0.3198,
1647
+ "step": 2570
1648
+ },
1649
+ {
1650
+ "epoch": 1.66,
1651
+ "learning_rate": 0.00020329641693811072,
1652
+ "loss": 0.3352,
1653
+ "step": 2580
1654
+ },
1655
+ {
1656
+ "epoch": 1.67,
1657
+ "learning_rate": 0.00020290553745928337,
1658
+ "loss": 0.3194,
1659
+ "step": 2590
1660
+ },
1661
+ {
1662
+ "epoch": 1.67,
1663
+ "learning_rate": 0.000202514657980456,
1664
+ "loss": 0.3277,
1665
+ "step": 2600
1666
+ },
1667
+ {
1668
+ "epoch": 1.67,
1669
+ "eval_loss": 0.33181408047676086,
1670
+ "eval_runtime": 506.9358,
1671
+ "eval_samples_per_second": 3.945,
1672
+ "eval_steps_per_second": 0.493,
1673
+ "step": 2600
1674
+ },
1675
+ {
1676
+ "epoch": 1.68,
1677
+ "learning_rate": 0.00020212377850162866,
1678
+ "loss": 0.3336,
1679
+ "step": 2610
1680
+ },
1681
+ {
1682
+ "epoch": 1.68,
1683
+ "learning_rate": 0.00020173289902280128,
1684
+ "loss": 0.3202,
1685
+ "step": 2620
1686
+ },
1687
+ {
1688
+ "epoch": 1.69,
1689
+ "learning_rate": 0.0002013420195439739,
1690
+ "loss": 0.3238,
1691
+ "step": 2630
1692
+ },
1693
+ {
1694
+ "epoch": 1.7,
1695
+ "learning_rate": 0.00020095114006514654,
1696
+ "loss": 0.3208,
1697
+ "step": 2640
1698
+ },
1699
+ {
1700
+ "epoch": 1.7,
1701
+ "learning_rate": 0.0002005602605863192,
1702
+ "loss": 0.3162,
1703
+ "step": 2650
1704
+ },
1705
+ {
1706
+ "epoch": 1.71,
1707
+ "learning_rate": 0.00020016938110749185,
1708
+ "loss": 0.3177,
1709
+ "step": 2660
1710
+ },
1711
+ {
1712
+ "epoch": 1.72,
1713
+ "learning_rate": 0.00019977850162866448,
1714
+ "loss": 0.3202,
1715
+ "step": 2670
1716
+ },
1717
+ {
1718
+ "epoch": 1.72,
1719
+ "learning_rate": 0.0001993876221498371,
1720
+ "loss": 0.3348,
1721
+ "step": 2680
1722
+ },
1723
+ {
1724
+ "epoch": 1.73,
1725
+ "learning_rate": 0.00019899674267100974,
1726
+ "loss": 0.3238,
1727
+ "step": 2690
1728
+ },
1729
+ {
1730
+ "epoch": 1.74,
1731
+ "learning_rate": 0.0001986058631921824,
1732
+ "loss": 0.3271,
1733
+ "step": 2700
1734
+ },
1735
+ {
1736
+ "epoch": 1.74,
1737
+ "learning_rate": 0.00019821498371335505,
1738
+ "loss": 0.3176,
1739
+ "step": 2710
1740
+ },
1741
+ {
1742
+ "epoch": 1.75,
1743
+ "learning_rate": 0.00019782410423452767,
1744
+ "loss": 0.3327,
1745
+ "step": 2720
1746
+ },
1747
+ {
1748
+ "epoch": 1.76,
1749
+ "learning_rate": 0.0001974332247557003,
1750
+ "loss": 0.3329,
1751
+ "step": 2730
1752
+ },
1753
+ {
1754
+ "epoch": 1.76,
1755
+ "learning_rate": 0.00019704234527687293,
1756
+ "loss": 0.3261,
1757
+ "step": 2740
1758
+ },
1759
+ {
1760
+ "epoch": 1.77,
1761
+ "learning_rate": 0.00019665146579804559,
1762
+ "loss": 0.325,
1763
+ "step": 2750
1764
+ },
1765
+ {
1766
+ "epoch": 1.77,
1767
+ "learning_rate": 0.00019626058631921821,
1768
+ "loss": 0.3122,
1769
+ "step": 2760
1770
+ },
1771
+ {
1772
+ "epoch": 1.78,
1773
+ "learning_rate": 0.00019586970684039087,
1774
+ "loss": 0.3323,
1775
+ "step": 2770
1776
+ },
1777
+ {
1778
+ "epoch": 1.79,
1779
+ "learning_rate": 0.00019547882736156352,
1780
+ "loss": 0.3307,
1781
+ "step": 2780
1782
+ },
1783
+ {
1784
+ "epoch": 1.79,
1785
+ "learning_rate": 0.00019508794788273613,
1786
+ "loss": 0.329,
1787
+ "step": 2790
1788
+ },
1789
+ {
1790
+ "epoch": 1.8,
1791
+ "learning_rate": 0.00019469706840390878,
1792
+ "loss": 0.3286,
1793
+ "step": 2800
1794
+ },
1795
+ {
1796
+ "epoch": 1.8,
1797
+ "eval_loss": 0.330875426530838,
1798
+ "eval_runtime": 512.0455,
1799
+ "eval_samples_per_second": 3.906,
1800
+ "eval_steps_per_second": 0.488,
1801
+ "step": 2800
1802
+ },
1803
+ {
1804
+ "epoch": 1.81,
1805
+ "learning_rate": 0.0001943061889250814,
1806
+ "loss": 0.3155,
1807
+ "step": 2810
1808
+ },
1809
+ {
1810
+ "epoch": 1.81,
1811
+ "learning_rate": 0.00019391530944625406,
1812
+ "loss": 0.3291,
1813
+ "step": 2820
1814
+ },
1815
+ {
1816
+ "epoch": 1.82,
1817
+ "learning_rate": 0.00019352442996742672,
1818
+ "loss": 0.3229,
1819
+ "step": 2830
1820
+ },
1821
+ {
1822
+ "epoch": 1.83,
1823
+ "learning_rate": 0.00019313355048859932,
1824
+ "loss": 0.3298,
1825
+ "step": 2840
1826
+ },
1827
+ {
1828
+ "epoch": 1.83,
1829
+ "learning_rate": 0.00019274267100977198,
1830
+ "loss": 0.31,
1831
+ "step": 2850
1832
+ },
1833
+ {
1834
+ "epoch": 1.84,
1835
+ "learning_rate": 0.0001923517915309446,
1836
+ "loss": 0.3177,
1837
+ "step": 2860
1838
+ },
1839
+ {
1840
+ "epoch": 1.85,
1841
+ "learning_rate": 0.00019196091205211726,
1842
+ "loss": 0.3175,
1843
+ "step": 2870
1844
+ },
1845
+ {
1846
+ "epoch": 1.85,
1847
+ "learning_rate": 0.0001915700325732899,
1848
+ "loss": 0.3321,
1849
+ "step": 2880
1850
+ },
1851
+ {
1852
+ "epoch": 1.86,
1853
+ "learning_rate": 0.00019117915309446252,
1854
+ "loss": 0.3257,
1855
+ "step": 2890
1856
+ },
1857
+ {
1858
+ "epoch": 1.86,
1859
+ "learning_rate": 0.00019078827361563514,
1860
+ "loss": 0.324,
1861
+ "step": 2900
1862
+ },
1863
+ {
1864
+ "epoch": 1.87,
1865
+ "learning_rate": 0.0001903973941368078,
1866
+ "loss": 0.3202,
1867
+ "step": 2910
1868
+ },
1869
+ {
1870
+ "epoch": 1.88,
1871
+ "learning_rate": 0.00019000651465798045,
1872
+ "loss": 0.3206,
1873
+ "step": 2920
1874
+ },
1875
+ {
1876
+ "epoch": 1.88,
1877
+ "learning_rate": 0.00018961563517915308,
1878
+ "loss": 0.3332,
1879
+ "step": 2930
1880
+ },
1881
+ {
1882
+ "epoch": 1.89,
1883
+ "learning_rate": 0.0001892247557003257,
1884
+ "loss": 0.3246,
1885
+ "step": 2940
1886
+ },
1887
+ {
1888
+ "epoch": 1.9,
1889
+ "learning_rate": 0.00018883387622149834,
1890
+ "loss": 0.3194,
1891
+ "step": 2950
1892
+ },
1893
+ {
1894
+ "epoch": 1.9,
1895
+ "learning_rate": 0.000188442996742671,
1896
+ "loss": 0.3204,
1897
+ "step": 2960
1898
+ },
1899
+ {
1900
+ "epoch": 1.91,
1901
+ "learning_rate": 0.00018805211726384365,
1902
+ "loss": 0.3201,
1903
+ "step": 2970
1904
+ },
1905
+ {
1906
+ "epoch": 1.92,
1907
+ "learning_rate": 0.00018766123778501628,
1908
+ "loss": 0.3219,
1909
+ "step": 2980
1910
+ },
1911
+ {
1912
+ "epoch": 1.92,
1913
+ "learning_rate": 0.0001872703583061889,
1914
+ "loss": 0.3104,
1915
+ "step": 2990
1916
+ },
1917
+ {
1918
+ "epoch": 1.93,
1919
+ "learning_rate": 0.00018687947882736153,
1920
+ "loss": 0.3257,
1921
+ "step": 3000
1922
+ },
1923
+ {
1924
+ "epoch": 1.93,
1925
+ "eval_loss": 0.329958438873291,
1926
+ "eval_runtime": 514.7505,
1927
+ "eval_samples_per_second": 3.885,
1928
+ "eval_steps_per_second": 0.486,
1929
+ "step": 3000
1930
+ },
1931
+ {
1932
+ "epoch": 1.94,
1933
+ "learning_rate": 0.0001864885993485342,
1934
+ "loss": 0.3238,
1935
+ "step": 3010
1936
+ },
1937
+ {
1938
+ "epoch": 1.94,
1939
+ "learning_rate": 0.00018609771986970682,
