Hugging Face's logo

totally-not-an-llm
/
EverythingLM-13b-V2-peft

PEFT
Model card Files Community
1
totally-not-an-llm commited on
Commit
573ff1c
·
1 Parent(s): 3b49de6

Upload folder using huggingface_hub

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. README.md +19 -1
  2. adapter_config.json +26 -0
  3. adapter_model.bin +3 -0
  4. added_tokens.json +3 -0
  5. checkpoint-301/README.md +21 -0
  6. checkpoint-301/adapter_config.json +26 -0
  7. checkpoint-301/adapter_model.bin +3 -0
  8. checkpoint-301/adapter_model/README.md +21 -0
  9. checkpoint-301/adapter_model/adapter_config.json +26 -0
  10. checkpoint-301/adapter_model/adapter_model.bin +3 -0
  11. checkpoint-301/optimizer.pt +3 -0
  12. checkpoint-301/rng_state.pth +3 -0
  13. checkpoint-301/scheduler.pt +3 -0
  14. checkpoint-301/trainer_state.json +1849 -0
  15. checkpoint-301/training_args.bin +3 -0
  16. checkpoint-319/README.md +21 -0
  17. checkpoint-319/adapter_config.json +26 -0
  18. checkpoint-319/adapter_model.bin +3 -0
  19. checkpoint-319/adapter_model/README.md +21 -0
  20. checkpoint-319/adapter_model/adapter_config.json +26 -0
  21. checkpoint-319/adapter_model/adapter_model.bin +3 -0
  22. checkpoint-319/optimizer.pt +3 -0
  23. checkpoint-319/rng_state.pth +3 -0
  24. checkpoint-319/scheduler.pt +3 -0
  25. checkpoint-319/trainer_state.json +1957 -0
  26. checkpoint-319/training_args.bin +3 -0
  27. checkpoint-337/README.md +21 -0
  28. checkpoint-337/adapter_config.json +26 -0
  29. checkpoint-337/adapter_model.bin +3 -0
  30. checkpoint-337/adapter_model/README.md +21 -0
  31. checkpoint-337/adapter_model/adapter_config.json +26 -0
  32. checkpoint-337/adapter_model/adapter_model.bin +3 -0
  33. checkpoint-337/optimizer.pt +3 -0
  34. checkpoint-337/rng_state.pth +3 -0
  35. checkpoint-337/scheduler.pt +3 -0
  36. checkpoint-337/trainer_state.json +2065 -0
  37. checkpoint-337/training_args.bin +3 -0
  38. checkpoint-340/README.md +21 -0
  39. checkpoint-340/adapter_config.json +26 -0
  40. checkpoint-340/adapter_model.bin +3 -0
  41. checkpoint-340/adapter_model/README.md +21 -0
  42. checkpoint-340/adapter_model/adapter_config.json +26 -0
  43. checkpoint-340/adapter_model/adapter_model.bin +3 -0
  44. checkpoint-340/optimizer.pt +3 -0
  45. checkpoint-340/rng_state.pth +3 -0
  46. checkpoint-340/scheduler.pt +3 -0
  47. checkpoint-340/trainer_state.json +2083 -0
  48. checkpoint-340/training_args.bin +3 -0
  49. special_tokens_map.json +6 -0
  50. tokenizer.model +3 -0
README.md CHANGED
@@ -1,3 +1,21 @@
1
  ---
2
- license: llama2
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ library_name: peft
3
  ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0.dev0
adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
4
+ "bias": "none",
5
+ "fan_in_fan_out": null,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 32,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "k_proj",
18
+ "gate_proj",
19
+ "v_proj",
20
+ "o_proj",
21
+ "q_proj",
22
+ "down_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c36d31ee32ee247944f1ebfb0ae67945a2bc9ac80545e7536bf0a970a3c1455
3
+ size 500897101
added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "<pad>": 32000
3
+ }
checkpoint-301/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0.dev0
checkpoint-301/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
4
+ "bias": "none",
5
+ "fan_in_fan_out": null,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 32,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "k_proj",
18
+ "gate_proj",
19
+ "v_proj",
20
+ "o_proj",
21
+ "q_proj",
22
+ "down_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
checkpoint-301/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa799994af1564feb9bd33425def3ac70853e684aaa5809c6419e9911c087a42
3
+ size 500897101
checkpoint-301/adapter_model/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0.dev0
checkpoint-301/adapter_model/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
