Hugging Face's logo

mkhalifa
/
instrucode

Model card Files Community
Muhammad Khalifa commited on
Commit
99cb18a
·
1 Parent(s): e1fbab3

add low lr code llama-13b

Browse files
Files changed (7) hide show
  1. with_input/decomp_code_with_intermediates/code-llama-13b-low-lr/best_model/adapter_config.json +21 -0
  2. with_input/decomp_code_with_intermediates/code-llama-13b-low-lr/best_model/adapter_model.bin +3 -0
  3. with_input/decomp_code_with_intermediates/code-llama-13b-low-lr/best_model/optimizer.pt +3 -0
  4. with_input/decomp_code_with_intermediates/code-llama-13b-low-lr/best_model/rng_state.pth +3 -0
  5. with_input/decomp_code_with_intermediates/code-llama-13b-low-lr/best_model/scheduler.pt +3 -0
  6. with_input/decomp_code_with_intermediates/code-llama-13b-low-lr/best_model/trainer_state.json +3987 -0
  7. with_input/decomp_code_with_intermediates/code-llama-13b-low-lr/best_model/training_args.bin +3 -0
with_input/decomp_code_with_intermediates/code-llama-13b-low-lr/best_model/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model_name_or_path": "codellama/CodeLlama-13b-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-13b-low-lr/best_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a55b5ded6448a36b99730c3abc5edbf15dfcf4c3f8cafeeac07b578b57a71ed
3
+ size 104973389
with_input/decomp_code_with_intermediates/code-llama-13b-low-lr/best_model/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e0afadacf543631e4a61610efb198eef7a2e9a2521ef79e277cdd8b81b55ff97
3
+ size 209984517
with_input/decomp_code_with_intermediates/code-llama-13b-low-lr/best_model/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c9f8bc7bb1b9b304d80d42cdc089a3f831b9bef5dbaf2d43ccf4aafe12630473
3
+ size 14575
with_input/decomp_code_with_intermediates/code-llama-13b-low-lr/best_model/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:133dfe71a56f41e19c3a4be518576365f99b4124f17bf26faf14beea9f7bfd16
3
+ size 627
with_input/decomp_code_with_intermediates/code-llama-13b-low-lr/best_model/trainer_state.json ADDED
@@ -0,0 +1,3987 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.31802934408187866,
3
+ "best_model_checkpoint": "checkpoints/instrucode/with_input/decomp_code_with_intermediates/code-llama-13b-low-lr/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": 1e-05,
14
+ "loss": 0.9907,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.01,
19
+ "learning_rate": 2e-05,
20
+ "loss": 0.9837,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.02,
25
+ "learning_rate": 3e-05,
26
+ "loss": 0.963,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.03,
31
+ "learning_rate": 4e-05,
32
+ "loss": 0.9176,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.03,
37
+ "learning_rate": 5e-05,
38
+ "loss": 0.805,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.04,
43
+ "learning_rate": 6e-05,
44
+ "loss": 0.6042,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.05,
49
+ "learning_rate": 7e-05,
50
+ "loss": 0.4961,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.05,
55
+ "learning_rate": 8e-05,
56
+ "loss": 0.4373,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.06,
61
+ "learning_rate": 9e-05,
62
+ "loss": 0.4321,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.06,
67
+ "learning_rate": 0.0001,
68
+ "loss": 0.4097,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.07,
73
+ "learning_rate": 9.986970684039088e-05,
74
+ "loss": 0.4113,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.08,
79
+ "learning_rate": 9.973941368078177e-05,
80
+ "loss": 0.4015,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.08,
85
+ "learning_rate": 9.960912052117263e-05,
86
+ "loss": 0.4068,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.09,
91
+ "learning_rate": 9.947882736156352e-05,
92
+ "loss": 0.3793,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.1,
97
+ "learning_rate": 9.93485342019544e-05,
98
+ "loss": 0.3905,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.1,
103
+ "learning_rate": 9.921824104234529e-05,
104
+ "loss": 0.3838,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.11,
109
+ "learning_rate": 9.908794788273616e-05,
110
+ "loss": 0.3777,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.12,
115
+ "learning_rate": 9.895765472312704e-05,
116
+ "loss": 0.3744,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.12,
121
+ "learning_rate": 9.882736156351792e-05,
122
+ "loss": 0.3854,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.13,
127
+ "learning_rate": 9.869706840390879e-05,
128
+ "loss": 0.3769,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.13,
133
+ "eval_loss": 0.3723263740539551,
134
+ "eval_runtime": 913.877,
135
+ "eval_samples_per_second": 2.188,
136
+ "eval_steps_per_second": 0.274,
137
+ "step": 200
138
+ },
139
+ {
140
+ "epoch": 0.14,
141
+ "learning_rate": 9.856677524429968e-05,
142
+ "loss": 0.3771,
143
+ "step": 210
144
+ },
145
+ {
146
+ "epoch": 0.14,
147
+ "learning_rate": 9.843648208469056e-05,
148
+ "loss": 0.3792,
149
+ "step": 220
150
+ },
151
+ {
152
+ "epoch": 0.15,
153
+ "learning_rate": 9.830618892508143e-05,
154
+ "loss": 0.3707,
155
+ "step": 230
156
+ },
157
+ {
158
+ "epoch": 0.15,
159
+ "learning_rate": 9.817589576547231e-05,
160
+ "loss": 0.3687,
161
+ "step": 240
162
+ },
163
+ {
164
+ "epoch": 0.16,
165
+ "learning_rate": 9.80456026058632e-05,
166
+ "loss": 0.3605,
167
+ "step": 250
168
+ },
169
+ {
170
+ "epoch": 0.17,
171
+ "learning_rate": 9.791530944625408e-05,
172
+ "loss": 0.373,
173
+ "step": 260
174
+ },
175
+ {
176
+ "epoch": 0.17,
177
+ "learning_rate": 9.778501628664497e-05,
178
+ "loss": 0.36,
179
+ "step": 270
180
+ },
181
+ {
182
+ "epoch": 0.18,
183
+ "learning_rate": 9.765472312703583e-05,
184
+ "loss": 0.3716,
185
+ "step": 280
186
+ },
187
+ {
188
+ "epoch": 0.19,
189
+ "learning_rate": 9.752442996742672e-05,
190
+ "loss": 0.3647,
191
+ "step": 290
192
+ },
193
+ {
194
+ "epoch": 0.19,
195
+ "learning_rate": 9.73941368078176e-05,
196
+ "loss": 0.3655,
197
+ "step": 300
198
+ },
199
+ {
200
+ "epoch": 0.2,
201
+ "learning_rate": 9.726384364820847e-05,
202
+ "loss": 0.3568,
203
+ "step": 310
204
+ },
205
+ {
206
+ "epoch": 0.21,
207
+ "learning_rate": 9.713355048859936e-05,
208
+ "loss": 0.3573,
209
+ "step": 320
210
+ },
211
+ {
212
+ "epoch": 0.21,
213
+ "learning_rate": 9.700325732899024e-05,
214
+ "loss": 0.3635,
215
+ "step": 330
216
+ },
217
+ {
218
+ "epoch": 0.22,
219
+ "learning_rate": 9.687296416938111e-05,
220
+ "loss": 0.359,
221
+ "step": 340
222
+ },
223
+ {
224
+ "epoch": 0.23,
225
+ "learning_rate": 9.674267100977199e-05,
226
+ "loss": 0.3525,
227
+ "step": 350
228
+ },
229
+ {
230
+ "epoch": 0.23,
231
+ "learning_rate": 9.661237785016288e-05,
232
+ "loss": 0.3713,
233
+ "step": 360
234
+ },
235
+ {
236
+ "epoch": 0.24,
237
+ "learning_rate": 9.648208469055375e-05,
238
+ "loss": 0.3605,
239
+ "step": 370
240
+ },
241
+ {
242
+ "epoch": 0.24,
243
+ "learning_rate": 9.635179153094463e-05,
244
+ "loss": 0.3554,
245
+ "step": 380
246
+ },
247
+ {
248
+ "epoch": 0.25,
249
+ "learning_rate": 9.62214983713355e-05,
250
+ "loss": 0.3659,
251
+ "step": 390
252
+ },
253
+ {
254
+ "epoch": 0.26,
255
+ "learning_rate": 9.609120521172638e-05,
256
+ "loss": 0.3638,
257
+ "step": 400
258
+ },
259
+ {
260
+ "epoch": 0.26,
261
+ "eval_loss": 0.356454461812973,
262
+ "eval_runtime": 903.5298,
263
+ "eval_samples_per_second": 2.214,
264
+ "eval_steps_per_second": 0.277,
265
+ "step": 400
266
+ },
267
+ {
268
+ "epoch": 0.26,
269
+ "learning_rate": 9.596091205211727e-05,
270
+ "loss": 0.3479,
271
+ "step": 410
272
+ },
273
+ {
274
+ "epoch": 0.27,
275
+ "learning_rate": 9.583061889250815e-05,
276
+ "loss": 0.3577,
277
+ "step": 420
278
+ },
279
+ {
280
+ "epoch": 0.28,
281
+ "learning_rate": 9.570032573289904e-05,
282
+ "loss": 0.359,
283
+ "step": 430
284
+ },
285
+ {
286
+ "epoch": 0.28,
287
+ "learning_rate": 9.55700325732899e-05,
288
+ "loss": 0.3584,
289
+ "step": 440
290
+ },
291
+ {
292
+ "epoch": 0.29,
293
+ "learning_rate": 9.543973941368079e-05,
294
+ "loss": 0.3658,
295
+ "step": 450
296
+ },
297
+ {
298
+ "epoch": 0.3,
299
+ "learning_rate": 9.530944625407166e-05,
300
+ "loss": 0.3576,
301
+ "step": 460
302
+ },
303
+ {
304
+ "epoch": 0.3,
305
+ "learning_rate": 9.517915309446255e-05,
306
+ "loss": 0.3695,
307
+ "step": 470
308
+ },
309
+ {
310
+ "epoch": 0.31,
311
+ "learning_rate": 9.504885993485343e-05,
312
+ "loss": 0.3554,
313
+ "step": 480
314
+ },
315
+ {
316
+ "epoch": 0.32,
317
+ "learning_rate": 9.49185667752443e-05,
318
+ "loss": 0.3617,
319
+ "step": 490
320
+ },
321
+ {
322
+ "epoch": 0.32,
323
+ "learning_rate": 9.478827361563518e-05,
324
+ "loss": 0.3568,
325
+ "step": 500
326
+ },
327
+ {
328
+ "epoch": 0.33,
329
+ "learning_rate": 9.465798045602606e-05,
330
+ "loss": 0.3629,
331
+ "step": 510
332
+ },
333
+ {
334
+ "epoch": 0.33,
335
+ "learning_rate": 9.452768729641695e-05,
336
+ "loss": 0.3559,
337
+ "step": 520
338
+ },
339
+ {
340
+ "epoch": 0.34,
341
+ "learning_rate": 9.439739413680782e-05,
342
+ "loss": 0.3467,
343
+ "step": 530
344
+ },
345
+ {
346
+ "epoch": 0.35,
347
+ "learning_rate": 9.42671009771987e-05,
348
+ "loss": 0.3464,
349
+ "step": 540
350
+ },
351
+ {
352
+ "epoch": 0.35,
353
+ "learning_rate": 9.413680781758958e-05,
354
+ "loss": 0.3516,
355
+ "step": 550
356
+ },
357
+ {
358
+ "epoch": 0.36,
359
+ "learning_rate": 9.400651465798047e-05,
360
+ "loss": 0.3564,
361
+ "step": 560
362
+ },
363
+ {
364
+ "epoch": 0.37,
365
+ "learning_rate": 9.387622149837134e-05,
366
+ "loss": 0.3631,
367
+ "step": 570
368
+ },
369
+ {
370
+ "epoch": 0.37,
371
+ "learning_rate": 9.374592833876222e-05,
372
+ "loss": 0.3484,
373
+ "step": 580
374
+ },
375
+ {
376
+ "epoch": 0.38,
377
+ "learning_rate": 9.36156351791531e-05,
378
+ "loss": 0.3461,
379
+ "step": 590
380
+ },
381
+ {
382
+ "epoch": 0.39,
383
+ "learning_rate": 9.348534201954397e-05,
384
+ "loss": 0.3558,
385
+ "step": 600
386
+ },
387
+ {
388
+ "epoch": 0.39,
389
+ "eval_loss": 0.349516361951828,
390
+ "eval_runtime": 892.7558,
391
+ "eval_samples_per_second": 2.24,
392
+ "eval_steps_per_second": 0.28,
393
+ "step": 600
394
+ },
395
+ {
396
+ "epoch": 0.39,
397
+ "learning_rate": 9.335504885993486e-05,
398
+ "loss": 0.3537,
399
+ "step": 610
400
+ },
401
+ {
402
+ "epoch": 0.4,
403
+ "learning_rate": 9.322475570032574e-05,
404
+ "loss": 0.3522,
405
+ "step": 620
406
+ },
407
+ {
408
+ "epoch": 0.41,
409
+ "learning_rate": 9.309446254071663e-05,
410
+ "loss": 0.3557,
411
+ "step": 630
412
+ },
413
+ {
414
+ "epoch": 0.41,
415
+ "learning_rate": 9.296416938110749e-05,
416
+ "loss": 0.3547,
417
+ "step": 640
418
+ },
419
+ {
420
+ "epoch": 0.42,
421
+ "learning_rate": 9.283387622149838e-05,
422
+ "loss": 0.3538,
423
+ "step": 650
424
+ },
425
+ {
426
+ "epoch": 0.42,
427
+ "learning_rate": 9.270358306188925e-05,
428
+ "loss": 0.3532,
429
+ "step": 660
430
+ },
431
+ {
432
+ "epoch": 0.43,
433
+ "learning_rate": 9.257328990228013e-05,
434
+ "loss": 0.3672,
435
+ "step": 670
436
+ },
437
+ {
438
+ "epoch": 0.44,
439
+ "learning_rate": 9.244299674267102e-05,
440
+ "loss": 0.3437,
441
+ "step": 680
442
+ },
443
+ {
444
+ "epoch": 0.44,
445