1940
+ "loss": 0.3261,
1941
+ "step": 3020
1942
+ },
1943
+ {
1944
+ "epoch": 1.95,
1945
+ "learning_rate": 0.00018570684039087947,
1946
+ "loss": 0.3165,
1947
+ "step": 3030
1948
+ },
1949
+ {
1950
+ "epoch": 1.95,
1951
+ "learning_rate": 0.00018531596091205207,
1952
+ "loss": 0.3236,
1953
+ "step": 3040
1954
+ },
1955
+ {
1956
+ "epoch": 1.96,
1957
+ "learning_rate": 0.00018492508143322473,
1958
+ "loss": 0.3169,
1959
+ "step": 3050
1960
+ },
1961
+ {
1962
+ "epoch": 1.97,
1963
+ "learning_rate": 0.00018453420195439738,
1964
+ "loss": 0.3306,
1965
+ "step": 3060
1966
+ },
1967
+ {
1968
+ "epoch": 1.97,
1969
+ "learning_rate": 0.00018414332247557,
1970
+ "loss": 0.3246,
1971
+ "step": 3070
1972
+ },
1973
+ {
1974
+ "epoch": 1.98,
1975
+ "learning_rate": 0.00018375244299674267,
1976
+ "loss": 0.321,
1977
+ "step": 3080
1978
+ },
1979
+ {
1980
+ "epoch": 1.99,
1981
+ "learning_rate": 0.00018336156351791527,
1982
+ "loss": 0.3167,
1983
+ "step": 3090
1984
+ },
1985
+ {
1986
+ "epoch": 1.99,
1987
+ "learning_rate": 0.00018297068403908792,
1988
+ "loss": 0.3123,
1989
+ "step": 3100
1990
+ },
1991
+ {
1992
+ "epoch": 2.0,
1993
+ "learning_rate": 0.00018257980456026058,
1994
+ "loss": 0.3332,
1995
+ "step": 3110
1996
+ },
1997
+ {
1998
+ "epoch": 2.01,
1999
+ "learning_rate": 0.0001821889250814332,
2000
+ "loss": 0.3123,
2001
+ "step": 3120
2002
+ },
2003
+ {
2004
+ "epoch": 2.01,
2005
+ "learning_rate": 0.00018179804560260586,
2006
+ "loss": 0.3218,
2007
+ "step": 3130
2008
+ },
2009
+ {
2010
+ "epoch": 2.02,
2011
+ "learning_rate": 0.0001814071661237785,
2012
+ "loss": 0.3068,
2013
+ "step": 3140
2014
+ },
2015
+ {
2016
+ "epoch": 2.03,
2017
+ "learning_rate": 0.00018101628664495112,
2018
+ "loss": 0.3137,
2019
+ "step": 3150
2020
+ },
2021
+ {
2022
+ "epoch": 2.03,
2023
+ "learning_rate": 0.00018062540716612375,
2024
+ "loss": 0.3183,
2025
+ "step": 3160
2026
+ },
2027
+ {
2028
+ "epoch": 2.04,
2029
+ "learning_rate": 0.0001802345276872964,
2030
+ "loss": 0.3125,
2031
+ "step": 3170
2032
+ },
2033
+ {
2034
+ "epoch": 2.05,
2035
+ "learning_rate": 0.00017984364820846906,
2036
+ "loss": 0.3037,
2037
+ "step": 3180
2038
+ },
2039
+ {
2040
+ "epoch": 2.05,
2041
+ "learning_rate": 0.00017945276872964169,
2042
+ "loss": 0.3224,
2043
+ "step": 3190
2044
+ },
2045
+ {
2046
+ "epoch": 2.06,
2047
+ "learning_rate": 0.0001790618892508143,
2048
+ "loss": 0.3168,
2049
+ "step": 3200
2050
+ },
2051
+ {
2052
+ "epoch": 2.06,
2053
+ "eval_loss": 0.3301369845867157,
2054
+ "eval_runtime": 512.6559,
2055
+ "eval_samples_per_second": 3.901,
2056
+ "eval_steps_per_second": 0.488,
2057
+ "step": 3200
2058
+ },
2059
+ {
2060
+ "epoch": 2.06,
2061
+ "learning_rate": 0.00017867100977198694,
2062
+ "loss": 0.3127,
2063
+ "step": 3210
2064
+ },
2065
+ {
2066
+ "epoch": 2.07,
2067
+ "learning_rate": 0.0001782801302931596,
2068
+ "loss": 0.3152,
2069
+ "step": 3220
2070
+ },
2071
+ {
2072
+ "epoch": 2.08,
2073
+ "learning_rate": 0.00017788925081433225,
2074
+ "loss": 0.3092,
2075
+ "step": 3230
2076
+ },
2077
+ {
2078
+ "epoch": 2.08,
2079
+ "learning_rate": 0.00017749837133550488,
2080
+ "loss": 0.2967,
2081
+ "step": 3240
2082
+ },
2083
+ {
2084
+ "epoch": 2.09,
2085
+ "learning_rate": 0.0001771074918566775,
2086
+ "loss": 0.3182,
2087
+ "step": 3250
2088
+ },
2089
+ {
2090
+ "epoch": 2.1,
2091
+ "learning_rate": 0.00017671661237785014,
2092
+ "loss": 0.2959,
2093
+ "step": 3260
2094
+ },
2095
+ {
2096
+ "epoch": 2.1,
2097
+ "learning_rate": 0.0001763257328990228,
2098
+ "loss": 0.3035,
2099
+ "step": 3270
2100
+ },
2101
+ {
2102
+ "epoch": 2.11,
2103
+ "learning_rate": 0.00017593485342019542,
2104
+ "loss": 0.3236,
2105
+ "step": 3280
2106
+ },
2107
+ {
2108
+ "epoch": 2.12,
2109
+ "learning_rate": 0.00017554397394136808,
2110
+ "loss": 0.3117,
2111
+ "step": 3290
2112
+ },
2113
+ {
2114
+ "epoch": 2.12,
2115
+ "learning_rate": 0.00017515309446254068,
2116
+ "loss": 0.3064,
2117
+ "step": 3300
2118
+ },
2119
+ {
2120
+ "epoch": 2.13,
2121
+ "learning_rate": 0.00017476221498371333,
2122
+ "loss": 0.2977,
2123
+ "step": 3310
2124
+ },
2125
+ {
2126
+ "epoch": 2.14,
2127
+ "learning_rate": 0.000174371335504886,
2128
+ "loss": 0.3179,
2129
+ "step": 3320
2130
+ },
2131
+ {
2132
+ "epoch": 2.14,
2133
+ "learning_rate": 0.00017398045602605861,
2134
+ "loss": 0.307,
2135
+ "step": 3330
2136
+ },
2137
+ {
2138
+ "epoch": 2.15,
2139
+ "learning_rate": 0.00017358957654723127,
2140
+ "loss": 0.3086,
2141
+ "step": 3340
2142
+ },
2143
+ {
2144
+ "epoch": 2.15,
2145
+ "learning_rate": 0.00017319869706840387,
2146
+ "loss": 0.3116,
2147
+ "step": 3350
2148
+ },
2149
+ {
2150
+ "epoch": 2.16,
2151
+ "learning_rate": 0.00017280781758957653,
2152
+ "loss": 0.3071,
2153
+ "step": 3360
2154
+ },
2155
+ {
2156
+ "epoch": 2.17,
2157
+ "learning_rate": 0.00017241693811074918,
2158
+ "loss": 0.3168,
2159
+ "step": 3370
2160
+ },
2161
+ {
2162
+ "epoch": 2.17,
2163
+ "learning_rate": 0.0001720260586319218,
2164
+ "loss": 0.318,
2165
+ "step": 3380
2166
+ },
2167
+ {
2168
+ "epoch": 2.18,
2169
+ "learning_rate": 0.00017163517915309446,
2170
+ "loss": 0.3139,
2171
+ "step": 3390
2172
+ },
2173
+ {
2174
+ "epoch": 2.19,
2175
+ "learning_rate": 0.00017124429967426707,
2176
+ "loss": 0.3118,
2177
+ "step": 3400
2178
+ },
2179
+ {
2180
+ "epoch": 2.19,
2181
+ "eval_loss": 0.3299133777618408,
2182
+ "eval_runtime": 515.2806,
2183
+ "eval_samples_per_second": 3.881,
2184
+ "eval_steps_per_second": 0.485,
2185
+ "step": 3400
2186
+ },
2187
+ {
2188
+ "epoch": 2.19,
2189
+ "learning_rate": 0.00017085342019543972,
2190
+ "loss": 0.3128,
2191
+ "step": 3410
2192
+ },
2193
+ {
2194
+ "epoch": 2.2,
2195
+ "learning_rate": 0.00017046254071661235,
2196
+ "loss": 0.3067,
2197
+ "step": 3420
2198
+ },
2199
+ {
2200
+ "epoch": 2.21,
2201
+ "learning_rate": 0.000170071661237785,
2202
+ "loss": 0.2983,
2203
+ "step": 3430
2204
+ },
2205
+ {
2206
+ "epoch": 2.21,
2207
+ "learning_rate": 0.00016968078175895766,
2208
+ "loss": 0.3045,
2209
+ "step": 3440
2210
+ },
2211
+ {
2212
+ "epoch": 2.22,
2213
+ "learning_rate": 0.00016928990228013026,
2214
+ "loss": 0.3199,
2215
+ "step": 3450
2216
+ },
2217
+ {
2218
+ "epoch": 2.23,
2219
+ "learning_rate": 0.00016889902280130292,
2220
+ "loss": 0.313,
2221
+ "step": 3460
2222
+ },
2223
+ {
2224
+ "epoch": 2.23,
2225
+ "learning_rate": 0.00016850814332247554,
2226
+ "loss": 0.3191,
2227
+ "step": 3470
2228
+ },
2229
+ {
2230
+ "epoch": 2.24,
2231
+ "learning_rate": 0.0001681172638436482,
2232
+ "loss": 0.3083,
2233
+ "step": 3480
2234
+ },
2235
+ {
2236
+ "epoch": 2.24,
2237
+ "learning_rate": 0.00016772638436482085,
2238
+ "loss": 0.307,
2239
+ "step": 3490
2240
+ },
2241
+ {
2242
+ "epoch": 2.25,
2243
+ "learning_rate": 0.00016733550488599346,
2244
+ "loss": 0.3058,
2245
+ "step": 3500
2246
+ },
2247
+ {
2248
+ "epoch": 2.26,
2249
+ "learning_rate": 0.0001669446254071661,
2250
+ "loss": 0.3062,
2251
+ "step": 3510
2252
+ },
2253
+ {
2254
+ "epoch": 2.26,
2255
+ "learning_rate": 0.00016655374592833874,
2256
+ "loss": 0.3221,
2257
+ "step": 3520
2258
+ },
2259
+ {
2260
+ "epoch": 2.27,
2261
+ "learning_rate": 0.0001661628664495114,
2262
+ "loss": 0.3058,
2263
+ "step": 3530
2264
+ },
2265
+ {
2266
+ "epoch": 2.28,
2267
+ "learning_rate": 0.00016577198697068402,
2268
+ "loss": 0.3086,
2269
+ "step": 3540
2270
+ },
2271
+ {
2272
+ "epoch": 2.28,
2273
+ "learning_rate": 0.00016538110749185668,
2274
+ "loss": 0.3154,
2275
+ "step": 3550
2276
+ },
2277
+ {
2278
+ "epoch": 2.29,
2279
+ "learning_rate": 0.00016499022801302928,
2280
+ "loss": 0.3127,
2281
+ "step": 3560
2282
+ },
2283
+ {
2284
+ "epoch": 2.3,
2285
+ "learning_rate": 0.00016459934853420193,
2286
+ "loss": 0.3071,
2287
+ "step": 3570
2288
+ },
2289
+ {
2290
+ "epoch": 2.3,
2291
+ "learning_rate": 0.0001642084690553746,
2292
+ "loss": 0.3044,
2293
+ "step": 3580
2294
+ },
2295
+ {
2296
+ "epoch": 2.31,
2297
+ "learning_rate": 0.00016381758957654722,
2298
+ "loss": 0.3107,
2299
+ "step": 3590
2300
+ },
2301
+ {
2302
+ "epoch": 2.32,
2303
+ "learning_rate": 0.00016342671009771987,
2304
+ "loss": 0.3058,
2305
+ "step": 3600
2306
+ },
2307
+ {
2308
+ "epoch": 2.32,
2309
+ "eval_loss": 0.32923969626426697,
2310
+ "eval_runtime": 511.8058,
2311
+ "eval_samples_per_second": 3.908,
2312
+ "eval_steps_per_second": 0.488,
2313
+ "step": 3600
2314
+ },
2315
+ {
2316
+ "epoch": 2.32,
2317
+ "learning_rate": 0.00016303583061889247,
2318
+ "loss": 0.3205,
2319
+ "step": 3610
2320
+ },
2321
+ {
2322
+ "epoch": 2.33,
2323
+ "learning_rate": 0.00016264495114006513,
2324
+ "loss": 0.3032,
2325
+ "step": 3620
2326
+ },
2327
+ {
2328
+ "epoch": 2.33,
2329
+ "learning_rate": 0.00016225407166123778,
2330
+ "loss": 0.3153,
2331
+ "step": 3630
2332
+ },
2333
+ {
2334
+ "epoch": 2.34,
2335
+ "learning_rate": 0.0001618631921824104,
2336
+ "loss": 0.3126,
2337
+ "step": 3640
2338
+ },
2339
+ {
2340
+ "epoch": 2.35,
2341
+ "learning_rate": 0.00016147231270358307,
2342
+ "loss": 0.3115,
2343
+ "step": 3650
2344
+ },
2345
+ {
2346
+ "epoch": 2.35,
2347
+ "learning_rate": 0.00016108143322475567,
2348
+ "loss": 0.3085,
2349
+ "step": 3660
2350
+ },
2351
+ {
2352
+ "epoch": 2.36,
2353
+ "learning_rate": 0.00016069055374592832,
2354
+ "loss": 0.3138,
2355
+ "step": 3670
2356
+ },
2357
+ {
2358
+ "epoch": 2.37,
2359
+ "learning_rate": 0.00016029967426710095,
2360
+ "loss": 0.3101,
2361
+ "step": 3680
2362
+ },
2363
+ {
2364
+ "epoch": 2.37,
2365
+ "learning_rate": 0.0001599087947882736,
2366
+ "loss": 0.3098,
2367
+ "step": 3690
2368
+ },
2369
+ {
2370
+ "epoch": 2.38,
2371
+ "learning_rate": 0.00015951791530944626,
2372
+ "loss": 0.2962,
2373
+ "step": 3700
2374
+ },
2375
+ {
2376
+ "epoch": 2.39,
2377
+ "learning_rate": 0.00015912703583061886,
2378
+ "loss": 0.3207,
2379
+ "step": 3710
2380
+ },
2381
+ {
2382
+ "epoch": 2.39,
2383
+ "learning_rate": 0.00015873615635179152,
2384
+ "loss": 0.3181,
2385
+ "step": 3720
2386
+ },
2387
+ {
2388
+ "epoch": 2.4,
2389
+ "learning_rate": 0.00015834527687296415,
2390
+ "loss": 0.3057,
2391
+ "step": 3730
2392
+ },
2393
+ {
2394
+ "epoch": 2.41,
2395
+ "learning_rate": 0.0001579543973941368,
2396
+ "loss": 0.3103,
2397
+ "step": 3740
2398
+ },
2399
+ {
2400
+ "epoch": 2.41,
2401
+ "learning_rate": 0.00015756351791530946,
2402
+ "loss": 0.3115,
2403
+ "step": 3750
2404
+ },
2405
+ {
2406
+ "epoch": 2.42,
2407
+ "learning_rate": 0.00015717263843648206,
2408
+ "loss": 0.3102,
2409
+ "step": 3760
2410
+ },
2411
+ {
2412
+ "epoch": 2.42,
2413
+ "learning_rate": 0.00015678175895765471,
2414
+ "loss": 0.3051,
2415
+ "step": 3770
2416
+ },
2417
+ {
2418
+ "epoch": 2.43,
2419
+ "learning_rate": 0.00015639087947882734,
2420
+ "loss": 0.3159,
2421
+ "step": 3780
2422
+ },
2423
+ {
2424
+ "epoch": 2.44,
2425
+ "learning_rate": 0.000156,
2426
+ "loss": 0.2957,
2427
+ "step": 3790
2428
+ },
2429
+ {
2430
+ "epoch": 2.44,
2431
+ "learning_rate": 0.00015560912052117263,
2432
+ "loss": 0.3062,
2433
+ "step": 3800
2434
+ },
2435
+ {
2436
+ "epoch": 2.44,
2437
+ "eval_loss": 0.32817041873931885,
2438
+ "eval_runtime": 517.8112,
2439
+ "eval_samples_per_second": 3.862,
2440
+ "eval_steps_per_second": 0.483,
2441
+ "step": 3800
2442
+ },
2443
+ {
2444
+ "epoch": 2.45,
2445
+ "learning_rate": 0.00015521824104234525,
2446
+ "loss": 0.3148,
2447
+ "step": 3810
2448
+ },
2449
+ {
2450
+ "epoch": 2.46,
2451
+ "learning_rate": 0.00015482736156351788,
2452
+ "loss": 0.2992,
2453
+ "step": 3820
2454
+ },
2455
+ {
2456
+ "epoch": 2.46,
2457
+ "learning_rate": 0.00015443648208469054,
2458
+ "loss": 0.3056,
2459
+ "step": 3830
2460
+ },
2461
+ {
2462
+ "epoch": 2.47,
2463
+ "learning_rate": 0.0001540456026058632,
2464
+ "loss": 0.3079,
2465
+ "step": 3840
2466
+ },
2467
+ {
2468
+ "epoch": 2.48,
2469
+ "learning_rate": 0.00015365472312703582,
2470
+ "loss": 0.3103,
2471
+ "step": 3850
2472
+ },
2473
+ {
2474
+ "epoch": 2.48,
2475
+ "learning_rate": 0.00015326384364820845,
2476
+ "loss": 0.3094,
2477
+ "step": 3860
2478
+ },
2479
+ {
2480
+ "epoch": 2.49,
2481
+ "learning_rate": 0.00015287296416938108,
2482
+ "loss": 0.3218,
2483
+ "step": 3870
2484
+ },
2485
+ {
2486
+ "epoch": 2.5,
2487
+ "learning_rate": 0.00015248208469055373,
2488
+ "loss": 0.3143,
2489
+ "step": 3880
2490
+ },
2491
+ {
2492
+ "epoch": 2.5,
2493
+ "learning_rate": 0.0001520912052117264,
2494
+ "loss": 0.3076,
2495
+ "step": 3890
2496
+ },
2497
+ {
2498
+ "epoch": 2.51,
2499
+ "learning_rate": 0.00015170032573289902,
2500
+ "loss": 0.312,
2501
+ "step": 3900
2502
+ },
2503
+ {
2504
+ "epoch": 2.51,
2505
+ "learning_rate": 0.00015130944625407167,
2506
+ "loss": 0.3068,
2507
+ "step": 3910
2508
+ },
2509
+ {
2510
+ "epoch": 2.52,
2511
+ "learning_rate": 0.00015091856677524427,
2512
+ "loss": 0.3112,
2513
+ "step": 3920
2514
+ },
2515
+ {
2516
+ "epoch": 2.53,
2517
+ "learning_rate": 0.00015052768729641693,
2518
+ "loss": 0.3134,
2519
+ "step": 3930
2520
+ },
2521
+ {
2522
+ "epoch": 2.53,
2523
+ "learning_rate": 0.00015013680781758955,
2524
+ "loss": 0.3174,
2525
+ "step": 3940
2526
+ },
2527
+ {
2528
+ "epoch": 2.54,