+ "learning_rate": 9.23127035830619e-05,
446
+ "loss": 0.343,
447
+ "step": 690
448
+ },
449
+ {
450
+ "epoch": 0.45,
451
+ "learning_rate": 9.218241042345277e-05,
452
+ "loss": 0.3669,
453
+ "step": 700
454
+ },
455
+ {
456
+ "epoch": 0.46,
457
+ "learning_rate": 9.205211726384365e-05,
458
+ "loss": 0.3416,
459
+ "step": 710
460
+ },
461
+ {
462
+ "epoch": 0.46,
463
+ "learning_rate": 9.192182410423454e-05,
464
+ "loss": 0.3506,
465
+ "step": 720
466
+ },
467
+ {
468
+ "epoch": 0.47,
469
+ "learning_rate": 9.179153094462541e-05,
470
+ "loss": 0.3521,
471
+ "step": 730
472
+ },
473
+ {
474
+ "epoch": 0.48,
475
+ "learning_rate": 9.16612377850163e-05,
476
+ "loss": 0.3416,
477
+ "step": 740
478
+ },
479
+ {
480
+ "epoch": 0.48,
481
+ "learning_rate": 9.153094462540716e-05,
482
+ "loss": 0.3583,
483
+ "step": 750
484
+ },
485
+ {
486
+ "epoch": 0.49,
487
+ "learning_rate": 9.140065146579805e-05,
488
+ "loss": 0.3664,
489
+ "step": 760
490
+ },
491
+ {
492
+ "epoch": 0.5,
493
+ "learning_rate": 9.127035830618893e-05,
494
+ "loss": 0.3295,
495
+ "step": 770
496
+ },
497
+ {
498
+ "epoch": 0.5,
499
+ "learning_rate": 9.11400651465798e-05,
500
+ "loss": 0.3513,
501
+ "step": 780
502
+ },
503
+ {
504
+ "epoch": 0.51,
505
+ "learning_rate": 9.10097719869707e-05,
506
+ "loss": 0.349,
507
+ "step": 790
508
+ },
509
+ {
510
+ "epoch": 0.51,
511
+ "learning_rate": 9.087947882736156e-05,
512
+ "loss": 0.3512,
513
+ "step": 800
514
+ },
515
+ {
516
+ "epoch": 0.51,
517
+ "eval_loss": 0.3447827398777008,
518
+ "eval_runtime": 910.1893,
519
+ "eval_samples_per_second": 2.197,
520
+ "eval_steps_per_second": 0.275,
521
+ "step": 800
522
+ },
523
+ {
524
+ "epoch": 0.52,
525
+ "learning_rate": 9.074918566775245e-05,
526
+ "loss": 0.358,
527
+ "step": 810
528
+ },
529
+ {
530
+ "epoch": 0.53,
531
+ "learning_rate": 9.061889250814332e-05,
532
+ "loss": 0.339,
533
+ "step": 820
534
+ },
535
+ {
536
+ "epoch": 0.53,
537
+ "learning_rate": 9.048859934853421e-05,
538
+ "loss": 0.3459,
539
+ "step": 830
540
+ },
541
+ {
542
+ "epoch": 0.54,
543
+ "learning_rate": 9.035830618892509e-05,
544
+ "loss": 0.3474,
545
+ "step": 840
546
+ },
547
+ {
548
+ "epoch": 0.55,
549
+ "learning_rate": 9.022801302931597e-05,
550
+ "loss": 0.3464,
551
+ "step": 850
552
+ },
553
+ {
554
+ "epoch": 0.55,
555
+ "learning_rate": 9.009771986970684e-05,
556
+ "loss": 0.3476,
557
+ "step": 860
558
+ },
559
+ {
560
+ "epoch": 0.56,
561
+ "learning_rate": 8.996742671009772e-05,
562
+ "loss": 0.3499,
563
+ "step": 870
564
+ },
565
+ {
566
+ "epoch": 0.57,
567
+ "learning_rate": 8.983713355048861e-05,
568
+ "loss": 0.3371,
569
+ "step": 880
570
+ },
571
+ {
572
+ "epoch": 0.57,
573
+ "learning_rate": 8.970684039087948e-05,
574
+ "loss": 0.354,
575
+ "step": 890
576
+ },
577
+ {
578
+ "epoch": 0.58,
579
+ "learning_rate": 8.957654723127036e-05,
580
+ "loss": 0.3454,
581
+ "step": 900
582
+ },
583
+ {
584
+ "epoch": 0.59,
585
+ "learning_rate": 8.944625407166124e-05,
586
+ "loss": 0.3493,
587
+ "step": 910
588
+ },
589
+ {
590
+ "epoch": 0.59,
591
+ "learning_rate": 8.931596091205213e-05,
592
+ "loss": 0.3412,
593
+ "step": 920
594
+ },
595
+ {
596
+ "epoch": 0.6,
597
+ "learning_rate": 8.9185667752443e-05,
598
+ "loss": 0.3558,
599
+ "step": 930
600
+ },
601
+ {
602
+ "epoch": 0.6,
603
+ "learning_rate": 8.905537459283389e-05,
604
+ "loss": 0.346,
605
+ "step": 940
606
+ },
607
+ {
608
+ "epoch": 0.61,
609
+ "learning_rate": 8.892508143322475e-05,
610
+ "loss": 0.342,
611
+ "step": 950
612
+ },
613
+ {
614
+ "epoch": 0.62,
615
+ "learning_rate": 8.879478827361564e-05,
616
+ "loss": 0.3497,
617
+ "step": 960
618
+ },
619
+ {
620
+ "epoch": 0.62,
621
+ "learning_rate": 8.866449511400652e-05,
622
+ "loss": 0.337,
623
+ "step": 970
624
+ },
625
+ {
626
+ "epoch": 0.63,
627
+ "learning_rate": 8.85342019543974e-05,
628
+ "loss": 0.3435,
629
+ "step": 980
630
+ },
631
+ {
632
+ "epoch": 0.64,
633
+ "learning_rate": 8.840390879478828e-05,
634
+ "loss": 0.3415,
635
+ "step": 990
636
+ },
637
+ {
638
+ "epoch": 0.64,
639
+ "learning_rate": 8.827361563517915e-05,
640
+ "loss": 0.3626,
641
+ "step": 1000
642
+ },
643
+ {
644
+ "epoch": 0.64,
645
+ "eval_loss": 0.3409925103187561,
646
+ "eval_runtime": 932.007,
647
+ "eval_samples_per_second": 2.146,
648
+ "eval_steps_per_second": 0.268,
649
+ "step": 1000
650
+ },
651
+ {
652
+ "epoch": 0.65,
653
+ "learning_rate": 8.814332247557004e-05,
654
+ "loss": 0.3448,
655
+ "step": 1010
656
+ },
657
+ {
658
+ "epoch": 0.66,
659
+ "learning_rate": 8.801302931596091e-05,
660
+ "loss": 0.3459,
661
+ "step": 1020
662
+ },
663
+ {
664
+ "epoch": 0.66,
665
+ "learning_rate": 8.78827361563518e-05,
666
+ "loss": 0.3508,
667
+ "step": 1030
668
+ },
669
+ {
670
+ "epoch": 0.67,
671
+ "learning_rate": 8.775244299674268e-05,
672
+ "loss": 0.3488,
673
+ "step": 1040
674
+ },
675
+ {
676
+ "epoch": 0.68,
677
+ "learning_rate": 8.762214983713355e-05,
678
+ "loss": 0.3488,
679
+ "step": 1050
680
+ },
681
+ {
682
+ "epoch": 0.68,
683
+ "learning_rate": 8.749185667752443e-05,
684
+ "loss": 0.3354,
685
+ "step": 1060
686
+ },
687
+ {
688
+ "epoch": 0.69,
689
+ "learning_rate": 8.73615635179153e-05,
690
+ "loss": 0.3438,
691
+ "step": 1070
692
+ },
693
+ {
694
+ "epoch": 0.69,
695
+ "learning_rate": 8.72312703583062e-05,
696
+ "loss": 0.3519,
697
+ "step": 1080
698
+ },
699
+ {
700
+ "epoch": 0.7,
701
+ "learning_rate": 8.710097719869707e-05,
702
+ "loss": 0.3474,
703
+ "step": 1090
704
+ },
705
+ {
706
+ "epoch": 0.71,
707
+ "learning_rate": 8.697068403908796e-05,
708
+ "loss": 0.3241,
709
+ "step": 1100
710
+ },
711
+ {
712
+ "epoch": 0.71,
713
+ "learning_rate": 8.684039087947882e-05,
714
+ "loss": 0.3421,
715
+ "step": 1110
716
+ },
717
+ {
718
+ "epoch": 0.72,
719
+ "learning_rate": 8.671009771986971e-05,
720
+ "loss": 0.3375,
721
+ "step": 1120
722
+ },
723
+ {
724
+ "epoch": 0.73,
725
+ "learning_rate": 8.657980456026059e-05,
726
+ "loss": 0.3381,
727
+ "step": 1130
728
+ },
729
+ {
730
+ "epoch": 0.73,
731
+ "learning_rate": 8.644951140065147e-05,
732
+ "loss": 0.3456,
733
+ "step": 1140
734
+ },
735
+ {
736
+ "epoch": 0.74,
737
+ "learning_rate": 8.631921824104236e-05,
738
+ "loss": 0.3327,
739
+ "step": 1150
740
+ },
741
+ {
742
+ "epoch": 0.75,
743
+ "learning_rate": 8.618892508143323e-05,
744
+ "loss": 0.3247,
745
+ "step": 1160
746
+ },
747
+ {
748
+ "epoch": 0.75,
749
+ "learning_rate": 8.605863192182411e-05,
750
+ "loss": 0.3422,
751
+ "step": 1170
752
+ },
753
+ {
754
+ "epoch": 0.76,
755
+ "learning_rate": 8.592833876221498e-05,
756
+ "loss": 0.3581,
757
+ "step": 1180
758
+ },
759
+ {
760
+ "epoch": 0.77,
761
+ "learning_rate": 8.579804560260587e-05,
762
+ "loss": 0.3507,
763
+ "step": 1190
764
+ },
765
+ {
766
+ "epoch": 0.77,
767
+ "learning_rate": 8.566775244299675e-05,
768
+ "loss": 0.3438,
769
+ "step": 1200
770
+ },
771
+ {
772
+ "epoch": 0.77,
773
+ "eval_loss": 0.3380748927593231,
774
+ "eval_runtime": 914.1363,
775
+ "eval_samples_per_second": 2.188,
776
+ "eval_steps_per_second": 0.273,
777
+ "step": 1200
778
+ },
779
+ {
780
+ "epoch": 0.78,
781
+ "learning_rate": 8.553745928338763e-05,
782
+ "loss": 0.3502,
783
+ "step": 1210
784
+ },
785
+ {
786
+ "epoch": 0.78,
787
+ "learning_rate": 8.54071661237785e-05,
788
+ "loss": 0.3463,
789
+ "step": 1220
790
+ },
791
+ {
792
+ "epoch": 0.79,
793
+ "learning_rate": 8.527687296416939e-05,
794
+ "loss": 0.3414,
795
+ "step": 1230
796
+ },
797
+ {
798
+ "epoch": 0.8,
799
+ "learning_rate": 8.514657980456027e-05,
800
+ "loss": 0.3491,
801
+ "step": 1240
802
+ },
803
+ {
804
+ "epoch": 0.8,
805
+ "learning_rate": 8.501628664495114e-05,
806
+ "loss": 0.3469,
807
+ "step": 1250
808
+ },
809
+ {
810
+ "epoch": 0.81,
811
+ "learning_rate": 8.488599348534202e-05,
812
+ "loss": 0.3496,
813
+ "step": 1260
814
+ },
815
+ {
816
+ "epoch": 0.82,
817
+ "learning_rate": 8.47557003257329e-05,
818
+ "loss": 0.3436,
819
+ "step": 1270
820
+ },
821
+ {
822
+ "epoch": 0.82,
823
+ "learning_rate": 8.462540716612378e-05,
824
+ "loss": 0.3373,
825
+ "step": 1280
826
+ },
827
+ {
828
+ "epoch": 0.83,
829
+ "learning_rate": 8.449511400651466e-05,
830
+ "loss": 0.3451,
831
+ "step": 1290
832
+ },
833
+ {
834
+ "epoch": 0.84,
835
+ "learning_rate": 8.436482084690555e-05,
836
+ "loss": 0.3422,
837
+ "step": 1300
838
+ },
839
+ {
840
+ "epoch": 0.84,
841
+ "learning_rate": 8.423452768729641e-05,
842
+ "loss": 0.3554,
843
+ "step": 1310
844
+ },
845
+ {
846
+ "epoch": 0.85,
847
+ "learning_rate": 8.41042345276873e-05,
848
+ "loss": 0.3298,
849
+ "step": 1320
850
+ },
851
+ {
852
+ "epoch": 0.86,
853
+ "learning_rate": 8.397394136807818e-05,
854
+ "loss": 0.3489,
855
+ "step": 1330
856
+ },
857
+ {
858
+ "epoch": 0.86,
859
+ "learning_rate": 8.384364820846905e-05,
860
+ "loss": 0.3452,
861
+ "step": 1340
862
+ },
863
+ {
864
+ "epoch": 0.87,
865
+ "learning_rate": 8.371335504885994e-05,
866
+ "loss": 0.3501,
867
+ "step": 1350
868
+ },
869
+ {
870
+ "epoch": 0.87,
871
+ "learning_rate": 8.358306188925081e-05,
872
+ "loss": 0.3395,
873
+ "step": 1360
874
+ },
875
+ {
876
+ "epoch": 0.88,
877
+ "learning_rate": 8.34527687296417e-05,
878
+ "loss": 0.3406,
879
+ "step": 1370
880
+ },
881
+ {
882
+ "epoch": 0.89,
883
+ "learning_rate": 8.332247557003257e-05,
884
+ "loss": 0.3368,
885
+ "step": 1380
886
+ },
887
+ {
888
+ "epoch": 0.89,
889
+ "learning_rate": 8.319218241042346e-05,
890
+ "loss": 0.3252,
891
+ "step": 1390
892
+ },
893
+ {
894
+ "epoch": 0.9,
895
+ "learning_rate": 8.306188925081434e-05,
896
+ "loss": 0.3322,
897
+ "step": 1400
898
+ },
899
+ {
900
+ "epoch": 0.9,
901
+ "eval_loss": 0.33590540289878845,
902
+ "eval_runtime": 910.4397,
903
+ "eval_samples_per_second": 2.197,
904
+ "eval_steps_per_second": 0.275,
905
+ "step": 1400
906
+ },
907
+ {
908
+ "epoch": 0.91,
909
+ "learning_rate": 8.293159609120521e-05,
910
+ "loss": 0.3407,
911
+ "step": 1410
912
+ },
913
+ {
914
+ "epoch": 0.91,
915
+ "learning_rate": 8.280130293159609e-05,
916
+ "loss": 0.3364,
917
+ "step": 1420
918
+ },
919
+ {
920
+ "epoch": 0.92,
921
+ "learning_rate": 8.267100977198698e-05,
922
+ "loss": 0.3412,
923
+ "step": 1430
924
+ },
925
+ {
926
+ "epoch": 0.93,
927
+ "learning_rate": 8.254071661237786e-05,
928
+ "loss": 0.3402,
929
+ "step": 1440
930
+ },
931
+ {
932
+ "epoch": 0.93,
933
+ "learning_rate": 8.241042345276873e-05,
934
+ "loss": 0.3412,
935
+ "step": 1450
936
+ },
937
+ {
938
+ "epoch": 0.94,
939
+ "learning_rate": 8.228013029315962e-05,
940
+ "loss": 0.3268,
941
+ "step": 1460
942
+ },
943
+ {
944
+ "epoch": 0.95,
945
+ "learning_rate": 8.214983713355048e-05,
946
+ "loss": 0.338,
947
+ "step": 1470
948
+ },
949
+ {
950
+ "epoch": 0.95,
951
+ "learning_rate": 8.201954397394137e-05,
952
+ "loss": 0.3403,
953
+ "step": 1480
954
+ },
955
+ {
956
+ "epoch": 0.96,
957
+ "learning_rate": 8.188925081433225e-05,
958
+ "loss": 0.3407,
959
+ "step": 1490
960
+ },
961
+ {
962
+ "epoch": 0.96,
963
+ "learning_rate": 8.175895765472314e-05,
964
+ "loss": 0.34,
965
+ "step": 1500
966
+ },
967
+ {
968
+ "epoch": 0.97,
969
+ "learning_rate": 8.162866449511402e-05,
970
+ "loss": 0.3347,
971
+ "step": 1510
972
+ },
973
+ {
974
+ "epoch": 0.98,
975
+ "learning_rate": 8.149837133550489e-05,
976
+ "loss": 0.3303,
977
+ "step": 1520
978
+ },
979
+ {
980
+ "epoch": 0.98,
981
+ "learning_rate": 8.136807817589577e-05,
982
+ "loss": 0.345,
983
+ "step": 1530
984
+ },
985
+ {
986
+ "epoch": 0.99,
987
+ "learning_rate": 8.123778501628664e-05,
988
+ "loss": 0.3388,
989
+ "step": 1540
990
+ },
991
+ {
992
+ "epoch": 1.0,
993
+ "learning_rate": 8.110749185667753e-05,
994
+ "loss": 0.3303,
995
+ "step": 1550
996
+ },
997
+ {
998
+ "epoch": 1.0,
999
+ "learning_rate": 8.097719869706841e-05,
1000
+ "loss": 0.3308,
1001
+ "step": 1560
1002
+ },
1003
+ {
1004
+ "epoch": 1.01,
1005
+ "learning_rate": 8.084690553745929e-05,
1006
+ "loss": 0.3209,
1007
+ "step": 1570
1008
+ },
1009
+ {
1010
+ "epoch": 1.02,
1011
+ "learning_rate": 8.071661237785016e-05,
1012
+ "loss": 0.3317,
1013
+ "step": 1580
1014
+ },
1015
+ {
1016
+ "epoch": 1.02,
1017
+ "learning_rate": 8.058631921824105e-05,
1018
+ "loss": 0.3375,
1019
+ "step": 1590
1020
+ },
1021
+ {
1022
+ "epoch": 1.03,
1023
+ "learning_rate": 8.045602605863193e-05,
1024
+ "loss": 0.3363,
1025
+ "step": 1600
1026
+ },
1027
+ {
1028
+ "epoch": 1.03,
1029
+ "eval_loss": 0.3338090479373932,
1030
+ "eval_runtime": 887.1836,
1031
+ "eval_samples_per_second": 2.254,
1032
+ "eval_steps_per_second": 0.282,
1033
+ "step": 1600
1034
+ },
1035
+ {
1036
+ "epoch": 1.04,
1037
+ "learning_rate": 8.032573289902282e-05,
1038
+ "loss": 0.3355,
1039
+ "step": 1610
1040
+ },
1041
+ {
1042
+ "epoch": 1.04,
1043
+ "learning_rate": 8.019543973941368e-05,
1044
+ "loss": 0.3267,
1045
+ "step": 1620
1046
+ },
1047
+ {
1048
+ "epoch": 1.05,
1049
+ "learning_rate": 8.006514657980457e-05,
1050
+ "loss": 0.3396,
1051
+ "step": 1630
1052
+ },
1053
+ {
1054
+ "epoch": 1.05,
1055
+ "learning_rate": 7.993485342019544e-05,
1056
+ "loss": 0.327,
1057
+ "step": 1640
1058
+ },
1059
+ {
1060
+ "epoch": 1.06,
1061
+ "learning_rate": 7.980456026058632e-05,
1062
+ "loss": 0.3289,
1063
+ "step": 1650
1064
+ },
1065
+ {
1066
+ "epoch": 1.07,
1067
+ "learning_rate": 7.967426710097721e-05,
1068
+ "loss": 0.3323,
1069
+ "step": 1660
1070
+ },
1071
+ {
1072
+ "epoch": 1.07,
1073
+ "learning_rate": 7.954397394136807e-05,
1074
+ "loss": 0.3454,
1075
+ "step": 1670
1076
+ },
1077
+ {
1078
+ "epoch": 1.08,
1079
+ "learning_rate": 7.941368078175896e-05,
1080
+ "loss": 0.3388,
1081
+ "step": 1680
1082
+ },
1083
+ {
1084
+ "epoch": 1.09,
1085
+ "learning_rate": 7.928338762214984e-05,
1086
+ "loss": 0.3333,
1087
+ "step": 1690
1088
+ },
1089
+ {
1090
+ "epoch": 1.09,
1091
+ "learning_rate": 7.915309446254073e-05,
1092
+ "loss": 0.3398,
1093
+ "step": 1700
1094
+ },
1095
+ {
1096
+ "epoch": 1.1,
1097
+ "learning_rate": 7.90228013029316e-05,
1098
+ "loss": 0.3327,
1099
+ "step": 1710
1100
+ },
1101
+ {
1102
+ "epoch": 1.11,
1103
+ "learning_rate": 7.889250814332248e-05,
1104
+ "loss": 0.3338,
1105
+ "step": 1720
1106
+ },
1107
+ {
1108
+ "epoch": 1.11,
1109
+ "learning_rate": 7.876221498371336e-05,
1110
+ "loss": 0.3237,
1111
+ "step": 1730
1112
+ },
1113
+ {
1114
+ "epoch": 1.12,
1115
+ "learning_rate": 7.863192182410423e-05,
1116
+ "loss": 0.3322,
1117
+ "step": 1740
1118
+ },
1119
+ {
1120
+ "epoch": 1.13,
1121
+ "learning_rate": 7.850162866449512e-05,
1122
+ "loss": 0.3232,
1123
+ "step": 1750
1124
+ },
1125
+ {
1126
+ "epoch": 1.13,
1127
+ "learning_rate": 7.8371335504886e-05,
1128
+ "loss": 0.3385,
1129
+ "step": 1760
1130
+ },
1131
+ {
1132
+ "epoch": 1.14,
1133
+ "learning_rate": 7.824104234527687e-05,
1134
+ "loss": 0.332,
1135
+ "step": 1770
1136
+ },
1137
+ {
1138
+ "epoch": 1.14,
1139
+ "learning_rate": 7.811074918566775e-05,
1140
+ "loss": 0.335,
1141
+ "step": 1780
1142
+ },
1143
+ {
1144
+ "epoch": 1.15,
1145
+ "learning_rate": 7.798045602605864e-05,
1146
+ "loss": 0.3449,
1147
+ "step": 1790
1148
+ },
1149
+ {
1150
+ "epoch": 1.16,
1151
+ "learning_rate": 7.785016286644952e-05,
1152
+ "loss": 0.3322,
1153
+ "step": 1800
1154
+ },
1155
+ {
1156
+ "epoch": 1.16,
1157
+ "eval_loss": 0.3322593867778778,
1158
+ "eval_runtime": 884.0193,
1159
+ "eval_samples_per_second": 2.262,
1160
+ "eval_steps_per_second": 0.283,
1161
+ "step": 1800
1162
+ },
1163
+ {
1164
+ "epoch": 1.16,
1165
+ "learning_rate": 7.771986970684039e-05,
1166
+ "loss": 0.327,
1167
+ "step": 1810
1168
+ },
1169
+ {
1170
+ "epoch": 1.17,
1171
+ "learning_rate": 7.758957654723127e-05,
1172
+ "loss": 0.3532,
1173
+ "step": 1820
1174
+ },
1175
+ {
1176
+ "epoch": 1.18,
1177
+ "learning_rate": 7.745928338762214e-05,
1178
+ "loss": 0.329,
1179
+ "step": 1830
1180
+ },
1181
+ {
1182
+ "epoch": 1.18,
1183
+ "learning_rate": 7.732899022801303e-05,
1184
+ "loss": 0.3252,
1185
+ "step": 1840
1186
+ },
1187
+ {
1188
+ "epoch": 1.19,
1189
+ "learning_rate": 7.719869706840391e-05,
1190
+ "loss": 0.3211,
1191
+ "step": 1850
1192
+ },
1193
+ {
1194
+ "epoch": 1.2,
1195
+ "learning_rate": 7.70684039087948e-05,
1196
+ "loss": 0.3257,
1197
+ "step": 1860
1198
+ },
1199
+ {
1200
+ "epoch": 1.2,
1201
+ "learning_rate": 7.693811074918567e-05,
1202
+ "loss": 0.3251,
1203
+ "step": 1870
1204
+ },
1205
+ {
1206
+ "epoch": 1.21,
1207
+ "learning_rate": 7.680781758957655e-05,
1208
+ "loss": 0.3311,
1209
+ "step": 1880
1210
+ },
1211
+ {
1212
+ "epoch": 1.22,
1213
+ "learning_rate": 7.667752442996743e-05,
1214
+ "loss": 0.3312,
1215
+ "step": 1890
1216
+ },
1217
+ {
1218
+ "epoch": 1.22,
1219
+ "learning_rate": 7.654723127035832e-05,
1220
+ "loss": 0.3228,
1221
+ "step": 1900
1222
+ },
1223
+ {
1224
+ "epoch": 1.23,
1225
+ "learning_rate": 7.641693811074919e-05,
1226
+ "loss": 0.3301,
1227
+ "step": 1910
1228
+ },
1229
+ {
1230
+ "epoch": 1.23,
1231
+ "learning_rate": 7.628664495114007e-05,
1232
+ "loss": 0.3352,
1233
+ "step": 1920
1234
+ },
1235
+ {
1236
+ "epoch": 1.24,
1237
+ "learning_rate": 7.615635179153094e-05,
1238
+ "loss": 0.3355,
1239
+ "step": 1930
1240
+ },
1241
+ {
1242
+ "epoch": 1.25,
1243
+ "learning_rate": 7.602605863192182e-05,
1244
+ "loss": 0.3326,
1245
+ "step": 1940
1246
+ },
1247
+ {
1248
+ "epoch": 1.25,
1249
+ "learning_rate": 7.589576547231271e-05,
1250
+ "loss": 0.3323,
1251
+ "step": 1950
1252
+ },
1253
+ {
1254
+ "epoch": 1.26,
1255
+ "learning_rate": 7.576547231270359e-05,
1256
+ "loss": 0.3337,
1257
+ "step": 1960
1258
+ },
1259
+ {
1260
+ "epoch": 1.27,
1261
+ "learning_rate": 7.563517915309448e-05,
1262
+ "loss": 0.3393,
1263
+ "step": 1970
1264
+ },
1265
+ {
1266
+ "epoch": 1.27,
1267
+ "learning_rate": 7.550488599348534e-05,
1268
+ "loss": 0.349,
1269
+ "step": 1980
1270
+ },
1271
+ {
1272
+ "epoch": 1.28,
1273
+ "learning_rate": 7.537459283387623e-05,
1274
+ "loss": 0.3253,
1275
+ "step": 1990
1276
+ },
1277
+ {
1278
+ "epoch": 1.29,
1279
+ "learning_rate": 7.52442996742671e-05,
1280
+ "loss": 0.3285,
1281
+ "step": 2000
1282
+ },
1283
+ {
1284
+ "epoch": 1.29,
1285
+ "eval_loss": 0.33141371607780457,
1286
+ "eval_runtime": 888.1239,
1287
+ "eval_samples_per_second": 2.252,
1288
+ "eval_steps_per_second": 0.281,
1289
+ "step": 2000
1290
+ },
1291
+ {
1292
+ "epoch": 1.29,
1293
+ "learning_rate": 7.511400651465798e-05,
1294
+ "loss": 0.3363,
1295
+ "step": 2010
1296
+ },
1297
+ {
1298
+ "epoch": 1.3,
1299
+ "learning_rate": 7.498371335504887e-05,
1300
+ "loss": 0.3312,
1301
+ "step": 2020
1302
+ },
1303
+ {
1304
+ "epoch": 1.31,
1305
+ "learning_rate": 7.485342019543973e-05,
1306
+ "loss": 0.3201,
1307
+ "step": 2030
1308
+ },
1309
+ {
1310
+ "epoch": 1.31,
1311
+ "learning_rate": 7.472312703583062e-05,
1312
+ "loss": 0.3274,
1313
+ "step": 2040
1314
+ },
1315
+ {
1316
+ "epoch": 1.32,
1317
+ "learning_rate": 7.45928338762215e-05,
1318
+ "loss": 0.3366,
1319
+ "step": 2050
1320
+ },
1321
+ {
1322
+ "epoch": 1.32,
1323
+ "learning_rate": 7.446254071661239e-05,
1324
+ "loss": 0.3242,
1325
+ "step": 2060
1326
+ },
1327
+ {
1328
+ "epoch": 1.33,
1329
+ "learning_rate": 7.433224755700326e-05,
1330
+ "loss": 0.3261,
1331
+ "step": 2070
1332
+ },
1333
+ {
1334
+ "epoch": 1.34,
1335
+ "learning_rate": 7.420195439739414e-05,
1336
+ "loss": 0.3246,
1337
+ "step": 2080
1338
+ },
1339
+ {
1340
+ "epoch": 1.34,
1341
+ "learning_rate": 7.407166123778502e-05,
1342
+ "loss": 0.3264,
1343
+ "step": 2090
1344
+ },
1345
+ {
1346
+ "epoch": 1.35,
1347
+ "learning_rate": 7.39413680781759e-05,
1348
+ "loss": 0.3184,
1349
+ "step": 2100
1350
+ },
1351
+ {
1352
+ "epoch": 1.36,
1353
+ "learning_rate": 7.381107491856678e-05,
1354
+ "loss": 0.3332,
1355
+ "step": 2110
1356
+ },
1357
+ {
1358
+ "epoch": 1.36,
1359
+ "learning_rate": 7.368078175895766e-05,
1360
+ "loss": 0.3264,
1361
+ "step": 2120
1362
+ },
1363
+ {
1364
+ "epoch": 1.37,
1365
+ "learning_rate": 7.355048859934853e-05,
1366
+ "loss": 0.3331,
1367
+ "step": 2130
1368
+ },
1369
+ {
1370
+ "epoch": 1.38,
1371
+ "learning_rate": 7.342019543973941e-05,
1372
+ "loss": 0.3369,
1373
+ "step": 2140
1374
+ },
1375
+ {
1376
+ "epoch": 1.38,
1377
+ "learning_rate": 7.32899022801303e-05,
1378
+ "loss": 0.3288,
1379
+ "step": 2150
1380
+ },
1381
+ {
1382
+ "epoch": 1.39,
1383
+ "learning_rate": 7.315960912052118e-05,
1384
+ "loss": 0.3321,
1385
+ "step": 2160
1386
+ },
1387
+ {
1388
+ "epoch": 1.4,
1389
+ "learning_rate": 7.302931596091206e-05,
1390
+ "loss": 0.333,
1391
+ "step": 2170
1392
+ },
1393
+ {
1394
+ "epoch": 1.4,
1395
+ "learning_rate": 7.289902280130293e-05,
1396
+ "loss": 0.3181,
1397
+ "step": 2180
1398
+ },
1399
+ {
1400
+ "epoch": 1.41,
1401
+ "learning_rate": 7.276872964169382e-05,
1402
+ "loss": 0.3307,
1403
+ "step": 2190
1404
+ },
1405
+ {
1406
+ "epoch": 1.41,
1407
+ "learning_rate": 7.263843648208469e-05,
1408
+ "loss": 0.331,
1409
+ "step": 2200
1410
+ },
1411
+ {
1412
+ "epoch": 1.41,
1413
+ "eval_loss": 0.32971131801605225,
1414
+ "eval_runtime": 870.8838,
1415
+ "eval_samples_per_second": 2.297,
1416
+ "eval_steps_per_second": 0.287,
1417
+ "step": 2200
1418
+ },
1419
+ {
1420
+ "epoch": 1.42,
1421
+ "learning_rate": 7.250814332247557e-05,
1422
+ "loss": 0.3367,
1423
+ "step": 2210
1424
+ },
1425
+ {
1426
+ "epoch": 1.43,
1427
+ "learning_rate": 7.237785016286646e-05,
1428
+ "loss": 0.316,
1429
+ "step": 2220
1430
+ },
1431
+ {
1432
+ "epoch": 1.43,
1433
+ "learning_rate": 7.224755700325733e-05,
1434
+ "loss": 0.3266,
1435
+ "step": 2230
1436
+ },
1437
+ {
1438
+ "epoch": 1.44,
1439
+ "learning_rate": 7.211726384364821e-05,
1440
+ "loss": 0.3178,
1441
+ "step": 2240
1442
+ },
1443
+ {
1444
+ "epoch": 1.45,
1445
+ "learning_rate": 7.198697068403909e-05,
1446
+ "loss": 0.3333,
1447
+ "step": 2250
1448
+ },
1449
+ {
1450
+ "epoch": 1.45,
1451
+ "learning_rate": 7.185667752442998e-05,
1452
+ "loss": 0.3188,
1453
+ "step": 2260
1454
+ },
1455
+ {
1456
+ "epoch": 1.46,
1457
+ "learning_rate": 7.172638436482085e-05,
1458
+ "loss": 0.3415,
1459
+ "step": 2270
1460
+ },
1461
+ {
1462
+ "epoch": 1.47,
1463
+ "learning_rate": 7.159609120521173e-05,
1464
+ "loss": 0.3376,
1465
+ "step": 2280
1466
+ },
1467
+ {
1468
+ "epoch": 1.47,
1469
+ "learning_rate": 7.14657980456026e-05,
1470
+ "loss": 0.332,
1471
+ "step": 2290
1472
+ },
1473
+ {
1474
+ "epoch": 1.48,
1475
+ "learning_rate": 7.133550488599348e-05,
1476
+ "loss": 0.3277,
1477
+ "step": 2300
1478
+ },
1479
+ {
1480