2529
+ "learning_rate": 0.00014974592833876218,
2530
+ "loss": 0.3226,
2531
+ "step": 3950
2532
+ },
2533
+ {
2534
+ "epoch": 2.55,
2535
+ "learning_rate": 0.00014935504885993484,
2536
+ "loss": 0.3025,
2537
+ "step": 3960
2538
+ },
2539
+ {
2540
+ "epoch": 2.55,
2541
+ "learning_rate": 0.0001489641693811075,
2542
+ "loss": 0.3141,
2543
+ "step": 3970
2544
+ },
2545
+ {
2546
+ "epoch": 2.56,
2547
+ "learning_rate": 0.00014857328990228012,
2548
+ "loss": 0.3137,
2549
+ "step": 3980
2550
+ },
2551
+ {
2552
+ "epoch": 2.57,
2553
+ "learning_rate": 0.00014818241042345275,
2554
+ "loss": 0.306,
2555
+ "step": 3990
2556
+ },
2557
+ {
2558
+ "epoch": 2.57,
2559
+ "learning_rate": 0.00014779153094462538,
2560
+ "loss": 0.3108,
2561
+ "step": 4000
2562
+ },
2563
+ {
2564
+ "epoch": 2.57,
2565
+ "eval_loss": 0.32780957221984863,
2566
+ "eval_runtime": 512.0443,
2567
+ "eval_samples_per_second": 3.906,
2568
+ "eval_steps_per_second": 0.488,
2569
+ "step": 4000
2570
+ },
2571
+ {
2572
+ "epoch": 2.58,
2573
+ "learning_rate": 0.00014740065146579803,
2574
+ "loss": 0.3056,
2575
+ "step": 4010
2576
+ },
2577
+ {
2578
+ "epoch": 2.59,
2579
+ "learning_rate": 0.0001470097719869707,
2580
+ "loss": 0.3011,
2581
+ "step": 4020
2582
+ },
2583
+ {
2584
+ "epoch": 2.59,
2585
+ "learning_rate": 0.00014661889250814332,
2586
+ "loss": 0.3128,
2587
+ "step": 4030
2588
+ },
2589
+ {
2590
+ "epoch": 2.6,
2591
+ "learning_rate": 0.00014622801302931594,
2592
+ "loss": 0.3177,
2593
+ "step": 4040
2594
+ },
2595
+ {
2596
+ "epoch": 2.6,
2597
+ "learning_rate": 0.0001458371335504886,
2598
+ "loss": 0.3129,
2599
+ "step": 4050
2600
+ },
2601
+ {
2602
+ "epoch": 2.61,
2603
+ "learning_rate": 0.00014544625407166123,
2604
+ "loss": 0.3042,
2605
+ "step": 4060
2606
+ },
2607
+ {
2608
+ "epoch": 2.62,
2609
+ "learning_rate": 0.00014505537459283386,
2610
+ "loss": 0.3062,
2611
+ "step": 4070
2612
+ },
2613
+ {
2614
+ "epoch": 2.62,
2615
+ "learning_rate": 0.00014466449511400648,
2616
+ "loss": 0.332,
2617
+ "step": 4080
2618
+ },
2619
+ {
2620
+ "epoch": 2.63,
2621
+ "learning_rate": 0.00014427361563517914,
2622
+ "loss": 0.3044,
2623
+ "step": 4090
2624
+ },
2625
+ {
2626
+ "epoch": 2.64,
2627
+ "learning_rate": 0.0001438827361563518,
2628
+ "loss": 0.3168,
2629
+ "step": 4100
2630
+ },
2631
+ {
2632
+ "epoch": 2.64,
2633
+ "learning_rate": 0.00014349185667752442,
2634
+ "loss": 0.3056,
2635
+ "step": 4110
2636
+ },
2637
+ {
2638
+ "epoch": 2.65,
2639
+ "learning_rate": 0.00014310097719869705,
2640
+ "loss": 0.3163,
2641
+ "step": 4120
2642
+ },
2643
+ {
2644
+ "epoch": 2.66,
2645
+ "learning_rate": 0.00014271009771986968,
2646
+ "loss": 0.3163,
2647
+ "step": 4130
2648
+ },
2649
+ {
2650
+ "epoch": 2.66,
2651
+ "learning_rate": 0.00014231921824104233,
2652
+ "loss": 0.3046,
2653
+ "step": 4140
2654
+ },
2655
+ {
2656
+ "epoch": 2.67,
2657
+ "learning_rate": 0.000141928338762215,
2658
+ "loss": 0.326,
2659
+ "step": 4150
2660
+ },
2661
+ {
2662
+ "epoch": 2.68,
2663
+ "learning_rate": 0.00014153745928338762,
2664
+ "loss": 0.3056,
2665
+ "step": 4160
2666
+ },
2667
+ {
2668
+ "epoch": 2.68,
2669
+ "learning_rate": 0.00014114657980456025,
2670
+ "loss": 0.3175,
2671
+ "step": 4170
2672
+ },
2673
+ {
2674
+ "epoch": 2.69,
2675
+ "learning_rate": 0.00014075570032573287,
2676
+ "loss": 0.3081,
2677
+ "step": 4180
2678
+ },
2679
+ {
2680
+ "epoch": 2.69,
2681
+ "learning_rate": 0.00014036482084690553,
2682
+ "loss": 0.3127,
2683
+ "step": 4190
2684
+ },
2685
+ {
2686
+ "epoch": 2.7,
2687
+ "learning_rate": 0.00013997394136807816,
2688
+ "loss": 0.3214,
2689
+ "step": 4200
2690
+ },
2691
+ {
2692
+ "epoch": 2.7,
2693
+ "eval_loss": 0.32745644450187683,
2694
+ "eval_runtime": 516.9052,
2695
+ "eval_samples_per_second": 3.869,
2696
+ "eval_steps_per_second": 0.484,
2697
+ "step": 4200
2698
+ },
2699
+ {
2700
+ "epoch": 2.71,
2701
+ "learning_rate": 0.00013958306188925079,
2702
+ "loss": 0.2944,
2703
+ "step": 4210
2704
+ },
2705
+ {
2706
+ "epoch": 2.71,
2707
+ "learning_rate": 0.00013919218241042344,
2708
+ "loss": 0.307,
2709
+ "step": 4220
2710
+ },
2711
+ {
2712
+ "epoch": 2.72,
2713
+ "learning_rate": 0.0001388013029315961,
2714
+ "loss": 0.2973,
2715
+ "step": 4230
2716
+ },
2717
+ {
2718
+ "epoch": 2.73,
2719
+ "learning_rate": 0.00013841042345276872,
2720
+ "loss": 0.3165,
2721
+ "step": 4240
2722
+ },
2723
+ {
2724
+ "epoch": 2.73,
2725
+ "learning_rate": 0.00013801954397394135,
2726
+ "loss": 0.3247,
2727
+ "step": 4250
2728
+ },
2729
+ {
2730
+ "epoch": 2.74,
2731
+ "learning_rate": 0.00013762866449511398,
2732
+ "loss": 0.305,
2733
+ "step": 4260
2734
+ },
2735
+ {
2736
+ "epoch": 2.75,
2737
+ "learning_rate": 0.00013723778501628664,
2738
+ "loss": 0.307,
2739
+ "step": 4270
2740
+ },
2741
+ {
2742
+ "epoch": 2.75,
2743
+ "learning_rate": 0.0001368469055374593,
2744
+ "loss": 0.3047,
2745
+ "step": 4280
2746
+ },
2747
+ {
2748
+ "epoch": 2.76,
2749
+ "learning_rate": 0.00013645602605863192,
2750
+ "loss": 0.2988,
2751
+ "step": 4290
2752
+ },
2753
+ {
2754
+ "epoch": 2.77,
2755
+ "learning_rate": 0.00013606514657980455,
2756
+ "loss": 0.322,
2757
+ "step": 4300
2758
+ },
2759
+ {
2760
+ "epoch": 2.77,
2761
+ "learning_rate": 0.00013567426710097718,
2762
+ "loss": 0.3092,
2763
+ "step": 4310
2764
+ },
2765
+ {
2766
+ "epoch": 2.78,
2767
+ "learning_rate": 0.00013528338762214983,
2768
+ "loss": 0.3117,
2769
+ "step": 4320
2770
+ },
2771
+ {
2772
+ "epoch": 2.78,
2773
+ "learning_rate": 0.00013489250814332246,
2774
+ "loss": 0.3067,
2775
+ "step": 4330
2776
+ },
2777
+ {
2778
+ "epoch": 2.79,
2779
+ "learning_rate": 0.0001345016286644951,
2780
+ "loss": 0.3153,
2781
+ "step": 4340
2782
+ },
2783
+ {
2784
+ "epoch": 2.8,
2785
+ "learning_rate": 0.00013411074918566774,
2786
+ "loss": 0.3105,
2787
+ "step": 4350
2788
+ },
2789
+ {
2790
+ "epoch": 2.8,
2791
+ "learning_rate": 0.00013371986970684037,
2792
+ "loss": 0.3141,
2793
+ "step": 4360
2794
+ },
2795
+ {
2796
+ "epoch": 2.81,
2797
+ "learning_rate": 0.00013332899022801303,
2798
+ "loss": 0.3198,
2799
+ "step": 4370
2800
+ },
2801
+ {
2802
+ "epoch": 2.82,
2803
+ "learning_rate": 0.00013293811074918565,
2804
+ "loss": 0.3069,
2805
+ "step": 4380
2806
+ },
2807
+ {
2808
+ "epoch": 2.82,
2809
+ "learning_rate": 0.00013254723127035828,
2810
+ "loss": 0.2929,
2811
+ "step": 4390
2812
+ },
2813
+ {
2814
+ "epoch": 2.83,
2815
+ "learning_rate": 0.00013215635179153094,
2816
+ "loss": 0.3148,
2817
+ "step": 4400
2818
+ },
2819
+ {
2820
+ "epoch": 2.83,
2821
+ "eval_loss": 0.32642272114753723,
2822