+ "epoch": 1.49,
1481
+ "learning_rate": 7.120521172638437e-05,
1482
+ "loss": 0.3387,
1483
+ "step": 2310
1484
+ },
1485
+ {
1486
+ "epoch": 1.49,
1487
+ "learning_rate": 7.107491856677525e-05,
1488
+ "loss": 0.3243,
1489
+ "step": 2320
1490
+ },
1491
+ {
1492
+ "epoch": 1.5,
1493
+ "learning_rate": 7.094462540716614e-05,
1494
+ "loss": 0.3426,
1495
+ "step": 2330
1496
+ },
1497
+ {
1498
+ "epoch": 1.5,
1499
+ "learning_rate": 7.0814332247557e-05,
1500
+ "loss": 0.3169,
1501
+ "step": 2340
1502
+ },
1503
+ {
1504
+ "epoch": 1.51,
1505
+ "learning_rate": 7.068403908794789e-05,
1506
+ "loss": 0.3299,
1507
+ "step": 2350
1508
+ },
1509
+ {
1510
+ "epoch": 1.52,
1511
+ "learning_rate": 7.055374592833876e-05,
1512
+ "loss": 0.3174,
1513
+ "step": 2360
1514
+ },
1515
+ {
1516
+ "epoch": 1.52,
1517
+ "learning_rate": 7.042345276872965e-05,
1518
+ "loss": 0.3286,
1519
+ "step": 2370
1520
+ },
1521
+ {
1522
+ "epoch": 1.53,
1523
+ "learning_rate": 7.029315960912053e-05,
1524
+ "loss": 0.3128,
1525
+ "step": 2380
1526
+ },
1527
+ {
1528
+ "epoch": 1.54,
1529
+ "learning_rate": 7.01628664495114e-05,
1530
+ "loss": 0.3308,
1531
+ "step": 2390
1532
+ },
1533
+ {
1534
+ "epoch": 1.54,
1535
+ "learning_rate": 7.003257328990228e-05,
1536
+ "loss": 0.3474,
1537
+ "step": 2400
1538
+ },
1539
+ {
1540
+ "epoch": 1.54,
1541
+ "eval_loss": 0.3283294439315796,
1542
+ "eval_runtime": 868.3719,
1543
+ "eval_samples_per_second": 2.303,
1544
+ "eval_steps_per_second": 0.288,
1545
+ "step": 2400
1546
+ },
1547
+ {
1548
+ "epoch": 1.55,
1549
+ "learning_rate": 6.990228013029316e-05,
1550
+ "loss": 0.3145,
1551
+ "step": 2410
1552
+ },
1553
+ {
1554
+ "epoch": 1.56,
1555
+ "learning_rate": 6.977198697068405e-05,
1556
+ "loss": 0.3333,
1557
+ "step": 2420
1558
+ },
1559
+ {
1560
+ "epoch": 1.56,
1561
+ "learning_rate": 6.964169381107492e-05,
1562
+ "loss": 0.3307,
1563
+ "step": 2430
1564
+ },
1565
+ {
1566
+ "epoch": 1.57,
1567
+ "learning_rate": 6.95114006514658e-05,
1568
+ "loss": 0.3325,
1569
+ "step": 2440
1570
+ },
1571
+ {
1572
+ "epoch": 1.58,
1573
+ "learning_rate": 6.938110749185668e-05,
1574
+ "loss": 0.327,
1575
+ "step": 2450
1576
+ },
1577
+ {
1578
+ "epoch": 1.58,
1579
+ "learning_rate": 6.925081433224757e-05,
1580
+ "loss": 0.332,
1581
+ "step": 2460
1582
+ },
1583
+ {
1584
+ "epoch": 1.59,
1585
+ "learning_rate": 6.912052117263844e-05,
1586
+ "loss": 0.3229,
1587
+ "step": 2470
1588
+ },
1589
+ {
1590
+ "epoch": 1.59,
1591
+ "learning_rate": 6.899022801302932e-05,
1592
+ "loss": 0.3342,
1593
+ "step": 2480
1594
+ },
1595
+ {
1596
+ "epoch": 1.6,
1597
+ "learning_rate": 6.885993485342019e-05,
1598
+ "loss": 0.3267,
1599
+ "step": 2490
1600
+ },
1601
+ {
1602
+ "epoch": 1.61,
1603
+ "learning_rate": 6.872964169381107e-05,
1604
+ "loss": 0.3347,
1605
+ "step": 2500
1606
+ },
1607
+ {
1608
+ "epoch": 1.61,
1609
+ "learning_rate": 6.859934853420196e-05,
1610
+ "loss": 0.3334,
1611
+ "step": 2510
1612
+ },
1613
+ {
1614
+ "epoch": 1.62,
1615
+ "learning_rate": 6.846905537459283e-05,
1616
+ "loss": 0.3192,
1617
+ "step": 2520
1618
+ },
1619
+ {
1620
+ "epoch": 1.63,
1621
+ "learning_rate": 6.833876221498372e-05,
1622
+ "loss": 0.3299,
1623
+ "step": 2530
1624
+ },
1625
+ {
1626
+ "epoch": 1.63,
1627
+ "learning_rate": 6.820846905537459e-05,
1628
+ "loss": 0.3451,
1629
+ "step": 2540
1630
+ },
1631
+ {
1632
+ "epoch": 1.64,
1633
+ "learning_rate": 6.807817589576548e-05,
1634
+ "loss": 0.3161,
1635
+ "step": 2550
1636
+ },
1637
+ {
1638
+ "epoch": 1.65,
1639
+ "learning_rate": 6.794788273615635e-05,
1640
+ "loss": 0.3248,
1641
+ "step": 2560
1642
+ },
1643
+ {
1644
+ "epoch": 1.65,
1645
+ "learning_rate": 6.781758957654724e-05,
1646
+ "loss": 0.3206,
1647
+ "step": 2570
1648
+ },
1649
+ {
1650
+ "epoch": 1.66,
1651
+ "learning_rate": 6.768729641693812e-05,
1652
+ "loss": 0.339,
1653
+ "step": 2580
1654
+ },
1655
+ {
1656
+ "epoch": 1.67,
1657
+ "learning_rate": 6.7557003257329e-05,
1658
+ "loss": 0.3397,
1659
+ "step": 2590
1660
+ },
1661
+ {
1662
+ "epoch": 1.67,
1663
+ "learning_rate": 6.742671009771987e-05,
1664
+ "loss": 0.3263,
1665
+ "step": 2600
1666
+ },
1667
+ {
1668
+ "epoch": 1.67,
1669
+ "eval_loss": 0.3271417021751404,
1670
+ "eval_runtime": 871.5909,
1671
+ "eval_samples_per_second": 2.295,
1672
+ "eval_steps_per_second": 0.287,
1673
+ "step": 2600
1674
+ },
1675
+ {
1676
+ "epoch": 1.68,
1677
+ "learning_rate": 6.729641693811075e-05,
1678
+ "loss": 0.3235,
1679
+ "step": 2610
1680
+ },
1681
+ {
1682
+ "epoch": 1.68,
1683
+ "learning_rate": 6.716612377850164e-05,
1684
+ "loss": 0.3206,
1685
+ "step": 2620
1686
+ },
1687
+ {
1688
+ "epoch": 1.69,
1689
+ "learning_rate": 6.703583061889251e-05,
1690
+ "loss": 0.3229,
1691
+ "step": 2630
1692
+ },
1693
+ {
1694
+ "epoch": 1.7,
1695
+ "learning_rate": 6.69055374592834e-05,
1696
+ "loss": 0.3146,
1697
+ "step": 2640
1698
+ },
1699
+ {
1700
+ "epoch": 1.7,
1701
+ "learning_rate": 6.677524429967426e-05,
1702
+ "loss": 0.3192,
1703
+ "step": 2650
1704
+ },
1705
+ {
1706
+ "epoch": 1.71,
1707
+ "learning_rate": 6.664495114006515e-05,
1708
+ "loss": 0.3244,
1709
+ "step": 2660
1710
+ },
1711
+ {
1712
+ "epoch": 1.72,
1713
+ "learning_rate": 6.651465798045603e-05,
1714
+ "loss": 0.328,
1715
+ "step": 2670
1716
+ },
1717
+ {
1718
+ "epoch": 1.72,
1719
+ "learning_rate": 6.63843648208469e-05,
1720
+ "loss": 0.326,
1721
+ "step": 2680
1722
+ },
1723
+ {
1724
+ "epoch": 1.73,
1725
+ "learning_rate": 6.62540716612378e-05,
1726
+ "loss": 0.3402,
1727
+ "step": 2690
1728
+ },
1729
+ {
1730
+ "epoch": 1.74,
1731
+ "learning_rate": 6.612377850162866e-05,
1732
+ "loss": 0.3314,
1733
+ "step": 2700
1734
+ },
1735
+ {
1736
+ "epoch": 1.74,
1737
+ "learning_rate": 6.599348534201955e-05,
1738
+ "loss": 0.3196,
1739
+ "step": 2710
1740
+ },
1741
+ {
1742
+ "epoch": 1.75,
1743
+ "learning_rate": 6.586319218241042e-05,
1744
+ "loss": 0.3342,
1745
+ "step": 2720
1746
+ },
1747
+ {
1748
+ "epoch": 1.76,
1749
+ "learning_rate": 6.573289902280131e-05,
1750
+ "loss": 0.3211,
1751
+ "step": 2730
1752
+ },
1753
+ {
1754
+ "epoch": 1.76,
1755
+ "learning_rate": 6.560260586319219e-05,
1756
+ "loss": 0.33,
1757
+ "step": 2740
1758
+ },
1759
+ {
1760
+ "epoch": 1.77,
1761
+ "learning_rate": 6.547231270358307e-05,
1762
+ "loss": 0.3265,
1763
+ "step": 2750
1764
+ },
1765
+ {
1766
+ "epoch": 1.77,
1767
+ "learning_rate": 6.534201954397394e-05,
1768
+ "loss": 0.3367,
1769
+ "step": 2760
1770
+ },
1771
+ {
1772
+ "epoch": 1.78,
1773
+ "learning_rate": 6.521172638436482e-05,
1774
+ "loss": 0.3317,
1775
+ "step": 2770
1776
+ },
1777
+ {
1778
+ "epoch": 1.79,
1779
+ "learning_rate": 6.508143322475571e-05,
1780
+ "loss": 0.3272,
1781
+ "step": 2780
1782
+ },
1783
+ {
1784
+ "epoch": 1.79,
1785
+ "learning_rate": 6.495114006514658e-05,
1786
+ "loss": 0.3195,
1787
+ "step": 2790
1788
+ },
1789
+ {
1790
+ "epoch": 1.8,
1791
+ "learning_rate": 6.482084690553746e-05,
1792
+ "loss": 0.3347,
1793
+ "step": 2800
1794
+ },
1795
+ {
1796
+ "epoch": 1.8,
1797
+ "eval_loss": 0.3261171281337738,
1798
+ "eval_runtime": 873.3799,
1799
+ "eval_samples_per_second": 2.29,
1800
+ "eval_steps_per_second": 0.286,
1801
+ "step": 2800
1802
+ },
1803
+ {
1804
+ "epoch": 1.81,
1805
+ "learning_rate": 6.469055374592834e-05,
1806
+ "loss": 0.329,
1807
+ "step": 2810
1808
+ },
1809
+ {
1810
+ "epoch": 1.81,
1811
+ "learning_rate": 6.456026058631922e-05,
1812
+ "loss": 0.3143,
1813
+ "step": 2820
1814
+ },
1815
+ {
1816
+ "epoch": 1.82,
1817
+ "learning_rate": 6.44299674267101e-05,
1818
+ "loss": 0.3325,
1819
+ "step": 2830
1820
+ },
1821
+ {
1822
+ "epoch": 1.83,
1823
+ "learning_rate": 6.429967426710099e-05,
1824
+ "loss": 0.3362,
1825
+ "step": 2840
1826
+ },
1827
+ {
1828
+ "epoch": 1.83,
1829
+ "learning_rate": 6.416938110749185e-05,
1830
+ "loss": 0.3305,
1831
+ "step": 2850
1832
+ },
1833
+ {
1834
+ "epoch": 1.84,
1835
+ "learning_rate": 6.403908794788274e-05,
1836
+ "loss": 0.3272,
1837
+ "step": 2860
1838
+ },
1839
+ {
1840
+ "epoch": 1.85,
1841
+ "learning_rate": 6.390879478827362e-05,
1842
+ "loss": 0.3253,
1843
+ "step": 2870
1844
+ },
1845
+ {
1846
+ "epoch": 1.85,
1847
+ "learning_rate": 6.37785016286645e-05,
1848
+ "loss": 0.3325,
1849
+ "step": 2880
1850
+ },
1851
+ {
1852
+ "epoch": 1.86,
1853
+ "learning_rate": 6.364820846905538e-05,
1854
+ "loss": 0.3235,
1855
+ "step": 2890
1856
+ },
1857
+ {
1858
+ "epoch": 1.86,
1859
+ "learning_rate": 6.351791530944625e-05,
1860
+ "loss": 0.3357,
1861
+ "step": 2900
1862
+ },
1863
+ {
1864
+ "epoch": 1.87,
1865
+ "learning_rate": 6.338762214983714e-05,
1866
+ "loss": 0.3198,
1867
+ "step": 2910
1868
+ },
1869
+ {
1870
+ "epoch": 1.88,
1871
+ "learning_rate": 6.325732899022801e-05,
1872
+ "loss": 0.3232,
1873
+ "step": 2920
1874
+ },
1875
+ {
1876
+ "epoch": 1.88,
1877
+ "learning_rate": 6.31270358306189e-05,
1878
+ "loss": 0.3145,
1879
+ "step": 2930
1880
+ },
1881
+ {
1882
+ "epoch": 1.89,
1883
+ "learning_rate": 6.299674267100978e-05,
1884
+ "loss": 0.3282,
1885
+ "step": 2940
1886
+ },
1887
+ {
1888
+ "epoch": 1.9,
1889
+ "learning_rate": 6.286644951140065e-05,
1890
+ "loss": 0.3248,
1891
+ "step": 2950
1892
+ },
1893
+ {
1894
+ "epoch": 1.9,
1895
+ "learning_rate": 6.273615635179153e-05,
1896
+ "loss": 0.3311,
1897
+ "step": 2960
1898
+ },
1899
+ {
1900
+ "epoch": 1.91,
1901
+ "learning_rate": 6.26058631921824e-05,
1902
+ "loss": 0.3277,
1903
+ "step": 2970
1904
+ },
1905
+ {
1906
+ "epoch": 1.92,
1907
+ "learning_rate": 6.24755700325733e-05,
1908
+ "loss": 0.322,
1909
+ "step": 2980
1910
+ },
1911
+ {
1912
+ "epoch": 1.92,
1913
+ "learning_rate": 6.234527687296417e-05,
1914
+ "loss": 0.3272,
1915
+ "step": 2990
1916
+ },
1917
+ {
1918
+ "epoch": 1.93,
1919
+ "learning_rate": 6.221498371335506e-05,
1920
+ "loss": 0.3381,
1921
+ "step": 3000
1922
+ },
1923
+ {
1924
+ "epoch": 1.93,
1925
+ "eval_loss": 0.32507359981536865,
1926
+ "eval_runtime": 878.4405,
1927
+ "eval_samples_per_second": 2.277,
1928
+ "eval_steps_per_second": 0.285,
1929
+ "step": 3000
1930
+ },
1931
+ {
1932
+ "epoch": 1.94,
1933
+ "learning_rate": 6.208469055374592e-05,
1934
+ "loss": 0.3241,
1935
+ "step": 3010
1936
+ },
1937
+ {
1938
+ "epoch": 1.94,
1939
+ "learning_rate": 6.195439739413681e-05,
1940
+ "loss": 0.32,
1941
+ "step": 3020
1942
+ },
1943
+ {
1944
+ "epoch": 1.95,
1945
+ "learning_rate": 6.182410423452769e-05,
1946
+ "loss": 0.3166,
1947
+ "step": 3030
1948
+ },
1949
+ {
1950
+ "epoch": 1.95,
1951
+ "learning_rate": 6.169381107491858e-05,
1952
+ "loss": 0.3349,
1953
+ "step": 3040
1954
+ },
1955
+ {
1956
+ "epoch": 1.96,
1957
+ "learning_rate": 6.156351791530946e-05,
1958
+ "loss": 0.3269,
1959
+ "step": 3050
1960
+ },
1961
+ {
1962
+ "epoch": 1.97,
1963
+ "learning_rate": 6.143322475570033e-05,
1964
+ "loss": 0.34,
1965
+ "step": 3060
1966
+ },
1967
+ {
1968
+ "epoch": 1.97,
1969
+ "learning_rate": 6.130293159609121e-05,
1970
+ "loss": 0.3197,
1971
+ "step": 3070
1972
+ },
1973
+ {
1974
+ "epoch": 1.98,
1975
+ "learning_rate": 6.117263843648208e-05,
1976
+ "loss": 0.3203,
1977
+ "step": 3080
1978
+ },
1979
+ {
1980
+ "epoch": 1.99,
1981
+ "learning_rate": 6.104234527687297e-05,
1982
+ "loss": 0.3226,
1983
+ "step": 3090
1984
+ },
1985
+ {
1986
+ "epoch": 1.99,
1987
+ "learning_rate": 6.091205211726385e-05,
1988
+ "loss": 0.3192,
1989
+ "step": 3100
1990
+ },
1991
+ {
1992
+ "epoch": 2.0,
1993
+ "learning_rate": 6.0781758957654725e-05,
1994
+ "loss": 0.3321,
1995
+ "step": 3110
1996
+ },
1997
+ {
1998
+ "epoch": 2.01,
1999
+ "learning_rate": 6.06514657980456e-05,
2000
+ "loss": 0.3236,
2001
+ "step": 3120
2002
+ },
2003
+ {
2004
+ "epoch": 2.01,
2005
+ "learning_rate": 6.0521172638436484e-05,
2006
+ "loss": 0.3049,
2007
+ "step": 3130
2008
+ },
2009
+ {
2010
+ "epoch": 2.02,
2011
+ "learning_rate": 6.0390879478827367e-05,
2012
+ "loss": 0.3226,
2013
+ "step": 3140
2014
+ },
2015
+ {
2016
+ "epoch": 2.03,
2017
+ "learning_rate": 6.026058631921825e-05,
2018
+ "loss": 0.3376,
2019
+ "step": 3150
2020
+ },
2021
+ {
2022
+ "epoch": 2.03,
2023
+ "learning_rate": 6.013029315960912e-05,
2024
+ "loss": 0.3071,
2025
+ "step": 3160
2026
+ },
2027
+ {
2028
+ "epoch": 2.04,
2029
+ "learning_rate": 6e-05,
2030
+ "loss": 0.3177,
2031
+ "step": 3170
2032
+ },
2033
+ {
2034
+ "epoch": 2.05,
2035
+ "learning_rate": 5.9869706840390884e-05,
2036
+ "loss": 0.3155,
2037
+ "step": 3180
2038
+ },
2039
+ {
2040
+ "epoch": 2.05,
2041
+ "learning_rate": 5.973941368078177e-05,
2042
+ "loss": 0.3227,
2043
+ "step": 3190
2044
+ },
2045
+ {
2046
+ "epoch": 2.06,
2047
+ "learning_rate": 5.960912052117264e-05,
2048
+ "loss": 0.3064,
2049
+ "step": 3200
2050
+ },
2051
+ {
2052
+ "epoch": 2.06,
2053
+ "eval_loss": 0.3246903121471405,
2054
+ "eval_runtime": 877.3437,
2055
+ "eval_samples_per_second": 2.28,
2056
+ "eval_steps_per_second": 0.285,
2057
+ "step": 3200
2058
+ },
2059
+ {
2060
+ "epoch": 2.06,
2061
+ "learning_rate": 5.947882736156352e-05,
2062
+ "loss": 0.3207,
2063
+ "step": 3210
2064
+ },
2065
+ {
2066
+ "epoch": 2.07,
2067
+ "learning_rate": 5.9348534201954395e-05,
2068
+ "loss": 0.3288,
2069
+ "step": 3220
2070
+ },
2071
+ {
2072
+ "epoch": 2.08,
2073
+ "learning_rate": 5.921824104234528e-05,
2074
+ "loss": 0.3151,
2075
+ "step": 3230
2076
+ },
2077
+ {
2078
+ "epoch": 2.08,
2079
+ "learning_rate": 5.908794788273616e-05,
2080
+ "loss": 0.3134,
2081
+ "step": 3240
2082
+ },
2083
+ {
2084
+ "epoch": 2.09,
2085
+ "learning_rate": 5.8957654723127044e-05,
2086
+ "loss": 0.3009,
2087
+ "step": 3250
2088
+ },
2089
+ {
2090
+ "epoch": 2.1,
2091
+ "learning_rate": 5.882736156351791e-05,
2092
+ "loss": 0.3159,
2093
+ "step": 3260
2094
+ },
2095
+ {
2096
+ "epoch": 2.1,
2097
+ "learning_rate": 5.8697068403908796e-05,
2098
+ "loss": 0.3248,
2099
+ "step": 3270
2100
+ },
2101
+ {
2102
+ "epoch": 2.11,
2103
+ "learning_rate": 5.856677524429968e-05,
2104
+ "loss": 0.333,
2105
+ "step": 3280
2106
+ },
2107
+ {
2108
+ "epoch": 2.12,
2109
+ "learning_rate": 5.8436482084690555e-05,
2110
+ "loss": 0.3173,
2111
+ "step": 3290
2112
+ },
2113
+ {
2114
+ "epoch": 2.12,
2115
+ "learning_rate": 5.830618892508144e-05,
2116
+ "loss": 0.3145,
2117
+ "step": 3300
2118
+ },
2119
+ {
2120
+ "epoch": 2.13,
2121
+ "learning_rate": 5.8175895765472314e-05,
2122
+ "loss": 0.3273,
2123
+ "step": 3310
2124
+ },
2125
+ {
2126
+ "epoch": 2.14,
2127
+ "learning_rate": 5.804560260586319e-05,
2128
+ "loss": 0.338,
2129
+ "step": 3320
2130
+ },
2131
+ {
2132
+ "epoch": 2.14,
2133
+ "learning_rate": 5.791530944625407e-05,
2134
+ "loss": 0.3159,
2135
+ "step": 3330
2136
+ },
2137
+ {
2138
+ "epoch": 2.15,
2139
+ "learning_rate": 5.7785016286644955e-05,
2140
+ "loss": 0.3095,
2141
+ "step": 3340
2142
+ },
2143
+ {
2144
+ "epoch": 2.15,
2145
+ "learning_rate": 5.765472312703584e-05,
2146
+ "loss": 0.3174,
2147
+ "step": 3350
2148
+ },
2149
+ {
2150
+ "epoch": 2.16,
2151
+ "learning_rate": 5.752442996742671e-05,
2152
+ "loss": 0.3174,
2153
+ "step": 3360
2154
+ },
2155
+ {
2156
+ "epoch": 2.17,
2157
+ "learning_rate": 5.739413680781759e-05,
2158
+ "loss": 0.3126,
2159
+ "step": 3370
2160
+ },
2161
+ {
2162
+ "epoch": 2.17,
2163
+ "learning_rate": 5.726384364820847e-05,
2164
+ "loss": 0.3153,
2165
+ "step": 3380
2166
+ },
2167
+ {
2168
+ "epoch": 2.18,
2169
+ "learning_rate": 5.713355048859935e-05,
2170
+ "loss": 0.3231,
2171
+ "step": 3390
2172
+ },
2173
+ {
2174
+ "epoch": 2.19,
2175
+ "learning_rate": 5.700325732899023e-05,
2176
+ "loss": 0.3228,
2177
+ "step": 3400
2178
+ },
2179
+ {
2180
+ "epoch": 2.19,
2181
+ "eval_loss": 0.3241334855556488,
2182
+ "eval_runtime": 882.9043,
2183
+ "eval_samples_per_second": 2.265,
2184
+ "eval_steps_per_second": 0.283,
2185
+ "step": 3400
2186
+ },
2187
+ {
2188
+ "epoch": 2.19,
2189
+ "learning_rate": 5.6872964169381115e-05,
2190
+ "loss": 0.3246,
2191
+ "step": 3410
2192
+ },
2193
+ {
2194
+ "epoch": 2.2,
2195
+ "learning_rate": 5.6742671009771984e-05,
2196
+ "loss": 0.3053,
2197
+ "step": 3420
2198
+ },
2199
+ {
2200
+ "epoch": 2.21,
2201
+ "learning_rate": 5.661237785016287e-05,
2202
+ "loss": 0.3265,
2203
+ "step": 3430
2204
+ },
2205
+ {
2206
+ "epoch": 2.21,
2207
+ "learning_rate": 5.648208469055375e-05,
2208
+ "loss": 0.3229,
2209
+ "step": 3440
2210
+ },
2211
+ {
2212
+ "epoch": 2.22,
2213
+ "learning_rate": 5.635179153094463e-05,
2214
+ "loss": 0.3118,
2215
+ "step": 3450
2216
+ },
2217
+ {
2218
+ "epoch": 2.23,
2219
+ "learning_rate": 5.6221498371335515e-05,
2220
+ "loss": 0.3093,
2221
+ "step": 3460
2222
+ },
2223
+ {
2224
+ "epoch": 2.23,
2225
+ "learning_rate": 5.6091205211726385e-05,
2226
+ "loss": 0.3216,
2227
+ "step": 3470
2228
+ },
2229
+ {
2230
+ "epoch": 2.24,
2231
+ "learning_rate": 5.596091205211727e-05,
2232
+ "loss": 0.3226,
2233
+ "step": 3480
2234
+ },
2235
+ {
2236
+ "epoch": 2.24,
2237
+ "learning_rate": 5.5830618892508143e-05,
2238
+ "loss": 0.3248,
2239
+ "step": 3490
2240
+ },
2241
+ {
2242
+ "epoch": 2.25,
2243
+ "learning_rate": 5.5700325732899026e-05,
2244
+ "loss": 0.3188,
2245
+ "step": 3500
2246
+ },
2247
+ {
2248
+ "epoch": 2.26,
2249
+ "learning_rate": 5.557003257328991e-05,
2250
+ "loss": 0.3048,
2251
+ "step": 3510
2252
+ },
2253
+ {
2254
+ "epoch": 2.26,
2255
+ "learning_rate": 5.543973941368078e-05,
2256
+ "loss": 0.3211,
2257
+ "step": 3520
2258
+ },
2259
+ {
2260
+ "epoch": 2.27,
2261
+ "learning_rate": 5.530944625407166e-05,
2262
+ "loss": 0.3195,
2263
+ "step": 3530
2264
+ },
2265
+ {
2266
+ "epoch": 2.28,
2267
+ "learning_rate": 5.5179153094462544e-05,
2268
+ "loss": 0.3301,
2269
+ "step": 3540
2270
+ },
2271
+ {
2272
+ "epoch": 2.28,
2273
+ "learning_rate": 5.504885993485343e-05,
2274
+ "loss": 0.3251,
2275
+ "step": 3550
2276
+ },
2277
+ {
2278
+ "epoch": 2.29,
2279
+ "learning_rate": 5.491856677524431e-05,
2280
+ "loss": 0.3214,
2281
+ "step": 3560
2282
+ },
2283
+ {
2284
+ "epoch": 2.3,
2285
+ "learning_rate": 5.478827361563518e-05,
2286
+ "loss": 0.3081,
2287
+ "step": 3570
2288
+ },
2289
+ {
2290
+ "epoch": 2.3,
2291
+ "learning_rate": 5.465798045602606e-05,
2292
+ "loss": 0.3225,
2293
+ "step": 3580
2294
+ },
2295
+ {
2296
+ "epoch": 2.31,
2297
+ "learning_rate": 5.452768729641694e-05,
2298
+ "loss": 0.3263,
2299
+ "step": 3590
2300
+ },
2301
+ {
2302
+ "epoch": 2.32,
2303
+ "learning_rate": 5.439739413680782e-05,
2304
+ "loss": 0.311,
2305
+ "step": 3600
2306
+ },
2307
+ {
2308
+ "epoch": 2.32,
2309
+ "eval_loss": 0.3236062228679657,
2310
+ "eval_runtime": 878.9581,
2311
+ "eval_samples_per_second": 2.275,
2312
+ "eval_steps_per_second": 0.284,
2313
+ "step": 3600
2314
+ },
2315
+ {
2316
+ "epoch": 2.32,
2317
+ "learning_rate": 5.4267100977198703e-05,
2318
+ "loss": 0.3223,
2319
+ "step": 3610
2320
+ },
2321
+ {
2322
+ "epoch": 2.33,
2323
+ "learning_rate": 5.413680781758957e-05,
2324
+ "loss": 0.3244,
2325
+ "step": 3620
2326
+ },
2327
+ {
2328
+ "epoch": 2.33,
2329
+ "learning_rate": 5.4006514657980456e-05,
2330
+ "loss": 0.3251,
2331
+ "step": 3630
2332
+ },
2333
+ {
2334
+ "epoch": 2.34,
2335
+ "learning_rate": 5.387622149837134e-05,
2336
+ "loss": 0.3114,
2337
+ "step": 3640
2338
+ },
2339
+ {
2340
+ "epoch": 2.35,
2341
+ "learning_rate": 5.374592833876222e-05,
2342
+ "loss": 0.324,
2343
+ "step": 3650
2344
+ },
2345
+ {
2346
+ "epoch": 2.35,
2347
+ "learning_rate": 5.3615635179153104e-05,
2348
+ "loss": 0.3188,
2349
+ "step": 3660
2350
+ },
2351
+ {
2352
+ "epoch": 2.36,
2353
+ "learning_rate": 5.348534201954397e-05,
2354
+ "loss": 0.324,
2355
+ "step": 3670
2356
+ },
2357
+ {
2358
+ "epoch": 2.37,
2359
+ "learning_rate": 5.3355048859934856e-05,
2360
+ "loss": 0.324,
2361
+ "step": 3680
2362
+ },
2363
+ {
2364
+ "epoch": 2.37,
2365
+ "learning_rate": 5.322475570032573e-05,
2366
+ "loss": 0.3254,
2367
+ "step": 3690
2368
+ },
2369
+ {
2370
+ "epoch": 2.38,
2371
+ "learning_rate": 5.3094462540716615e-05,
2372
+ "loss": 0.3251,
2373
+ "step": 3700
2374
+ },
2375
+ {
2376
+ "epoch": 2.39,
2377
+ "learning_rate": 5.29641693811075e-05,
2378
+ "loss": 0.3162,
2379
+ "step": 3710
2380
+ },
2381
+ {
2382
+ "epoch": 2.39,
2383
+ "learning_rate": 5.283387622149837e-05,
2384
+ "loss": 0.326,
2385
+ "step": 3720
2386
+ },
2387
+ {
2388
+ "epoch": 2.4,
2389
+ "learning_rate": 5.270358306188925e-05,
2390
+ "loss": 0.3163,
2391
+ "step": 3730
2392
+ },
2393
+ {
2394
+ "epoch": 2.41,
2395
+ "learning_rate": 5.257328990228013e-05,
2396
+ "loss": 0.3262,
2397
+ "step": 3740
2398
+ },
2399
+ {
2400
+ "epoch": 2.41,
2401
+ "learning_rate": 5.2442996742671016e-05,
2402
+ "loss": 0.3227,
2403
+ "step": 3750
2404
+ },
2405
+ {
2406
+ "epoch": 2.42,
2407
+ "learning_rate": 5.23127035830619e-05,
2408
+ "loss": 0.3148,
2409
+ "step": 3760
2410
+ },
2411
+ {
2412
+ "epoch": 2.42,
2413
+ "learning_rate": 5.2182410423452774e-05,
2414
+ "loss": 0.3179,
2415
+ "step": 3770
2416
+ },
2417
+ {
2418
+ "epoch": 2.43,
2419
+ "learning_rate": 5.205211726384365e-05,
2420
+ "loss": 0.3159,
2421
+ "step": 3780
2422
+ },
2423
+ {
2424
+ "epoch": 2.44,
2425
+ "learning_rate": 5.1921824104234526e-05,
2426
+ "loss": 0.3268,
2427
+ "step": 3790
2428
+ },
2429
+ {
2430
+ "epoch": 2.44,
2431
+ "learning_rate": 5.179153094462541e-05,
2432
+ "loss": 0.3297,
2433
+ "step": 3800
2434
+ },
2435
+ {
2436
+ "epoch": 2.44,
2437
+ "eval_loss": 0.3224920332431793,
2438
+ "eval_runtime": 877.9437,
2439
+ "eval_samples_per_second": 2.278,
2440
+ "eval_steps_per_second": 0.285,
2441
+ "step": 3800
2442
+ },
2443
+ {
2444
+ "epoch": 2.45,
2445
+ "learning_rate": 5.166123778501629e-05,
2446
+ "loss": 0.3186,
2447
+ "step": 3810
2448
+ },
2449
+ {
2450
+ "epoch": 2.46,
2451
+ "learning_rate": 5.1530944625407175e-05,
2452
+ "loss": 0.3235,
2453
+ "step": 3820
2454
+ },
2455
+ {
2456
+ "epoch": 2.46,
2457
+ "learning_rate": 5.1400651465798044e-05,
2458
+ "loss": 0.313,
2459
+ "step": 3830
2460
+ },
2461
+ {
2462
+ "epoch": 2.47,
2463
+ "learning_rate": 5.127035830618893e-05,
2464
+ "loss": 0.3137,
2465
+ "step": 3840
2466
+ },
2467
+ {
2468
+ "epoch": 2.48,
2469
+ "learning_rate": 5.114006514657981e-05,
2470
+ "loss": 0.3207,
2471
+ "step": 3850
2472
+ },
2473
+ {
2474
+ "epoch": 2.48,
2475
+ "learning_rate": 5.1009771986970686e-05,
2476
+ "loss": 0.3159,
2477
+ "step": 3860
2478
+ },
2479
+ {
2480
+ "epoch": 2.49,
2481
+ "learning_rate": 5.087947882736157e-05,
2482
+ "loss": 0.3285,
2483
+ "step": 3870
2484
+ },
2485
+ {
2486
+ "epoch": 2.5,
2487