+ "eval_runtime": 517.6484,
2823
+ "eval_samples_per_second": 3.864,
2824
+ "eval_steps_per_second": 0.483,
2825
+ "step": 4400
2826
+ },
2827
+ {
2828
+ "epoch": 2.84,
2829
+ "learning_rate": 0.0001317654723127036,
2830
+ "loss": 0.294,
2831
+ "step": 4410
2832
+ },
2833
+ {
2834
+ "epoch": 2.84,
2835
+ "learning_rate": 0.00013137459283387622,
2836
+ "loss": 0.3155,
2837
+ "step": 4420
2838
+ },
2839
+ {
2840
+ "epoch": 2.85,
2841
+ "learning_rate": 0.00013098371335504885,
2842
+ "loss": 0.3044,
2843
+ "step": 4430
2844
+ },
2845
+ {
2846
+ "epoch": 2.86,
2847
+ "learning_rate": 0.00013059283387622148,
2848
+ "loss": 0.2997,
2849
+ "step": 4440
2850
+ },
2851
+ {
2852
+ "epoch": 2.86,
2853
+ "learning_rate": 0.00013020195439739413,
2854
+ "loss": 0.3119,
2855
+ "step": 4450
2856
+ },
2857
+ {
2858
+ "epoch": 2.87,
2859
+ "learning_rate": 0.00012981107491856676,
2860
+ "loss": 0.3206,
2861
+ "step": 4460
2862
+ },
2863
+ {
2864
+ "epoch": 2.87,
2865
+ "learning_rate": 0.0001294201954397394,
2866
+ "loss": 0.2994,
2867
+ "step": 4470
2868
+ },
2869
+ {
2870
+ "epoch": 2.88,
2871
+ "learning_rate": 0.00012902931596091204,
2872
+ "loss": 0.3285,
2873
+ "step": 4480
2874
+ },
2875
+ {
2876
+ "epoch": 2.89,
2877
+ "learning_rate": 0.00012863843648208467,
2878
+ "loss": 0.3021,
2879
+ "step": 4490
2880
+ },
2881
+ {
2882
+ "epoch": 2.89,
2883
+ "learning_rate": 0.00012824755700325733,
2884
+ "loss": 0.319,
2885
+ "step": 4500
2886
+ },
2887
+ {
2888
+ "epoch": 2.9,
2889
+ "learning_rate": 0.00012785667752442996,
2890
+ "loss": 0.3016,
2891
+ "step": 4510
2892
+ },
2893
+ {
2894
+ "epoch": 2.91,
2895
+ "learning_rate": 0.00012746579804560258,
2896
+ "loss": 0.3025,
2897
+ "step": 4520
2898
+ },
2899
+ {
2900
+ "epoch": 2.91,
2901
+ "learning_rate": 0.00012707491856677524,
2902
+ "loss": 0.3234,
2903
+ "step": 4530
2904
+ },
2905
+ {
2906
+ "epoch": 2.92,
2907
+ "learning_rate": 0.00012668403908794787,
2908
+ "loss": 0.3131,
2909
+ "step": 4540
2910
+ },
2911
+ {
2912
+ "epoch": 2.93,
2913
+ "learning_rate": 0.00012629315960912052,
2914
+ "loss": 0.3171,
2915
+ "step": 4550
2916
+ },
2917
+ {
2918
+ "epoch": 2.93,
2919
+ "learning_rate": 0.00012590228013029315,
2920
+ "loss": 0.2973,
2921
+ "step": 4560
2922
+ },
2923
+ {
2924
+ "epoch": 2.94,
2925
+ "learning_rate": 0.00012551140065146578,
2926
+ "loss": 0.3104,
2927
+ "step": 4570
2928
+ },
2929
+ {
2930
+ "epoch": 2.95,
2931
+ "learning_rate": 0.00012512052117263843,
2932
+ "loss": 0.2971,
2933
+ "step": 4580
2934
+ },
2935
+ {
2936
+ "epoch": 2.95,
2937
+ "learning_rate": 0.00012472964169381106,
2938
+ "loss": 0.311,
2939
+ "step": 4590
2940
+ },
2941
+ {
2942
+ "epoch": 2.96,
2943
+ "learning_rate": 0.0001243387622149837,
2944
+ "loss": 0.3162,
2945
+ "step": 4600
2946
+ },
2947
+ {
2948
+ "epoch": 2.96,
2949
+ "eval_loss": 0.3261347711086273,
2950
+ "eval_runtime": 518.85,
2951
+ "eval_samples_per_second": 3.855,
2952
+ "eval_steps_per_second": 0.482,
2953
+ "step": 4600
2954
+ },
2955
+ {
2956
+ "epoch": 2.96,
2957
+ "learning_rate": 0.00012394788273615634,
2958
+ "loss": 0.3169,
2959
+ "step": 4610
2960
+ },
2961
+ {
2962
+ "epoch": 2.97,
2963
+ "learning_rate": 0.00012355700325732897,
2964
+ "loss": 0.3189,
2965
+ "step": 4620
2966
+ },
2967
+ {
2968
+ "epoch": 2.98,
2969
+ "learning_rate": 0.00012316612377850163,
2970
+ "loss": 0.3041,
2971
+ "step": 4630
2972
+ },
2973
+ {
2974
+ "epoch": 2.98,
2975
+ "learning_rate": 0.00012277524429967426,
2976
+ "loss": 0.3153,
2977
+ "step": 4640
2978
+ },
2979
+ {
2980
+ "epoch": 2.99,
2981
+ "learning_rate": 0.00012238436482084688,
2982
+ "loss": 0.3127,
2983
+ "step": 4650
2984
+ },
2985
+ {
2986
+ "epoch": 3.0,
2987
+ "learning_rate": 0.00012199348534201954,
2988
+ "loss": 0.3016,
2989
+ "step": 4660
2990
+ },
2991
+ {
2992
+ "epoch": 3.0,
2993
+ "learning_rate": 0.00012160260586319217,
2994
+ "loss": 0.2938,
2995
+ "step": 4670
2996
+ },
2997
+ {
2998
+ "epoch": 3.01,
2999
+ "learning_rate": 0.00012121172638436481,
3000
+ "loss": 0.2971,
3001
+ "step": 4680
3002
+ },
3003
+ {
3004
+ "epoch": 3.02,
3005
+ "learning_rate": 0.00012082084690553745,
3006
+ "loss": 0.3045,
3007
+ "step": 4690
3008
+ },
3009
+ {
3010
+ "epoch": 3.02,
3011
+ "learning_rate": 0.00012042996742671008,
3012
+ "loss": 0.3054,
3013
+ "step": 4700
3014
+ },
3015
+ {
3016
+ "epoch": 3.03,
3017
+ "learning_rate": 0.00012003908794788273,
3018
+ "loss": 0.3018,
3019
+ "step": 4710
3020
+ },
3021
+ {
3022
+ "epoch": 3.04,
3023
+ "learning_rate": 0.00011964820846905536,
3024
+ "loss": 0.2913,
3025
+ "step": 4720
3026
+ },
3027
+ {
3028
+ "epoch": 3.04,
3029
+ "learning_rate": 0.000119257328990228,
3030
+ "loss": 0.2911,
3031
+ "step": 4730
3032
+ },
3033
+ {
3034
+ "epoch": 3.05,
3035
+ "learning_rate": 0.00011886644951140065,
3036
+ "loss": 0.2968,
3037
+ "step": 4740
3038
+ },
3039
+ {
3040
+ "epoch": 3.05,
3041
+ "learning_rate": 0.00011847557003257327,
3042
+ "loss": 0.2954,
3043
+ "step": 4750
3044
+ },
3045
+ {
3046
+ "epoch": 3.06,
3047
+ "learning_rate": 0.00011808469055374592,
3048
+ "loss": 0.2897,
3049
+ "step": 4760
3050
+ },
3051
+ {
3052
+ "epoch": 3.07,
3053
+ "learning_rate": 0.00011769381107491857,
3054
+ "loss": 0.3026,
3055
+ "step": 4770
3056
+ },
3057
+ {
3058
+ "epoch": 3.07,
3059
+ "learning_rate": 0.0001173029315960912,
3060
+ "loss": 0.2981,
3061
+ "step": 4780
3062
+ },
3063
+ {
3064
+ "epoch": 3.08,
3065
+ "learning_rate": 0.00011691205211726384,
3066
+ "loss": 0.2948,
3067
+ "step": 4790
3068
+ },
3069
+ {
3070
+ "epoch": 3.09,
3071
+ "learning_rate": 0.00011652117263843647,
3072
+ "loss": 0.2939,
3073
+ "step": 4800
3074
+ },
3075
+ {
3076
+ "epoch": 3.09,
3077
+ "eval_loss": 0.32907190918922424,
3078
+ "eval_runtime": 519.2943,
3079
+ "eval_samples_per_second": 3.851,
3080
+ "eval_steps_per_second": 0.481,
3081
+ "step": 4800
3082
+ },
3083
+ {
3084
+ "epoch": 3.09,
3085
+ "learning_rate": 0.00011613029315960911,
3086
+ "loss": 0.2999,
3087
+ "step": 4810
3088
+ },
3089
+ {
3090
+ "epoch": 3.1,
3091
+ "learning_rate": 0.00011573941368078175,
3092
+ "loss": 0.289,
3093
+ "step": 4820
3094
+ },
3095
+ {
3096
+ "epoch": 3.11,
3097
+ "learning_rate": 0.00011534853420195438,
3098
+ "loss": 0.2974,
3099
+ "step": 4830
3100
+ },
3101
+ {
3102
+ "epoch": 3.11,
3103
+ "learning_rate": 0.00011495765472312704,
3104
+ "loss": 0.3036,
3105
+ "step": 4840
3106
+ },
3107
+ {
3108
+ "epoch": 3.12,
3109
+ "learning_rate": 0.00011456677524429966,
3110
+ "loss": 0.2907,
3111
+ "step": 4850
3112
+ },
3113
+ {
3114
+ "epoch": 3.13,