+ "learning_rate": 5.074918566775244e-05,
2488
+ "loss": 0.3258,
2489
+ "step": 3880
2490
+ },
2491
+ {
2492
+ "epoch": 2.5,
2493
+ "learning_rate": 5.061889250814332e-05,
2494
+ "loss": 0.3317,
2495
+ "step": 3890
2496
+ },
2497
+ {
2498
+ "epoch": 2.51,
2499
+ "learning_rate": 5.0488599348534204e-05,
2500
+ "loss": 0.3083,
2501
+ "step": 3900
2502
+ },
2503
+ {
2504
+ "epoch": 2.51,
2505
+ "learning_rate": 5.0358306188925086e-05,
2506
+ "loss": 0.327,
2507
+ "step": 3910
2508
+ },
2509
+ {
2510
+ "epoch": 2.52,
2511
+ "learning_rate": 5.022801302931597e-05,
2512
+ "loss": 0.3147,
2513
+ "step": 3920
2514
+ },
2515
+ {
2516
+ "epoch": 2.53,
2517
+ "learning_rate": 5.009771986970684e-05,
2518
+ "loss": 0.3104,
2519
+ "step": 3930
2520
+ },
2521
+ {
2522
+ "epoch": 2.53,
2523
+ "learning_rate": 4.996742671009772e-05,
2524
+ "loss": 0.3181,
2525
+ "step": 3940
2526
+ },
2527
+ {
2528
+ "epoch": 2.54,
2529
+ "learning_rate": 4.9837133550488604e-05,
2530
+ "loss": 0.3205,
2531
+ "step": 3950
2532
+ },
2533
+ {
2534
+ "epoch": 2.55,
2535
+ "learning_rate": 4.970684039087948e-05,
2536
+ "loss": 0.3063,
2537
+ "step": 3960
2538
+ },
2539
+ {
2540
+ "epoch": 2.55,
2541
+ "learning_rate": 4.9576547231270356e-05,
2542
+ "loss": 0.3197,
2543
+ "step": 3970
2544
+ },
2545
+ {
2546
+ "epoch": 2.56,
2547
+ "learning_rate": 4.944625407166124e-05,
2548
+ "loss": 0.308,
2549
+ "step": 3980
2550
+ },
2551
+ {
2552
+ "epoch": 2.57,
2553
+ "learning_rate": 4.9315960912052115e-05,
2554
+ "loss": 0.322,
2555
+ "step": 3990
2556
+ },
2557
+ {
2558
+ "epoch": 2.57,
2559
+ "learning_rate": 4.9185667752443e-05,
2560
+ "loss": 0.3217,
2561
+ "step": 4000
2562
+ },
2563
+ {
2564
+ "epoch": 2.57,
2565
+ "eval_loss": 0.32212400436401367,
2566
+ "eval_runtime": 868.7384,
2567
+ "eval_samples_per_second": 2.302,
2568
+ "eval_steps_per_second": 0.288,
2569
+ "step": 4000
2570
+ },
2571
+ {
2572
+ "epoch": 2.58,
2573
+ "learning_rate": 4.905537459283388e-05,
2574
+ "loss": 0.3153,
2575
+ "step": 4010
2576
+ },
2577
+ {
2578
+ "epoch": 2.59,
2579
+ "learning_rate": 4.892508143322476e-05,
2580
+ "loss": 0.3278,
2581
+ "step": 4020
2582
+ },
2583
+ {
2584
+ "epoch": 2.59,
2585
+ "learning_rate": 4.879478827361564e-05,
2586
+ "loss": 0.3128,
2587
+ "step": 4030
2588
+ },
2589
+ {
2590
+ "epoch": 2.6,
2591
+ "learning_rate": 4.866449511400652e-05,
2592
+ "loss": 0.3243,
2593
+ "step": 4040
2594
+ },
2595
+ {
2596
+ "epoch": 2.6,
2597
+ "learning_rate": 4.85342019543974e-05,
2598
+ "loss": 0.3175,
2599
+ "step": 4050
2600
+ },
2601
+ {
2602
+ "epoch": 2.61,
2603
+ "learning_rate": 4.8403908794788275e-05,
2604
+ "loss": 0.3104,
2605
+ "step": 4060
2606
+ },
2607
+ {
2608
+ "epoch": 2.62,
2609
+ "learning_rate": 4.827361563517915e-05,
2610
+ "loss": 0.3322,
2611
+ "step": 4070
2612
+ },
2613
+ {
2614
+ "epoch": 2.62,
2615
+ "learning_rate": 4.8143322475570033e-05,
2616
+ "loss": 0.3128,
2617
+ "step": 4080
2618
+ },
2619
+ {
2620
+ "epoch": 2.63,
2621
+ "learning_rate": 4.8013029315960916e-05,
2622
+ "loss": 0.3299,
2623
+ "step": 4090
2624
+ },
2625
+ {
2626
+ "epoch": 2.64,
2627
+ "learning_rate": 4.788273615635179e-05,
2628
+ "loss": 0.3154,
2629
+ "step": 4100
2630
+ },
2631
+ {
2632
+ "epoch": 2.64,
2633
+ "learning_rate": 4.7752442996742675e-05,
2634
+ "loss": 0.327,
2635
+ "step": 4110
2636
+ },
2637
+ {
2638
+ "epoch": 2.65,
2639
+ "learning_rate": 4.762214983713355e-05,
2640
+ "loss": 0.3175,
2641
+ "step": 4120
2642
+ },
2643
+ {
2644
+ "epoch": 2.66,
2645
+ "learning_rate": 4.7491856677524434e-05,
2646
+ "loss": 0.3115,
2647
+ "step": 4130
2648
+ },
2649
+ {
2650
+ "epoch": 2.66,
2651
+ "learning_rate": 4.736156351791532e-05,
2652
+ "loss": 0.3077,
2653
+ "step": 4140
2654
+ },
2655
+ {
2656
+ "epoch": 2.67,
2657
+ "learning_rate": 4.723127035830619e-05,
2658
+ "loss": 0.3171,
2659
+ "step": 4150
2660
+ },
2661
+ {
2662
+ "epoch": 2.68,
2663
+ "learning_rate": 4.710097719869707e-05,
2664
+ "loss": 0.3156,
2665
+ "step": 4160
2666
+ },
2667
+ {
2668
+ "epoch": 2.68,
2669
+ "learning_rate": 4.6970684039087945e-05,
2670
+ "loss": 0.3122,
2671
+ "step": 4170
2672
+ },
2673
+ {
2674
+ "epoch": 2.69,
2675
+ "learning_rate": 4.684039087947883e-05,
2676
+ "loss": 0.332,
2677
+ "step": 4180
2678
+ },
2679
+ {
2680
+ "epoch": 2.69,
2681
+ "learning_rate": 4.671009771986971e-05,
2682
+ "loss": 0.3211,
2683
+ "step": 4190
2684
+ },
2685
+ {
2686
+ "epoch": 2.7,
2687
+ "learning_rate": 4.657980456026059e-05,
2688
+ "loss": 0.3099,
2689
+ "step": 4200
2690
+ },
2691
+ {
2692
+ "epoch": 2.7,
2693
+ "eval_loss": 0.3215048015117645,
2694
+ "eval_runtime": 866.6195,
2695
+ "eval_samples_per_second": 2.308,
2696
+ "eval_steps_per_second": 0.288,
2697
+ "step": 4200
2698
+ },
2699
+ {
2700
+ "epoch": 2.71,
2701
+ "learning_rate": 4.644951140065147e-05,
2702
+ "loss": 0.326,
2703
+ "step": 4210
2704
+ },
2705
+ {
2706
+ "epoch": 2.71,
2707
+ "learning_rate": 4.6319218241042346e-05,
2708
+ "loss": 0.3113,
2709
+ "step": 4220
2710
+ },
2711
+ {
2712
+ "epoch": 2.72,
2713
+ "learning_rate": 4.618892508143323e-05,
2714
+ "loss": 0.3093,
2715
+ "step": 4230
2716
+ },
2717
+ {
2718
+ "epoch": 2.73,
2719
+ "learning_rate": 4.605863192182411e-05,
2720
+ "loss": 0.3161,
2721
+ "step": 4240
2722
+ },
2723
+ {
2724
+ "epoch": 2.73,
2725
+ "learning_rate": 4.592833876221499e-05,
2726
+ "loss": 0.3064,
2727
+ "step": 4250
2728
+ },
2729
+ {
2730
+ "epoch": 2.74,
2731
+ "learning_rate": 4.579804560260586e-05,
2732
+ "loss": 0.3098,
2733
+ "step": 4260
2734
+ },
2735
+ {
2736
+ "epoch": 2.75,
2737
+ "learning_rate": 4.5667752442996746e-05,
2738
+ "loss": 0.3274,
2739
+ "step": 4270
2740
+ },
2741
+ {
2742
+ "epoch": 2.75,
2743
+ "learning_rate": 4.553745928338762e-05,
2744
+ "loss": 0.3118,
2745
+ "step": 4280
2746
+ },
2747
+ {
2748
+ "epoch": 2.76,
2749
+ "learning_rate": 4.5407166123778505e-05,
2750
+ "loss": 0.3212,
2751
+ "step": 4290
2752
+ },
2753
+ {
2754
+ "epoch": 2.77,
2755
+ "learning_rate": 4.527687296416938e-05,
2756
+ "loss": 0.3107,
2757
+ "step": 4300
2758
+ },
2759
+ {
2760
+ "epoch": 2.77,
2761
+ "learning_rate": 4.5146579804560264e-05,
2762
+ "loss": 0.3066,
2763
+ "step": 4310
2764
+ },
2765
+ {
2766
+ "epoch": 2.78,
2767
+ "learning_rate": 4.501628664495115e-05,
2768
+ "loss": 0.3161,
2769
+ "step": 4320
2770
+ },
2771
+ {
2772
+ "epoch": 2.78,
2773
+ "learning_rate": 4.488599348534202e-05,
2774
+ "loss": 0.3127,
2775
+ "step": 4330
2776
+ },
2777
+ {
2778
+ "epoch": 2.79,
2779
+ "learning_rate": 4.47557003257329e-05,
2780
+ "loss": 0.3278,
2781
+ "step": 4340
2782
+ },
2783
+ {
2784
+ "epoch": 2.8,
2785
+ "learning_rate": 4.462540716612378e-05,
2786
+ "loss": 0.3199,
2787
+ "step": 4350
2788
+ },
2789
+ {
2790
+ "epoch": 2.8,
2791
+ "learning_rate": 4.449511400651466e-05,
2792
+ "loss": 0.3082,
2793
+ "step": 4360
2794
+ },
2795
+ {
2796
+ "epoch": 2.81,
2797
+ "learning_rate": 4.436482084690554e-05,
2798
+ "loss": 0.3217,
2799
+ "step": 4370
2800
+ },
2801
+ {
2802
+ "epoch": 2.82,
2803
+ "learning_rate": 4.4234527687296417e-05,
2804
+ "loss": 0.3017,
2805
+ "step": 4380
2806
+ },
2807
+ {
2808
+ "epoch": 2.82,
2809
+ "learning_rate": 4.41042345276873e-05,
2810
+ "loss": 0.3111,
2811
+ "step": 4390
2812
+ },
2813
+ {
2814
+ "epoch": 2.83,
2815
+ "learning_rate": 4.3973941368078175e-05,
2816
+ "loss": 0.3141,
2817
+ "step": 4400
2818
+ },
2819
+ {
2820
+ "epoch": 2.83,
2821
+ "eval_loss": 0.3212812542915344,
2822
+ "eval_runtime": 865.9391,
2823
+ "eval_samples_per_second": 2.31,
2824
+ "eval_steps_per_second": 0.289,
2825
+ "step": 4400
2826
+ },
2827
+ {
2828
+ "epoch": 2.84,
2829
+ "learning_rate": 4.384364820846906e-05,
2830
+ "loss": 0.3121,
2831
+ "step": 4410
2832
+ },
2833
+ {
2834
+ "epoch": 2.84,
2835
+ "learning_rate": 4.371335504885994e-05,
2836
+ "loss": 0.3168,
2837
+ "step": 4420
2838
+ },
2839
+ {
2840
+ "epoch": 2.85,
2841
+ "learning_rate": 4.358306188925082e-05,
2842
+ "loss": 0.3093,
2843
+ "step": 4430
2844
+ },
2845
+ {
2846
+ "epoch": 2.86,
2847
+ "learning_rate": 4.345276872964169e-05,
2848
+ "loss": 0.3092,
2849
+ "step": 4440
2850
+ },
2851
+ {
2852
+ "epoch": 2.86,
2853
+ "learning_rate": 4.3322475570032576e-05,
2854
+ "loss": 0.3268,
2855
+ "step": 4450
2856
+ },
2857
+ {
2858
+ "epoch": 2.87,
2859
+ "learning_rate": 4.319218241042345e-05,
2860
+ "loss": 0.3214,
2861
+ "step": 4460
2862
+ },
2863
+ {
2864
+ "epoch": 2.87,
2865
+ "learning_rate": 4.3061889250814335e-05,
2866
+ "loss": 0.3047,
2867
+ "step": 4470
2868
+ },
2869
+ {
2870
+ "epoch": 2.88,
2871
+ "learning_rate": 4.293159609120521e-05,
2872
+ "loss": 0.3207,
2873
+ "step": 4480
2874
+ },
2875
+ {
2876
+ "epoch": 2.89,
2877
+ "learning_rate": 4.2801302931596094e-05,
2878
+ "loss": 0.3209,
2879
+ "step": 4490
2880
+ },
2881
+ {
2882
+ "epoch": 2.89,
2883
+ "learning_rate": 4.2671009771986977e-05,
2884
+ "loss": 0.3147,
2885
+ "step": 4500
2886
+ },
2887
+ {
2888
+ "epoch": 2.9,
2889
+ "learning_rate": 4.254071661237785e-05,
2890
+ "loss": 0.3109,
2891
+ "step": 4510
2892
+ },
2893
+ {
2894
+ "epoch": 2.91,
2895
+ "learning_rate": 4.2410423452768735e-05,
2896
+ "loss": 0.3254,
2897
+ "step": 4520
2898
+ },
2899
+ {
2900
+ "epoch": 2.91,
2901
+ "learning_rate": 4.228013029315961e-05,
2902
+ "loss": 0.3166,
2903
+ "step": 4530
2904
+ },
2905
+ {
2906
+ "epoch": 2.92,
2907
+ "learning_rate": 4.214983713355049e-05,
2908
+ "loss": 0.3211,
2909
+ "step": 4540
2910
+ },
2911
+ {
2912
+ "epoch": 2.93,
2913
+ "learning_rate": 4.201954397394137e-05,
2914
+ "loss": 0.3216,
2915
+ "step": 4550
2916
+ },
2917
+ {
2918
+ "epoch": 2.93,
2919
+ "learning_rate": 4.1889250814332246e-05,
2920
+ "loss": 0.3196,
2921
+ "step": 4560
2922
+ },
2923
+ {
2924
+ "epoch": 2.94,
2925
+ "learning_rate": 4.175895765472313e-05,
2926
+ "loss": 0.3123,
2927
+ "step": 4570
2928
+ },
2929
+ {
2930
+ "epoch": 2.95,
2931
+ "learning_rate": 4.1628664495114005e-05,
2932
+ "loss": 0.3163,
2933
+ "step": 4580
2934
+ },
2935
+ {
2936
+ "epoch": 2.95,
2937
+ "learning_rate": 4.149837133550489e-05,
2938
+ "loss": 0.3071,
2939
+ "step": 4590
2940
+ },
2941
+ {
2942
+ "epoch": 2.96,
2943
+ "learning_rate": 4.136807817589577e-05,
2944
+ "loss": 0.3179,
2945
+ "step": 4600
2946
+ },
2947
+ {
2948
+ "epoch": 2.96,
2949
+ "eval_loss": 0.32042255997657776,
2950
+ "eval_runtime": 868.0765,
2951
+ "eval_samples_per_second": 2.304,
2952
+ "eval_steps_per_second": 0.288,
2953
+ "step": 4600
2954
+ },
2955
+ {
2956
+ "epoch": 2.96,
2957
+ "learning_rate": 4.123778501628665e-05,
2958
+ "loss": 0.3194,
2959
+ "step": 4610
2960
+ },
2961
+ {
2962
+ "epoch": 2.97,
2963
+ "learning_rate": 4.110749185667753e-05,
2964
+ "loss": 0.3281,
2965
+ "step": 4620
2966
+ },
2967
+ {
2968
+ "epoch": 2.98,
2969
+ "learning_rate": 4.0977198697068406e-05,
2970
+ "loss": 0.3146,
2971
+ "step": 4630
2972
+ },
2973
+ {
2974
+ "epoch": 2.98,
2975
+ "learning_rate": 4.084690553745928e-05,
2976
+ "loss": 0.31,
2977
+ "step": 4640
2978
+ },
2979
+ {
2980
+ "epoch": 2.99,
2981
+ "learning_rate": 4.0716612377850165e-05,
2982
+ "loss": 0.3182,
2983
+ "step": 4650
2984
+ },
2985
+ {
2986
+ "epoch": 3.0,
2987
+ "learning_rate": 4.058631921824104e-05,
2988
+ "loss": 0.3191,
2989
+ "step": 4660
2990
+ },
2991
+ {
2992
+ "epoch": 3.0,
2993
+ "learning_rate": 4.0456026058631924e-05,
2994
+ "loss": 0.3139,
2995
+ "step": 4670
2996
+ },
2997
+ {
2998
+ "epoch": 3.01,
2999
+ "learning_rate": 4.0325732899022806e-05,
3000
+ "loss": 0.3151,
3001
+ "step": 4680
3002
+ },
3003
+ {
3004
+ "epoch": 3.02,
3005
+ "learning_rate": 4.019543973941368e-05,
3006
+ "loss": 0.3196,
3007
+ "step": 4690
3008
+ },
3009
+ {
3010
+ "epoch": 3.02,
3011
+ "learning_rate": 4.0065146579804565e-05,
3012
+ "loss": 0.3036,
3013
+ "step": 4700
3014
+ },
3015
+ {
3016
+ "epoch": 3.03,
3017
+ "learning_rate": 3.993485342019544e-05,
3018
+ "loss": 0.3069,
3019
+ "step": 4710
3020
+ },
3021
+ {
3022
+ "epoch": 3.04,
3023
+ "learning_rate": 3.9804560260586324e-05,
3024
+ "loss": 0.3144,
3025
+ "step": 4720
3026
+ },
3027
+ {
3028
+ "epoch": 3.04,
3029
+ "learning_rate": 3.96742671009772e-05,
3030
+ "loss": 0.308,
3031
+ "step": 4730
3032
+ },
3033
+ {
3034
+ "epoch": 3.05,
3035
+ "learning_rate": 3.9543973941368076e-05,
3036
+ "loss": 0.3006,
3037
+ "step": 4740
3038
+ },
3039
+ {
3040
+ "epoch": 3.05,
3041
+ "learning_rate": 3.941368078175896e-05,
3042
+ "loss": 0.3124,
3043
+ "step": 4750
3044
+ },
3045
+ {
3046
+ "epoch": 3.06,
3047
+ "learning_rate": 3.9283387622149835e-05,
3048
+ "loss": 0.3027,
3049
+ "step": 4760
3050
+ },
3051
+ {
3052
+ "epoch": 3.07,
3053
+ "learning_rate": 3.915309446254072e-05,
3054
+ "loss": 0.3098,
3055
+ "step": 4770
3056
+ },
3057
+ {
3058
+ "epoch": 3.07,
3059
+ "learning_rate": 3.90228013029316e-05,
3060
+ "loss": 0.3128,
3061
+ "step": 4780
3062
+ },
3063
+ {
3064
+ "epoch": 3.08,
3065
+ "learning_rate": 3.889250814332248e-05,
3066
+ "loss": 0.3068,
3067
+ "step": 4790
3068
+ },
3069
+ {
3070
+ "epoch": 3.09,
3071
+ "learning_rate": 3.876221498371336e-05,
3072
+ "loss": 0.3049,
3073
+ "step": 4800
3074
+ },
3075
+ {
3076
+ "epoch": 3.09,
3077
+ "eval_loss": 0.3204311430454254,
3078
+ "eval_runtime": 865.119,
3079
+ "eval_samples_per_second": 2.312,
3080
+ "eval_steps_per_second": 0.289,
3081
+ "step": 4800
3082
+ },
3083
+ {
3084
+ "epoch": 3.09,
3085
+ "learning_rate": 3.863192182410424e-05,
3086
+ "loss": 0.3082,
3087
+ "step": 4810
3088
+ },
3089
+ {
3090
+ "epoch": 3.1,
3091
+ "learning_rate": 3.850162866449512e-05,
3092
+ "loss": 0.3036,
3093
+ "step": 4820
3094
+ },
3095
+ {
3096
+ "epoch": 3.11,
3097
+ "learning_rate": 3.8371335504885995e-05,
3098
+ "loss": 0.3106,
3099
+ "step": 4830
3100
+ },
3101
+ {
3102
+ "epoch": 3.11,
3103
+ "learning_rate": 3.824104234527687e-05,
3104
+ "loss": 0.3052,
3105
+ "step": 4840
3106
+ },
3107
+ {
3108
+ "epoch": 3.12,
3109
+ "learning_rate": 3.811074918566775e-05,
3110
+ "loss": 0.3129,
3111
+ "step": 4850
3112
+ },
3113
+ {
3114
+ "epoch": 3.13,
3115
+ "learning_rate": 3.7980456026058636e-05,
3116
+ "loss": 0.3119,
3117
+ "step": 4860
3118
+ },
3119
+ {
3120
+ "epoch": 3.13,
3121
+ "learning_rate": 3.785016286644951e-05,
3122
+ "loss": 0.3203,
3123
+ "step": 4870
3124
+ },
3125
+ {
3126
+ "epoch": 3.14,
3127
+ "learning_rate": 3.7719869706840395e-05,
3128
+ "loss": 0.306,
3129
+ "step": 4880
3130
+ },
3131
+ {
3132
+ "epoch": 3.14,
3133
+ "learning_rate": 3.758957654723127e-05,
3134
+ "loss": 0.3122,
3135
+ "step": 4890
3136
+ },
3137
+ {
3138
+ "epoch": 3.15,
3139
+ "learning_rate": 3.7459283387622154e-05,
3140
+ "loss": 0.3069,
3141
+ "step": 4900
3142
+ },
3143
+ {
3144
+ "epoch": 3.16,
3145
+ "learning_rate": 3.732899022801303e-05,
3146
+ "loss": 0.3208,
3147
+ "step": 4910
3148
+ },
3149
+ {
3150
+ "epoch": 3.16,
3151
+ "learning_rate": 3.7198697068403906e-05,
3152
+ "loss": 0.3282,
3153
+ "step": 4920
3154
+ },
3155
+ {
3156
+ "epoch": 3.17,
3157
+ "learning_rate": 3.706840390879479e-05,
3158
+ "loss": 0.3146,
3159
+ "step": 4930
3160
+ },
3161
+ {
3162
+ "epoch": 3.18,
3163
+ "learning_rate": 3.6938110749185665e-05,
3164
+ "loss": 0.3069,
3165
+ "step": 4940
3166
+ },
3167
+ {
3168
+ "epoch": 3.18,
3169
+ "learning_rate": 3.680781758957655e-05,
3170
+ "loss": 0.3122,
3171
+ "step": 4950
3172
+ },
3173
+ {
3174
+ "epoch": 3.19,
3175
+ "learning_rate": 3.667752442996743e-05,
3176
+ "loss": 0.3071,
3177
+ "step": 4960
3178
+ },
3179
+ {
3180
+ "epoch": 3.2,
3181
+ "learning_rate": 3.6547231270358307e-05,
3182
+ "loss": 0.3163,
3183
+ "step": 4970
3184
+ },
3185
+ {
3186
+ "epoch": 3.2,
3187
+ "learning_rate": 3.641693811074919e-05,
3188
+ "loss": 0.3017,
3189
+ "step": 4980
3190
+ },
3191
+ {
3192
+ "epoch": 3.21,
3193
+ "learning_rate": 3.6286644951140065e-05,
3194
+ "loss": 0.312,
3195
+ "step": 4990
3196
+ },
3197
+ {
3198
+ "epoch": 3.22,
3199
+ "learning_rate": 3.615635179153095e-05,
3200
+ "loss": 0.293,
3201
+ "step": 5000
3202
+ },
3203
+ {
3204
+ "epoch": 3.22,
3205
+ "eval_loss": 0.32003089785575867,
3206
+ "eval_runtime": 866.3295,
3207
+ "eval_samples_per_second": 2.309,
3208
+ "eval_steps_per_second": 0.289,
3209
+ "step": 5000
3210
+ },
3211
+ {
3212
+ "epoch": 3.22,
3213
+ "learning_rate": 3.6026058631921824e-05,
3214
+ "loss": 0.3054,
3215
+ "step": 5010
3216
+ },
3217
+ {
3218
+ "epoch": 3.23,
3219
+ "learning_rate": 3.58957654723127e-05,
3220
+ "loss": 0.3113,
3221
+ "step": 5020
3222
+ },
3223
+ {
3224
+ "epoch": 3.23,
3225
+ "learning_rate": 3.576547231270358e-05,
3226
+ "loss": 0.3027,
3227
+ "step": 5030
3228
+ },
3229
+ {
3230
+ "epoch": 3.24,
3231
+ "learning_rate": 3.5635179153094466e-05,
3232
+ "loss": 0.3027,
3233
+ "step": 5040
3234
+ },
3235
+ {
3236
+ "epoch": 3.25,
3237
+ "learning_rate": 3.550488599348534e-05,
3238
+ "loss": 0.3083,
3239
+ "step": 5050
3240
+ },
3241
+ {
3242
+ "epoch": 3.25,
3243
+ "learning_rate": 3.5374592833876225e-05,
3244
+ "loss": 0.3158,
3245
+ "step": 5060
3246
+ },
3247
+ {
3248
+ "epoch": 3.26,
3249
+ "learning_rate": 3.52442996742671e-05,
3250
+ "loss": 0.3258,
3251
+ "step": 5070
3252
+ },
3253
+ {
3254
+ "epoch": 3.27,
3255
+ "learning_rate": 3.5114006514657984e-05,
3256
+ "loss": 0.3099,
3257
+ "step": 5080
3258
+ },
3259
+ {
3260
+ "epoch": 3.27,
3261
+ "learning_rate": 3.4983713355048867e-05,
3262
+ "loss": 0.301,
3263
+ "step": 5090
3264
+ },
3265
+ {
3266
+ "epoch": 3.28,
3267
+ "learning_rate": 3.485342019543974e-05,
3268
+ "loss": 0.3113,
3269
+ "step": 5100
3270
+ },
3271
+ {
3272
+ "epoch": 3.29,
3273
+ "learning_rate": 3.472312703583062e-05,
3274
+ "loss": 0.3,
3275
+ "step": 5110
3276
+ },
3277
+ {
3278
+ "epoch": 3.29,
3279
+ "learning_rate": 3.4592833876221495e-05,
3280
+ "loss": 0.3077,
3281
+ "step": 5120
3282
+ },
3283
+ {
3284
+ "epoch": 3.3,
3285
+ "learning_rate": 3.446254071661238e-05,
3286
+ "loss": 0.3224,
3287
+ "step": 5130
3288
+ },
3289
+ {
3290
+ "epoch": 3.31,
3291
+ "learning_rate": 3.433224755700326e-05,
3292
+ "loss": 0.303,
3293
+ "step": 5140
3294
+ },
3295
+ {
3296
+ "epoch": 3.31,
3297
+ "learning_rate": 3.4201954397394136e-05,
3298
+ "loss": 0.3069,
3299
+ "step": 5150
3300
+ },
3301
+ {
3302
+ "epoch": 3.32,
3303
+ "learning_rate": 3.407166123778502e-05,
3304
+ "loss": 0.316,
3305
+ "step": 5160
3306
+ },
3307
+ {
3308
+ "epoch": 3.32,
3309
+ "learning_rate": 3.3941368078175895e-05,
3310
+ "loss": 0.31,
3311
+ "step": 5170
3312
+ },
3313
+ {
3314
+ "epoch": 3.33,
3315
+ "learning_rate": 3.381107491856678e-05,
3316
+ "loss": 0.3103,
3317
+ "step": 5180
3318
+ },
3319
+ {
3320
+ "epoch": 3.34,
3321
+ "learning_rate": 3.368078175895766e-05,
3322
+ "loss": 0.3114,
3323
+ "step": 5190
3324
+ },
3325
+ {
3326
+ "epoch": 3.34,
3327
+ "learning_rate": 3.355048859934854e-05,
3328
+ "loss": 0.3038,
3329
+ "step": 5200
3330
+ },
3331
+ {
3332
+ "epoch": 3.34,
3333
+ "eval_loss": 0.3200548589229584,
3334
+ "eval_runtime": 866.3756,
3335
+ "eval_samples_per_second": 2.308,
3336
+ "eval_steps_per_second": 0.289,
3337
+ "step": 5200
3338
+ },
3339
+ {
3340
+ "epoch": 3.35,
3341
+ "learning_rate": 3.342019543973941e-05,
3342
+ "loss": 0.3146,
3343
+ "step": 5210
3344
+ },
3345
+ {
3346
+ "epoch": 3.36,
3347
+ "learning_rate": 3.3289902280130296e-05,
3348
+ "loss": 0.3075,
3349
+ "step": 5220
3350
+ },
3351
+ {
3352
+ "epoch": 3.36,
3353
+ "learning_rate": 3.315960912052117e-05,
3354
+ "loss": 0.3164,
3355
+ "step": 5230
3356
+ },
3357
+ {
3358
+ "epoch": 3.37,
3359
+ "learning_rate": 3.3029315960912055e-05,
3360
+ "loss": 0.3201,
3361
+ "step": 5240
3362
+ },
3363
+ {
3364
+ "epoch": 3.38,
3365
+ "learning_rate": 3.289902280130293e-05,
3366
+ "loss": 0.299,
3367
+ "step": 5250
3368
+ },
3369
+ {
3370
+ "epoch": 3.38,
3371
+ "learning_rate": 3.2768729641693814e-05,
3372
+ "loss": 0.3201,
3373
+ "step": 5260
3374
+ },
3375
+ {
3376
+ "epoch": 3.39,
3377
+ "learning_rate": 3.2638436482084696e-05,
3378
+ "loss": 0.3124,
3379
+ "step": 5270
3380
+ },
3381
+ {
3382
+ "epoch": 3.4,
3383
+ "learning_rate": 3.250814332247557e-05,
3384
+ "loss": 0.3081,
3385
+ "step": 5280
3386
+ },
3387
+ {
3388
+ "epoch": 3.4,
3389
+ "learning_rate": 3.2377850162866455e-05,
3390
+ "loss": 0.31,
3391
+ "step": 5290
3392
+ },
3393
+ {
3394
+ "epoch": 3.41,
3395
+ "learning_rate": 3.224755700325733e-05,
3396
+ "loss": 0.3143,
3397
+ "step": 5300
3398
+ },
3399
+ {
3400
+ "epoch": 3.41,
3401
+ "learning_rate": 3.211726384364821e-05,
3402
+ "loss": 0.3081,
3403
+ "step": 5310
3404
+ },
3405
+ {
3406
+ "epoch": 3.42,
3407
+ "learning_rate": 3.198697068403909e-05,
3408
+ "loss": 0.3086,
3409
+ "step": 5320
3410
+ },
3411
+ {
3412
+ "epoch": 3.43,
3413
+ "learning_rate": 3.1856677524429966e-05,
3414
+ "loss": 0.3106,
3415
+ "step": 5330
3416
+ },
3417
+ {
3418
+ "epoch": 3.43,
3419
+ "learning_rate": 3.172638436482085e-05,
3420
+ "loss": 0.3062,
3421
+ "step": 5340
3422
+ },
3423
+ {
3424
+ "epoch": 3.44,
3425
+ "learning_rate": 3.1596091205211725e-05,
3426
+ "loss": 0.3022,
3427
+ "step": 5350
3428
+ },
3429
+ {
3430
+ "epoch": 3.45,
3431
+ "learning_rate": 3.146579804560261e-05,
3432
+ "loss": 0.3089,
3433
+ "step": 5360
3434
+ },
3435
+ {
3436
+ "epoch": 3.45,
3437
+ "learning_rate": 3.133550488599349e-05,
3438
+ "loss": 0.3092,
3439
+ "step": 5370
3440
+ },
3441
+ {
3442
+ "epoch": 3.46,
3443
+ "learning_rate": 3.120521172638437e-05,
3444
+ "loss": 0.3097,
3445
+ "step": 5380
3446
+ },
3447
+ {
3448
+ "epoch": 3.47,
3449
+ "learning_rate": 3.107491856677524e-05,
3450
+ "loss": 0.3263,
3451
+ "step": 5390
3452
+ },
3453
+ {
3454
+ "epoch": 3.47,
3455
+ "learning_rate": 3.0944625407166126e-05,
3456
+ "loss": 0.313,
3457
+ "step": 5400
3458
+ },
3459
+ {
3460
+ "epoch": 3.47,
3461
+ "eval_loss": 0.3192284405231476,
3462
+ "eval_runtime": 867.7162,
3463
+ "eval_samples_per_second": 2.305,
3464
+ "eval_steps_per_second": 0.288,
3465
+ "step": 5400
3466
+ },
3467
+ {
3468
+ "epoch": 3.48,
3469
+ "learning_rate": 3.0814332247557e-05,
3470
+ "loss": 0.315,
3471
+ "step": 5410
3472
+ },
3473
+ {
3474
+ "epoch": 3.49,
3475
+ "learning_rate": 3.0684039087947885e-05,
3476
+ "loss": 0.3183,
3477
+ "step": 5420
3478
+ },
3479
+ {
3480
+ "epoch": 3.49,
3481
+ "learning_rate": 3.055374592833876e-05,
3482
+ "loss": 0.313,
3483
+ "step": 5430
3484
+ },
3485
+ {
3486
+ "epoch": 3.5,
3487
+ "learning_rate": 3.0423452768729643e-05,
3488
+ "loss": 0.3118,
3489
+ "step": 5440
3490
+ },
3491
+ {
3492
+ "epoch": 3.5,