3115
+ "learning_rate": 0.0001141758957654723,
3116
+ "loss": 0.3061,
3117
+ "step": 4860
3118
+ },
3119
+ {
3120
+ "epoch": 3.13,
3121
+ "learning_rate": 0.00011378501628664495,
3122
+ "loss": 0.305,
3123
+ "step": 4870
3124
+ },
3125
+ {
3126
+ "epoch": 3.14,
3127
+ "learning_rate": 0.00011339413680781758,
3128
+ "loss": 0.2956,
3129
+ "step": 4880
3130
+ },
3131
+ {
3132
+ "epoch": 3.14,
3133
+ "learning_rate": 0.00011300325732899022,
3134
+ "loss": 0.2921,
3135
+ "step": 4890
3136
+ },
3137
+ {
3138
+ "epoch": 3.15,
3139
+ "learning_rate": 0.00011261237785016285,
3140
+ "loss": 0.2924,
3141
+ "step": 4900
3142
+ },
3143
+ {
3144
+ "epoch": 3.16,
3145
+ "learning_rate": 0.0001122214983713355,
3146
+ "loss": 0.282,
3147
+ "step": 4910
3148
+ },
3149
+ {
3150
+ "epoch": 3.16,
3151
+ "learning_rate": 0.00011183061889250814,
3152
+ "loss": 0.3045,
3153
+ "step": 4920
3154
+ },
3155
+ {
3156
+ "epoch": 3.17,
3157
+ "learning_rate": 0.00011143973941368077,
3158
+ "loss": 0.2975,
3159
+ "step": 4930
3160
+ },
3161
+ {
3162
+ "epoch": 3.18,
3163
+ "learning_rate": 0.00011104885993485341,
3164
+ "loss": 0.2985,
3165
+ "step": 4940
3166
+ },
3167
+ {
3168
+ "epoch": 3.18,
3169
+ "learning_rate": 0.00011065798045602604,
3170
+ "loss": 0.2972,
3171
+ "step": 4950
3172
+ },
3173
+ {
3174
+ "epoch": 3.19,
3175
+ "learning_rate": 0.00011026710097719868,
3176
+ "loss": 0.2895,
3177
+ "step": 4960
3178
+ },
3179
+ {
3180
+ "epoch": 3.2,
3181
+ "learning_rate": 0.00010987622149837132,
3182
+ "loss": 0.2971,
3183
+ "step": 4970
3184
+ },
3185
+ {
3186
+ "epoch": 3.2,
3187
+ "learning_rate": 0.00010948534201954397,
3188
+ "loss": 0.3026,
3189
+ "step": 4980
3190
+ },
3191
+ {
3192
+ "epoch": 3.21,
3193
+ "learning_rate": 0.00010909446254071661,
3194
+ "loss": 0.2913,
3195
+ "step": 4990
3196
+ },
3197
+ {
3198
+ "epoch": 3.22,
3199
+ "learning_rate": 0.00010870358306188925,
3200
+ "loss": 0.3076,
3201
+ "step": 5000
3202
+ },
3203
+ {
3204
+ "epoch": 3.22,
3205
+ "eval_loss": 0.3277848958969116,
3206
+ "eval_runtime": 520.0003,
3207
+ "eval_samples_per_second": 3.846,
3208
+ "eval_steps_per_second": 0.481,
3209
+ "step": 5000
3210
+ },
3211
+ {
3212
+ "epoch": 3.22,
3213
+ "learning_rate": 0.00010831270358306188,
3214
+ "loss": 0.2947,
3215
+ "step": 5010
3216
+ },
3217
+ {
3218
+ "epoch": 3.23,
3219
+ "learning_rate": 0.00010792182410423452,
3220
+ "loss": 0.3057,
3221
+ "step": 5020
3222
+ },
3223
+ {
3224
+ "epoch": 3.23,
3225
+ "learning_rate": 0.00010753094462540715,
3226
+ "loss": 0.2815,
3227
+ "step": 5030
3228
+ },
3229
+ {
3230
+ "epoch": 3.24,
3231
+ "learning_rate": 0.00010714006514657979,
3232
+ "loss": 0.2906,
3233
+ "step": 5040
3234
+ },
3235
+ {
3236
+ "epoch": 3.25,
3237
+ "learning_rate": 0.00010674918566775244,
3238
+ "loss": 0.3094,
3239
+ "step": 5050
3240
+ },
3241
+ {
3242
+ "epoch": 3.25,
3243
+ "learning_rate": 0.00010635830618892507,
3244
+ "loss": 0.2952,
3245
+ "step": 5060
3246
+ },
3247
+ {
3248
+ "epoch": 3.26,
3249
+ "learning_rate": 0.00010596742671009771,
3250
+ "loss": 0.2999,
3251
+ "step": 5070
3252
+ },
3253
+ {
3254
+ "epoch": 3.27,
3255
+ "learning_rate": 0.00010557654723127034,
3256
+ "loss": 0.2944,
3257
+ "step": 5080
3258
+ },
3259
+ {
3260
+ "epoch": 3.27,
3261
+ "learning_rate": 0.00010518566775244298,
3262
+ "loss": 0.294,
3263
+ "step": 5090
3264
+ },
3265
+ {
3266
+ "epoch": 3.28,
3267
+ "learning_rate": 0.00010479478827361563,
3268
+ "loss": 0.3064,
3269
+ "step": 5100
3270
+ },
3271
+ {
3272
+ "epoch": 3.29,
3273
+ "learning_rate": 0.00010440390879478825,
3274
+ "loss": 0.2915,
3275
+ "step": 5110
3276
+ },
3277
+ {
3278
+ "epoch": 3.29,
3279
+ "learning_rate": 0.00010401302931596091,
3280
+ "loss": 0.3013,
3281
+ "step": 5120
3282
+ },
3283
+ {
3284
+ "epoch": 3.3,
3285
+ "learning_rate": 0.00010362214983713354,
3286
+ "loss": 0.2934,
3287
+ "step": 5130
3288
+ },
3289
+ {
3290
+ "epoch": 3.31,
3291
+ "learning_rate": 0.00010323127035830618,
3292
+ "loss": 0.285,
3293
+ "step": 5140
3294
+ },
3295
+ {
3296
+ "epoch": 3.31,
3297
+ "learning_rate": 0.00010284039087947882,
3298
+ "loss": 0.3002,
3299
+ "step": 5150
3300
+ },
3301
+ {
3302
+ "epoch": 3.32,
3303
+ "learning_rate": 0.00010244951140065145,
3304
+ "loss": 0.3005,
3305
+ "step": 5160
3306
+ },
3307
+ {
3308
+ "epoch": 3.32,
3309
+ "learning_rate": 0.00010205863192182409,
3310
+ "loss": 0.2935,
3311
+ "step": 5170
3312
+ },
3313
+ {
3314
+ "epoch": 3.33,
3315
+ "learning_rate": 0.00010166775244299675,
3316
+ "loss": 0.2933,
3317
+ "step": 5180
3318
+ },
3319
+ {
3320
+ "epoch": 3.34,
3321
+ "learning_rate": 0.00010127687296416937,
3322
+ "loss": 0.302,
3323
+ "step": 5190
3324
+ },
3325
+ {
3326
+ "epoch": 3.34,
3327
+ "learning_rate": 0.00010088599348534202,
3328
+ "loss": 0.2807,
3329
+ "step": 5200
3330
+ },
3331
+ {
3332
+ "epoch": 3.34,
3333
+ "eval_loss": 0.3280768394470215,
3334
+ "eval_runtime": 516.0013,
3335
+ "eval_samples_per_second": 3.876,
3336
+ "eval_steps_per_second": 0.484,
3337
+ "step": 5200
3338
+ },
3339
+ {
3340
+ "epoch": 3.35,
3341
+ "learning_rate": 0.00010049511400651464,
3342
+ "loss": 0.3021,
3343
+ "step": 5210
3344
+ },
3345
+ {
3346
+ "epoch": 3.36,
3347
+ "learning_rate": 0.00010010423452768728,
3348
+ "loss": 0.2941,
3349
+ "step": 5220
3350
+ },
3351
+ {
3352
+ "epoch": 3.36,
3353
+ "learning_rate": 9.971335504885993e-05,
3354
+ "loss": 0.2938,
3355
+ "step": 5230
3356
+ },
3357
+ {
3358
+ "epoch": 3.37,
3359
+ "learning_rate": 9.932247557003255e-05,
3360
+ "loss": 0.2975,
3361
+ "step": 5240
3362
+ },
3363
+ {
3364
+ "epoch": 3.38,
3365
+ "learning_rate": 9.893159609120521e-05,
3366
+ "loss": 0.2946,
3367
+ "step": 5250
3368
+ },
3369
+ {
3370
+ "epoch": 3.38,
3371
+ "learning_rate": 9.854071661237784e-05,
3372
+ "loss": 0.2965,
3373
+ "step": 5260
3374
+ },
3375
+ {
3376
+ "epoch": 3.39,
3377
+ "learning_rate": 9.814983713355048e-05,
3378
+ "loss": 0.2998,
3379
+ "step": 5270
3380
+ },
3381
+ {
3382
+ "epoch": 3.4,
3383
+ "learning_rate": 9.775895765472312e-05,
3384
+ "loss": 0.3109,
3385
+ "step": 5280
3386
+ },
3387
+ {
3388
+ "epoch": 3.4,
3389
+ "learning_rate": 9.736807817589575e-05,
3390
+ "loss": 0.2848,
3391
+ "step": 5290
3392
+ },
3393
+ {
3394
+ "epoch": 3.41,
3395
+ "learning_rate": 9.697719869706839e-05,
3396
+ "loss": 0.3021,
3397
+ "step": 5300
3398
+ },
3399
+ {
3400
+ "epoch": 3.41,
3401
+ "learning_rate": 9.658631921824102e-05,
3402
+ "loss": 0.3019,
3403
+ "step": 5310
3404
+ },
3405
+ {
3406
+ "epoch": 3.42,
3407
+ "learning_rate": 9.619543973941367e-05,
3408
+ "loss": 0.2992,
3409
+ "step": 5320
3410
+ },
3411
+ {
3412
+ "epoch": 3.43,
3413
+ "learning_rate": 9.580456026058632e-05,
3414
+ "loss": 0.29,
3415
+ "step": 5330
3416
+ },
3417
+ {
3418
+ "epoch": 3.43,
3419
+ "learning_rate": 9.541368078175894e-05,
3420
+ "loss": 0.2817,
3421
+ "step": 5340
3422
+ },
3423
+ {
3424
+ "epoch": 3.44,
3425
+ "learning_rate": 9.502280130293159e-05,
3426
+ "loss": 0.3014,
3427
+ "step": 5350
3428
+ },
3429
+ {
3430
+ "epoch": 3.45,
3431
+ "learning_rate": 9.463192182410423e-05,
3432
+ "loss": 0.2975,
3433
+ "step": 5360
3434
+ },
3435
+ {
3436
+ "epoch": 3.45,
3437
+ "learning_rate": 9.424104234527686e-05,
3438
+ "loss": 0.295,
3439
+ "step": 5370
3440
+ },
3441
+ {
3442
+ "epoch": 3.46,
3443
+ "learning_rate": 9.385016286644951e-05,
3444
+ "loss": 0.3049,
3445
+ "step": 5380
3446
+ },
3447
+ {
3448
+ "epoch": 3.47,
3449
+ "learning_rate": 9.345928338762214e-05,
3450
+ "loss": 0.3047,
3451
+ "step": 5390
3452
+ },
3453
+ {
3454
+ "epoch": 3.47,
3455
+ "learning_rate": 9.306840390879478e-05,
3456
+ "loss": 0.3013,
3457
+ "step": 5400
3458
+ },
3459
+ {
3460
+ "epoch": 3.47,
3461
+ "eval_loss": 0.32735779881477356,
3462
+ "eval_runtime": 506.9924,
3463
+ "eval_samples_per_second": 3.945,
3464
+ "eval_steps_per_second": 0.493,
3465
+ "step": 5400
3466
+ },
3467
+ {
3468
+ "epoch": 3.48,
3469
+ "learning_rate": 9.267752442996742e-05,
3470
+ "loss": 0.2979,
3471
+ "step": 5410
3472
+ },
3473
+ {
3474
+ "epoch": 3.49,
3475
+ "learning_rate": 9.228664495114005e-05,
3476
+ "loss": 0.2957,
3477
+ "step": 5420
3478
+ },
3479
+ {
3480
+ "epoch": 3.49,
3481
+ "learning_rate": 9.189576547231269e-05,
3482
+ "loss": 0.2998,
3483
+ "step": 5430
3484
+ },
3485
+ {
3486
+ "epoch": 3.5,
3487
+ "learning_rate": 9.150488599348532e-05,
3488
+ "loss": 0.3027,
3489
+ "step": 5440
3490
+ },
3491
+ {
3492
+ "epoch": 3.5,
3493
+ "learning_rate": 9.111400651465798e-05,
3494
+ "loss": 0.2841,
3495
+ "step": 5450
3496
+ },
3497
+ {
3498
+ "epoch": 3.51,
3499
+ "learning_rate": 9.072312703583062e-05,
3500
+ "loss": 0.2919,
3501
+ "step": 5460
3502
+ },
3503
+ {
3504
+ "epoch": 3.52,
3505
+ "learning_rate": 9.033224755700325e-05,
3506
+ "loss": 0.308,
3507
+ "step": 5470
3508
+ },
3509
+ {
3510
+ "epoch": 3.52,
3511
+ "learning_rate": 8.994136807817589e-05,
3512
+ "loss": 0.2871,
3513
+ "step": 5480
3514
+ },
3515
+ {
3516
+ "epoch": 3.53,
3517
+ "learning_rate": 8.955048859934852e-05,
3518
+ "loss": 0.2883,
3519
+ "step": 5490
3520
+ },
3521
+ {
3522
+ "epoch": 3.54,
3523
+ "learning_rate": 8.915960912052116e-05,
3524
+ "loss": 0.2975,
3525
+ "step": 5500
3526
+ },
3527
+ {
3528
+ "epoch": 3.54,
3529
+ "learning_rate": 8.876872964169381e-05,
3530
+ "loss": 0.2962,
3531
+ "step": 5510
3532
+ },
3533
+ {
3534
+ "epoch": 3.55,
3535
+ "learning_rate": 8.837785016286644e-05,
3536
+ "loss": 0.2933,
3537
+ "step": 5520
3538
+ },
3539
+ {
3540
+ "epoch": 3.56,
3541
+ "learning_rate": 8.798697068403908e-05,
3542
+ "loss": 0.2967,
3543
+ "step": 5530
3544
+ },
3545
+ {
3546
+ "epoch": 3.56,
3547
+ "learning_rate": 8.759609120521172e-05,
3548
+ "loss": 0.2902,
3549
+ "step": 5540
3550
+ },
3551
+ {
3552
+ "epoch": 3.57,
3553
+ "learning_rate": 8.720521172638435e-05,
3554
+ "loss": 0.2969,
3555
+ "step": 5550
3556
+ },
3557
+ {
3558
+ "epoch": 3.58,
3559
+ "learning_rate": 8.6814332247557e-05,
3560
+ "loss": 0.2967,
3561
+ "step": 5560
3562
+ },
3563
+ {
3564
+ "epoch": 3.58,
3565
+ "learning_rate": 8.642345276872962e-05,
3566
+ "loss": 0.2923,
3567
+ "step": 5570
3568
+ },
3569
+ {
3570
+ "epoch": 3.59,
3571
+ "learning_rate": 8.603257328990228e-05,
3572
+ "loss": 0.2938,
3573
+ "step": 5580
3574
+ },
3575
+ {
3576
+ "epoch": 3.59,
3577
+ "learning_rate": 8.564169381107492e-05,
3578
+ "loss": 0.2941,
3579
+ "step": 5590
3580
+ },
3581
+ {
3582
+ "epoch": 3.6,
3583
+ "learning_rate": 8.525081433224755e-05,
3584
+ "loss": 0.3092,
3585
+ "step": 5600
3586
+ },
3587
+ {
3588
+ "epoch": 3.6,
3589
+ "eval_loss": 0.3266870081424713,
3590
+ "eval_runtime": 504.0288,
3591
+ "eval_samples_per_second": 3.968,
3592
+ "eval_steps_per_second": 0.496,
3593
+ "step": 5600
3594
+ },
3595
+ {
3596
+ "epoch": 3.61,
3597
+ "learning_rate": 8.485993485342019e-05,
3598
+ "loss": 0.2834,
3599
+ "step": 5610
3600
+ },
3601
+ {
3602
+ "epoch": 3.61,
3603
+ "learning_rate": 8.446905537459282e-05,
3604
+ "loss": 0.2953,
3605
+ "step": 5620
3606
+ },
3607
+ {
3608
+ "epoch": 3.62,
3609
+ "learning_rate": 8.407817589576546e-05,
3610
+ "loss": 0.2918,
3611
+ "step": 5630
3612
+ },
3613
+ {
3614
+ "epoch": 3.63,
3615
+ "learning_rate": 8.368729641693811e-05,
3616
+ "loss": 0.2968,
3617
+ "step": 5640
3618
+ },
3619
+ {
3620
+ "epoch": 3.63,
3621
+ "learning_rate": 8.329641693811074e-05,
3622
+ "loss": 0.2891,
3623
+ "step": 5650
3624
+ },
3625
+ {
3626
+ "epoch": 3.64,
3627
+ "learning_rate": 8.290553745928338e-05,
3628
+ "loss": 0.2927,
3629
+ "step": 5660
3630
+ },
3631
+ {
3632
+ "epoch": 3.65,
3633
+ "learning_rate": 8.251465798045601e-05,
3634
+ "loss": 0.3057,
3635
+ "step": 5670
3636
+ },
3637
+ {
3638
+ "epoch": 3.65,
3639
+ "learning_rate": 8.212377850162865e-05,
3640
+ "loss": 0.2946,
3641
+ "step": 5680
3642
+ },
3643
+ {
3644
+ "epoch": 3.66,
3645
+ "learning_rate": 8.17328990228013e-05,
3646
+ "loss": 0.296,
3647
+ "step": 5690
3648
+ },
3649
+ {
3650
+ "epoch": 3.67,
3651
+ "learning_rate": 8.134201954397392e-05,
3652
+ "loss": 0.2864,
3653
+ "step": 5700
3654
+ },
3655
+ {
3656
+ "epoch": 3.67,
3657
+ "learning_rate": 8.095114006514658e-05,
3658
+ "loss": 0.2967,
3659
+ "step": 5710
3660
+ },
3661
+ {
3662
+ "epoch": 3.68,
3663
+ "learning_rate": 8.056026058631921e-05,
3664
+ "loss": 0.2928,
3665
+ "step": 5720
3666
+ },
3667
+ {
3668
+ "epoch": 3.68,
3669
+ "learning_rate": 8.016938110749185e-05,
3670
+ "loss": 0.2986,
3671
+ "step": 5730
3672
+ },
3673
+ {
3674
+ "epoch": 3.69,
3675
+ "learning_rate": 7.977850162866449e-05,
3676
+ "loss": 0.2948,
3677
+ "step": 5740
3678
+ },
3679
+ {
3680
+ "epoch": 3.7,
3681
+ "learning_rate": 7.938762214983712e-05,
3682
+ "loss": 0.2936,
3683
+ "step": 5750
3684
+ },
3685
+ {
3686
+ "epoch": 3.7,
3687
+ "learning_rate": 7.899674267100976e-05,
3688
+ "loss": 0.2961,
3689
+ "step": 5760
3690
+ },
3691
+ {
3692
+ "epoch": 3.71,
3693
+ "learning_rate": 7.860586319218242e-05,
3694
+ "loss": 0.2959,
3695
+ "step": 5770
3696
+ },
3697
+ {
3698
+ "epoch": 3.72,
3699
+ "learning_rate": 7.821498371335504e-05,
3700
+ "loss": 0.3029,
3701
+ "step": 5780
3702
+ },
3703
+ {
3704
+ "epoch": 3.72,