3493
+ "learning_rate": 3.0293159609120526e-05,
3494
+ "loss": 0.3233,
3495
+ "step": 5450
3496
+ },
3497
+ {
3498
+ "epoch": 3.51,
3499
+ "learning_rate": 3.0162866449511402e-05,
3500
+ "loss": 0.3269,
3501
+ "step": 5460
3502
+ },
3503
+ {
3504
+ "epoch": 3.52,
3505
+ "learning_rate": 3.0032573289902282e-05,
3506
+ "loss": 0.3051,
3507
+ "step": 5470
3508
+ },
3509
+ {
3510
+ "epoch": 3.52,
3511
+ "learning_rate": 2.9902280130293158e-05,
3512
+ "loss": 0.3052,
3513
+ "step": 5480
3514
+ },
3515
+ {
3516
+ "epoch": 3.53,
3517
+ "learning_rate": 2.977198697068404e-05,
3518
+ "loss": 0.3213,
3519
+ "step": 5490
3520
+ },
3521
+ {
3522
+ "epoch": 3.54,
3523
+ "learning_rate": 2.9641693811074923e-05,
3524
+ "loss": 0.3096,
3525
+ "step": 5500
3526
+ },
3527
+ {
3528
+ "epoch": 3.54,
3529
+ "learning_rate": 2.95114006514658e-05,
3530
+ "loss": 0.311,
3531
+ "step": 5510
3532
+ },
3533
+ {
3534
+ "epoch": 3.55,
3535
+ "learning_rate": 2.938110749185668e-05,
3536
+ "loss": 0.3048,
3537
+ "step": 5520
3538
+ },
3539
+ {
3540
+ "epoch": 3.56,
3541
+ "learning_rate": 2.9250814332247555e-05,
3542
+ "loss": 0.308,
3543
+ "step": 5530
3544
+ },
3545
+ {
3546
+ "epoch": 3.56,
3547
+ "learning_rate": 2.9120521172638438e-05,
3548
+ "loss": 0.3074,
3549
+ "step": 5540
3550
+ },
3551
+ {
3552
+ "epoch": 3.57,
3553
+ "learning_rate": 2.899022801302932e-05,
3554
+ "loss": 0.3208,
3555
+ "step": 5550
3556
+ },
3557
+ {
3558
+ "epoch": 3.58,
3559
+ "learning_rate": 2.8859934853420197e-05,
3560
+ "loss": 0.318,
3561
+ "step": 5560
3562
+ },
3563
+ {
3564
+ "epoch": 3.58,
3565
+ "learning_rate": 2.8729641693811076e-05,
3566
+ "loss": 0.3175,
3567
+ "step": 5570
3568
+ },
3569
+ {
3570
+ "epoch": 3.59,
3571
+ "learning_rate": 2.859934853420196e-05,
3572
+ "loss": 0.315,
3573
+ "step": 5580
3574
+ },
3575
+ {
3576
+ "epoch": 3.59,
3577
+ "learning_rate": 2.8469055374592835e-05,
3578
+ "loss": 0.3075,
3579
+ "step": 5590
3580
+ },
3581
+ {
3582
+ "epoch": 3.6,
3583
+ "learning_rate": 2.8338762214983718e-05,
3584
+ "loss": 0.3157,
3585
+ "step": 5600
3586
+ },
3587
+ {
3588
+ "epoch": 3.6,
3589
+ "eval_loss": 0.31872594356536865,
3590
+ "eval_runtime": 866.8553,
3591
+ "eval_samples_per_second": 2.307,
3592
+ "eval_steps_per_second": 0.288,
3593
+ "step": 5600
3594
+ },
3595
+ {
3596
+ "epoch": 3.61,
3597
+ "learning_rate": 2.8208469055374594e-05,
3598
+ "loss": 0.3135,
3599
+ "step": 5610
3600
+ },
3601
+ {
3602
+ "epoch": 3.61,
3603
+ "learning_rate": 2.8078175895765473e-05,
3604
+ "loss": 0.3095,
3605
+ "step": 5620
3606
+ },
3607
+ {
3608
+ "epoch": 3.62,
3609
+ "learning_rate": 2.7947882736156356e-05,
3610
+ "loss": 0.2958,
3611
+ "step": 5630
3612
+ },
3613
+ {
3614
+ "epoch": 3.63,
3615
+ "learning_rate": 2.7817589576547232e-05,
3616
+ "loss": 0.3071,
3617
+ "step": 5640
3618
+ },
3619
+ {
3620
+ "epoch": 3.63,
3621
+ "learning_rate": 2.768729641693811e-05,
3622
+ "loss": 0.3257,
3623
+ "step": 5650
3624
+ },
3625
+ {
3626
+ "epoch": 3.64,
3627
+ "learning_rate": 2.755700325732899e-05,
3628
+ "loss": 0.3092,
3629
+ "step": 5660
3630
+ },
3631
+ {
3632
+ "epoch": 3.65,
3633
+ "learning_rate": 2.742671009771987e-05,
3634
+ "loss": 0.3025,
3635
+ "step": 5670
3636
+ },
3637
+ {
3638
+ "epoch": 3.65,
3639
+ "learning_rate": 2.7296416938110753e-05,
3640
+ "loss": 0.3179,
3641
+ "step": 5680
3642
+ },
3643
+ {
3644
+ "epoch": 3.66,
3645
+ "learning_rate": 2.716612377850163e-05,
3646
+ "loss": 0.3121,
3647
+ "step": 5690
3648
+ },
3649
+ {
3650
+ "epoch": 3.67,
3651
+ "learning_rate": 2.703583061889251e-05,
3652
+ "loss": 0.3222,
3653
+ "step": 5700
3654
+ },
3655
+ {
3656
+ "epoch": 3.67,
3657
+ "learning_rate": 2.6905537459283385e-05,
3658
+ "loss": 0.3016,
3659
+ "step": 5710
3660
+ },
3661
+ {
3662
+ "epoch": 3.68,
3663
+ "learning_rate": 2.6775244299674268e-05,
3664
+ "loss": 0.3092,
3665
+ "step": 5720
3666
+ },
3667
+ {
3668
+ "epoch": 3.68,
3669
+ "learning_rate": 2.664495114006515e-05,
3670
+ "loss": 0.3129,
3671
+ "step": 5730
3672
+ },
3673
+ {
3674
+ "epoch": 3.69,
3675
+ "learning_rate": 2.6514657980456026e-05,
3676
+ "loss": 0.3098,
3677
+ "step": 5740
3678
+ },
3679
+ {
3680
+ "epoch": 3.7,
3681
+ "learning_rate": 2.6384364820846906e-05,
3682
+ "loss": 0.2959,
3683
+ "step": 5750
3684
+ },
3685
+ {
3686
+ "epoch": 3.7,
3687
+ "learning_rate": 2.6254071661237782e-05,
3688
+ "loss": 0.3136,
3689
+ "step": 5760
3690
+ },
3691
+ {
3692
+ "epoch": 3.71,
3693
+ "learning_rate": 2.6123778501628665e-05,
3694
+ "loss": 0.2931,
3695
+ "step": 5770
3696
+ },
3697
+ {
3698
+ "epoch": 3.72,
3699
+ "learning_rate": 2.5993485342019548e-05,
3700
+ "loss": 0.3104,
3701
+ "step": 5780
3702
+ },
3703
+ {
3704
+ "epoch": 3.72,
3705
+ "learning_rate": 2.5863192182410424e-05,
3706
+ "loss": 0.3072,
3707
+ "step": 5790
3708
+ },
3709
+ {
3710
+ "epoch": 3.73,
3711
+ "learning_rate": 2.5732899022801303e-05,
3712
+ "loss": 0.3041,
3713
+ "step": 5800
3714
+ },
3715
+ {
3716
+ "epoch": 3.73,
3717
+ "eval_loss": 0.31860673427581787,
3718
+ "eval_runtime": 862.7977,
3719
+ "eval_samples_per_second": 2.318,
3720
+ "eval_steps_per_second": 0.29,
3721
+ "step": 5800
3722
+ },
3723
+ {
3724
+ "epoch": 3.74,
3725
+ "learning_rate": 2.5602605863192186e-05,
3726
+ "loss": 0.3136,
3727
+ "step": 5810
3728
+ },
3729
+ {
3730
+ "epoch": 3.74,
3731
+ "learning_rate": 2.5472312703583062e-05,
3732
+ "loss": 0.2986,
3733
+ "step": 5820
3734
+ },
3735
+ {
3736
+ "epoch": 3.75,
3737
+ "learning_rate": 2.5342019543973945e-05,
3738
+ "loss": 0.3081,
3739
+ "step": 5830
3740
+ },
3741
+ {
3742
+ "epoch": 3.76,
3743
+ "learning_rate": 2.521172638436482e-05,
3744
+ "loss": 0.3137,
3745
+ "step": 5840
3746
+ },
3747
+ {
3748
+ "epoch": 3.76,
3749
+ "learning_rate": 2.50814332247557e-05,
3750
+ "loss": 0.3165,
3751
+ "step": 5850
3752
+ },
3753
+ {
3754
+ "epoch": 3.77,
3755
+ "learning_rate": 2.495114006514658e-05,
3756
+ "loss": 0.3132,
3757
+ "step": 5860
3758
+ },
3759
+ {
3760
+ "epoch": 3.77,
3761
+ "learning_rate": 2.482084690553746e-05,
3762
+ "loss": 0.3087,
3763
+ "step": 5870
3764
+ },
3765
+ {
3766
+ "epoch": 3.78,
3767
+ "learning_rate": 2.4690553745928342e-05,
3768
+ "loss": 0.3202,
3769
+ "step": 5880
3770
+ },
3771
+ {
3772
+ "epoch": 3.79,
3773
+ "learning_rate": 2.456026058631922e-05,
3774
+ "loss": 0.3084,
3775
+ "step": 5890
3776
+ },
3777
+ {
3778
+ "epoch": 3.79,
3779
+ "learning_rate": 2.4429967426710097e-05,
3780
+ "loss": 0.3146,
3781
+ "step": 5900
3782
+ },
3783
+ {
3784
+ "epoch": 3.8,
3785
+ "learning_rate": 2.4299674267100977e-05,
3786
+ "loss": 0.3038,
3787
+ "step": 5910
3788
+ },
3789
+ {
3790
+ "epoch": 3.81,
3791
+ "learning_rate": 2.416938110749186e-05,
3792
+ "loss": 0.3088,
3793
+ "step": 5920
3794
+ },
3795
+ {
3796
+ "epoch": 3.81,
3797
+ "learning_rate": 2.403908794788274e-05,
3798
+ "loss": 0.3081,
3799
+ "step": 5930
3800
+ },
3801
+ {
3802
+ "epoch": 3.82,
3803
+ "learning_rate": 2.3908794788273615e-05,
3804
+ "loss": 0.3143,
3805
+ "step": 5940
3806
+ },
3807
+ {
3808
+ "epoch": 3.83,
3809
+ "learning_rate": 2.3778501628664495e-05,
3810
+ "loss": 0.3194,
3811
+ "step": 5950
3812
+ },
3813
+ {
3814
+ "epoch": 3.83,
3815
+ "learning_rate": 2.3648208469055374e-05,
3816
+ "loss": 0.3074,
3817
+ "step": 5960
3818
+ },
3819
+ {
3820
+ "epoch": 3.84,
3821
+ "learning_rate": 2.3517915309446257e-05,
3822
+ "loss": 0.3027,
3823
+ "step": 5970
3824
+ },
3825
+ {
3826
+ "epoch": 3.85,
3827
+ "learning_rate": 2.3387622149837136e-05,
3828
+ "loss": 0.3303,
3829
+ "step": 5980
3830
+ },
3831
+ {
3832
+ "epoch": 3.85,
3833
+ "learning_rate": 2.3257328990228012e-05,
3834
+ "loss": 0.3095,
3835
+ "step": 5990
3836
+ },
3837
+ {
3838
+ "epoch": 3.86,
3839
+ "learning_rate": 2.3127035830618892e-05,
3840
+ "loss": 0.3124,
3841
+ "step": 6000
3842
+ },
3843
+ {
3844
+ "epoch": 3.86,
3845
+ "eval_loss": 0.3180752694606781,
3846
+ "eval_runtime": 863.1247,
3847
+ "eval_samples_per_second": 2.317,
3848
+ "eval_steps_per_second": 0.29,
3849
+ "step": 6000
3850
+ },
3851
+ {
3852
+ "epoch": 3.86,
3853
+ "learning_rate": 2.2996742671009775e-05,
3854
+ "loss": 0.311,
3855
+ "step": 6010
3856
+ },
3857
+ {
3858
+ "epoch": 3.87,
3859
+ "learning_rate": 2.2866449511400654e-05,
3860
+ "loss": 0.314,
3861
+ "step": 6020
3862
+ },
3863
+ {
3864
+ "epoch": 3.88,
3865
+ "learning_rate": 2.2736156351791533e-05,
3866
+ "loss": 0.3083,
3867
+ "step": 6030
3868
+ },
3869
+ {
3870
+ "epoch": 3.88,
3871
+ "learning_rate": 2.260586319218241e-05,
3872
+ "loss": 0.3023,
3873
+ "step": 6040
3874
+ },
3875
+ {
3876
+ "epoch": 3.89,
3877
+ "learning_rate": 2.247557003257329e-05,
3878
+ "loss": 0.3103,
3879
+ "step": 6050
3880
+ },
3881
+ {
3882
+ "epoch": 3.9,
3883
+ "learning_rate": 2.2345276872964172e-05,
3884
+ "loss": 0.3233,
3885
+ "step": 6060
3886
+ },
3887
+ {
3888
+ "epoch": 3.9,
3889
+ "learning_rate": 2.221498371335505e-05,
3890
+ "loss": 0.3102,
3891
+ "step": 6070
3892
+ },
3893
+ {
3894
+ "epoch": 3.91,
3895
+ "learning_rate": 2.208469055374593e-05,
3896
+ "loss": 0.3118,
3897
+ "step": 6080
3898
+ },
3899
+ {
3900
+ "epoch": 3.92,
3901
+ "learning_rate": 2.1954397394136807e-05,
3902
+ "loss": 0.3086,
3903
+ "step": 6090
3904
+ },
3905
+ {
3906
+ "epoch": 3.92,
3907
+ "learning_rate": 2.182410423452769e-05,
3908
+ "loss": 0.3133,
3909
+ "step": 6100
3910
+ },
3911
+ {
3912
+ "epoch": 3.93,
3913
+ "learning_rate": 2.169381107491857e-05,
3914
+ "loss": 0.3317,
3915
+ "step": 6110
3916
+ },
3917
+ {
3918
+ "epoch": 3.94,
3919
+ "learning_rate": 2.156351791530945e-05,
3920
+ "loss": 0.2972,
3921
+ "step": 6120
3922
+ },
3923
+ {
3924
+ "epoch": 3.94,
3925
+ "learning_rate": 2.1433224755700328e-05,
3926
+ "loss": 0.2986,
3927
+ "step": 6130
3928
+ },
3929
+ {
3930
+ "epoch": 3.95,
3931
+ "learning_rate": 2.1302931596091204e-05,
3932
+ "loss": 0.3128,
3933
+ "step": 6140
3934
+ },
3935
+ {
3936
+ "epoch": 3.95,
3937
+ "learning_rate": 2.1172638436482087e-05,
3938
+ "loss": 0.3078,
3939
+ "step": 6150
3940
+ },
3941
+ {
3942
+ "epoch": 3.96,
3943
+ "learning_rate": 2.1042345276872966e-05,
3944
+ "loss": 0.313,
3945
+ "step": 6160
3946
+ },
3947
+ {
3948
+ "epoch": 3.97,
3949
+ "learning_rate": 2.0912052117263846e-05,
3950
+ "loss": 0.3202,
3951
+ "step": 6170
3952
+ },
3953
+ {
3954
+ "epoch": 3.97,
3955
+ "learning_rate": 2.0781758957654725e-05,
3956
+ "loss": 0.308,
3957
+ "step": 6180
3958
+ },
3959
+ {
3960
+ "epoch": 3.98,
3961
+ "learning_rate": 2.06514657980456e-05,
3962
+ "loss": 0.311,
3963
+ "step": 6190
3964
+ },
3965
+ {
3966
+ "epoch": 3.99,
3967
+ "learning_rate": 2.0521172638436484e-05,
3968
+ "loss": 0.3178,
3969
+ "step": 6200
3970
+ },
3971
+ {
3972
+ "epoch": 3.99,
3973
+ "eval_loss": 0.31802934408187866,
3974
+ "eval_runtime": 861.7623,
3975
+ "eval_samples_per_second": 2.321,
3976
+ "eval_steps_per_second": 0.29,
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": 1.0929527210907156e+19,
3985
+ "trial_name": null,
3986
+ "trial_params": null
3987
+ }
with_input/decomp_code_with_intermediates/code-llama-13b-low-lr/best_model/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fad19c0560d6abe2f9c8bff0bcf8facd99399d87d881cd034de8e305f28be028
3
+ size 4155