3705
+ "learning_rate": 7.782410423452769e-05,
3706
+ "loss": 0.2922,
3707
+ "step": 5790
3708
+ },
3709
+ {
3710
+ "epoch": 3.73,
3711
+ "learning_rate": 7.743322475570031e-05,
3712
+ "loss": 0.2997,
3713
+ "step": 5800
3714
+ },
3715
+ {
3716
+ "epoch": 3.73,
3717
+ "eval_loss": 0.3263789117336273,
3718
+ "eval_runtime": 503.8664,
3719
+ "eval_samples_per_second": 3.969,
3720
+ "eval_steps_per_second": 0.496,
3721
+ "step": 5800
3722
+ },
3723
+ {
3724
+ "epoch": 3.74,
3725
+ "learning_rate": 7.704234527687296e-05,
3726
+ "loss": 0.2885,
3727
+ "step": 5810
3728
+ },
3729
+ {
3730
+ "epoch": 3.74,
3731
+ "learning_rate": 7.66514657980456e-05,
3732
+ "loss": 0.3113,
3733
+ "step": 5820
3734
+ },
3735
+ {
3736
+ "epoch": 3.75,
3737
+ "learning_rate": 7.626058631921822e-05,
3738
+ "loss": 0.2886,
3739
+ "step": 5830
3740
+ },
3741
+ {
3742
+ "epoch": 3.76,
3743
+ "learning_rate": 7.586970684039088e-05,
3744
+ "loss": 0.2976,
3745
+ "step": 5840
3746
+ },
3747
+ {
3748
+ "epoch": 3.76,
3749
+ "learning_rate": 7.547882736156351e-05,
3750
+ "loss": 0.305,
3751
+ "step": 5850
3752
+ },
3753
+ {
3754
+ "epoch": 3.77,
3755
+ "learning_rate": 7.508794788273615e-05,
3756
+ "loss": 0.3059,
3757
+ "step": 5860
3758
+ },
3759
+ {
3760
+ "epoch": 3.77,
3761
+ "learning_rate": 7.469706840390879e-05,
3762
+ "loss": 0.2992,
3763
+ "step": 5870
3764
+ },
3765
+ {
3766
+ "epoch": 3.78,
3767
+ "learning_rate": 7.430618892508143e-05,
3768
+ "loss": 0.2943,
3769
+ "step": 5880
3770
+ },
3771
+ {
3772
+ "epoch": 3.79,
3773
+ "learning_rate": 7.391530944625406e-05,
3774
+ "loss": 0.2861,
3775
+ "step": 5890
3776
+ },
3777
+ {
3778
+ "epoch": 3.79,
3779
+ "learning_rate": 7.35244299674267e-05,
3780
+ "loss": 0.2862,
3781
+ "step": 5900
3782
+ },
3783
+ {
3784
+ "epoch": 3.8,
3785
+ "learning_rate": 7.313355048859934e-05,
3786
+ "loss": 0.2938,
3787
+ "step": 5910
3788
+ },
3789
+ {
3790
+ "epoch": 3.81,
3791
+ "learning_rate": 7.274267100977197e-05,
3792
+ "loss": 0.2993,
3793
+ "step": 5920
3794
+ },
3795
+ {
3796
+ "epoch": 3.81,
3797
+ "learning_rate": 7.235179153094463e-05,
3798
+ "loss": 0.2955,
3799
+ "step": 5930
3800
+ },
3801
+ {
3802
+ "epoch": 3.82,
3803
+ "learning_rate": 7.196091205211726e-05,
3804
+ "loss": 0.2996,
3805
+ "step": 5940
3806
+ },
3807
+ {
3808
+ "epoch": 3.83,
3809
+ "learning_rate": 7.15700325732899e-05,
3810
+ "loss": 0.3063,
3811
+ "step": 5950
3812
+ },
3813
+ {
3814
+ "epoch": 3.83,
3815
+ "learning_rate": 7.117915309446253e-05,
3816
+ "loss": 0.2998,
3817
+ "step": 5960
3818
+ },
3819
+ {
3820
+ "epoch": 3.84,
3821
+ "learning_rate": 7.078827361563518e-05,
3822
+ "loss": 0.3062,
3823
+ "step": 5970
3824
+ },
3825
+ {
3826
+ "epoch": 3.85,
3827
+ "learning_rate": 7.039739413680781e-05,
3828
+ "loss": 0.3001,
3829
+ "step": 5980
3830
+ },
3831
+ {
3832
+ "epoch": 3.85,
3833
+ "learning_rate": 7.000651465798045e-05,
3834
+ "loss": 0.3072,
3835
+ "step": 5990
3836
+ },
3837
+ {
3838
+ "epoch": 3.86,
3839
+ "learning_rate": 6.961563517915309e-05,
3840
+ "loss": 0.3046,
3841
+ "step": 6000
3842
+ },
3843
+ {
3844
+ "epoch": 3.86,
3845
+ "eval_loss": 0.3262394964694977,
3846
+ "eval_runtime": 505.9386,
3847
+ "eval_samples_per_second": 3.953,
3848
+ "eval_steps_per_second": 0.494,
3849
+ "step": 6000
3850
+ },
3851
+ {
3852
+ "epoch": 3.86,
3853
+ "learning_rate": 6.922475570032572e-05,
3854
+ "loss": 0.2936,
3855
+ "step": 6010
3856
+ },
3857
+ {
3858
+ "epoch": 3.87,
3859
+ "learning_rate": 6.883387622149836e-05,
3860
+ "loss": 0.2981,
3861
+ "step": 6020
3862
+ },
3863
+ {
3864
+ "epoch": 3.88,
3865
+ "learning_rate": 6.8442996742671e-05,
3866
+ "loss": 0.2868,
3867
+ "step": 6030
3868
+ },
3869
+ {
3870
+ "epoch": 3.88,
3871
+ "learning_rate": 6.805211726384365e-05,
3872
+ "loss": 0.2895,
3873
+ "step": 6040
3874
+ },
3875
+ {
3876
+ "epoch": 3.89,
3877
+ "learning_rate": 6.766123778501627e-05,
3878
+ "loss": 0.2851,
3879
+ "step": 6050
3880
+ },
3881
+ {
3882
+ "epoch": 3.9,
3883
+ "learning_rate": 6.727035830618892e-05,
3884
+ "loss": 0.2997,
3885
+ "step": 6060
3886
+ },
3887
+ {
3888
+ "epoch": 3.9,
3889
+ "learning_rate": 6.687947882736156e-05,
3890
+ "loss": 0.2954,
3891
+ "step": 6070
3892
+ },
3893
+ {
3894
+ "epoch": 3.91,
3895
+ "learning_rate": 6.64885993485342e-05,
3896
+ "loss": 0.2969,
3897
+ "step": 6080
3898
+ },
3899
+ {
3900
+ "epoch": 3.92,
3901
+ "learning_rate": 6.609771986970683e-05,
3902
+ "loss": 0.2926,
3903
+ "step": 6090
3904
+ },
3905
+ {
3906
+ "epoch": 3.92,
3907
+ "learning_rate": 6.570684039087947e-05,
3908
+ "loss": 0.2934,
3909
+ "step": 6100
3910
+ },
3911
+ {
3912
+ "epoch": 3.93,
3913
+ "learning_rate": 6.531596091205211e-05,
3914
+ "loss": 0.3088,
3915
+ "step": 6110
3916
+ },
3917
+ {
3918
+ "epoch": 3.94,
3919
+ "learning_rate": 6.492508143322475e-05,
3920
+ "loss": 0.3065,
3921
+ "step": 6120
3922
+ },
3923
+ {
3924
+ "epoch": 3.94,
3925
+ "learning_rate": 6.45342019543974e-05,
3926
+ "loss": 0.2921,
3927
+ "step": 6130
3928
+ },
3929
+ {
3930
+ "epoch": 3.95,
3931
+ "learning_rate": 6.414332247557002e-05,
3932
+ "loss": 0.283,
3933
+ "step": 6140
3934
+ },
3935
+ {
3936
+ "epoch": 3.95,
3937
+ "learning_rate": 6.375244299674266e-05,
3938
+ "loss": 0.3073,
3939
+ "step": 6150
3940
+ },
3941
+ {
3942
+ "epoch": 3.96,
3943
+ "learning_rate": 6.33615635179153e-05,
3944
+ "loss": 0.2856,
3945
+ "step": 6160
3946
+ },
3947
+ {
3948
+ "epoch": 3.97,
3949
+ "learning_rate": 6.297068403908795e-05,
3950
+ "loss": 0.3007,
3951
+ "step": 6170
3952
+ },
3953
+ {
3954
+ "epoch": 3.97,
3955
+ "learning_rate": 6.257980456026058e-05,
3956
+ "loss": 0.3029,
3957
+ "step": 6180
3958
+ },
3959
+ {
3960
+ "epoch": 3.98,
3961
+ "learning_rate": 6.218892508143322e-05,
3962
+ "loss": 0.3057,
3963
+ "step": 6190
3964
+ },
3965
+ {
3966
+ "epoch": 3.99,
3967
+ "learning_rate": 6.179804560260586e-05,
3968
+ "loss": 0.2968,
3969
+ "step": 6200
3970
+ },
3971
+ {
3972
+ "epoch": 3.99,
3973
+ "eval_loss": 0.3255977928638458,
3974
+ "eval_runtime": 507.4967,
3975
+ "eval_samples_per_second": 3.941,
3976
+ "eval_steps_per_second": 0.493,
3977
+ "step": 6200
3978
+ }
3979
+ ],
3980
+ "logging_steps": 10,
3981
+ "max_steps": 7775,
3982
+ "num_train_epochs": 5,
3983
+ "save_steps": 200,
3984
+ "total_flos": 5.622440445909271e+18,
3985
+ "trial_name": null,
3986
+ "trial_params": null
3987
+ }
with_input/decomp_code_with_intermediates/code-llama-7b/best_model/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93dc15bb0e8ad18ed14f792ce2aee2f3b91f969031eaf46c387f15ace0de77d0
3
+ size 4091