4
+ "bias": "none",
5
+ "fan_in_fan_out": null,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 32,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "k_proj",
18
+ "gate_proj",
19
+ "v_proj",
20
+ "o_proj",
21
+ "q_proj",
22
+ "down_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
checkpoint-301/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa799994af1564feb9bd33425def3ac70853e684aaa5809c6419e9911c087a42
3
+ size 500897101
checkpoint-301/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:816e9ad360a4d2c1b341d088523849f7e0b49e9e68921d7534d6f7928b064a35
3
+ size 1001737021
checkpoint-301/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9cba857f49c675e3f5d71fa45a477c84c5a39c2dd075c4d65a2f2cf8ae879a77
3
+ size 14575
checkpoint-301/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1ede778dc51796a72a59cd5a7e42dd9fb7142e70f287d7bc13dc2a60a03ceb0f
3
+ size 627
checkpoint-301/trainer_state.json ADDED
@@ -0,0 +1,1849 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 16.95774647887324,
5
+ "eval_steps": 100,
6
+ "global_step": 301,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.06,
13
+ "learning_rate": 5e-06,
14
+ "loss": 1.5896,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.11,
19
+ "learning_rate": 1e-05,
20
+ "loss": 1.6521,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.17,
25
+ "learning_rate": 1.5e-05,
26
+ "loss": 1.3654,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.23,
31
+ "learning_rate": 2e-05,
32
+ "loss": 1.3569,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.28,
37
+ "learning_rate": 2.5e-05,
38
+ "loss": 1.4352,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.34,
43
+ "learning_rate": 3e-05,
44
+ "loss": 1.425,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.39,
49
+ "learning_rate": 3.5e-05,
50
+ "loss": 1.3529,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.45,
55
+ "learning_rate": 4e-05,
56
+ "loss": 1.384,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.51,
61
+ "learning_rate": 4.5e-05,
62
+ "loss": 1.3881,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.56,
67
+ "learning_rate": 5e-05,
68
+ "loss": 1.5593,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.62,
73
+ "learning_rate": 4.999886713385432e-05,
74
+ "loss": 1.3423,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.68,
79
+ "learning_rate": 4.999546863808815e-05,
80
+ "loss": 1.4725,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.73,
85
+ "learning_rate": 4.9989804820704735e-05,
86
+ "loss": 1.4342,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.79,
91
+ "learning_rate": 4.9981876195011844e-05,
92
+ "loss": 1.3769,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.85,
97
+ "learning_rate": 4.99716834795752e-05,
98
+ "loss": 1.3976,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.9,
103
+ "learning_rate": 4.995922759815339e-05,
104
+ "loss": 1.328,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.96,
109
+ "learning_rate": 4.994450967961413e-05,
110
+ "loss": 1.266,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 1.01,
115
+ "learning_rate": 4.992753105783194e-05,
116
+ "loss": 1.4671,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 1.07,
121
+ "learning_rate": 4.9908293271567286e-05,
122
+ "loss": 1.3973,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 1.13,
127
+ "learning_rate": 4.988679806432712e-05,
128
+ "loss": 1.3835,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 1.18,
133
+ "learning_rate": 4.9863047384206835e-05,
134
+ "loss": 1.2203,
135
+ "step": 21
136
+ },
137
+ {
138
+ "epoch": 1.24,
139
+ "learning_rate": 4.9837043383713753e-05,
140
+ "loss": 1.3752,
141
+ "step": 22
142
+ },
143
+ {
144
+ "epoch": 1.3,
145
+ "learning_rate": 4.980878841957203e-05,
146
+ "loss": 1.4142,
147
+ "step": 23
148
+ },
149
+ {
150
+ "epoch": 1.35,
151
+ "learning_rate": 4.977828505250903e-05,
152
+ "loss": 1.3657,
153
+ "step": 24
154
+ },
155
+ {
156
+ "epoch": 1.41,
157
+ "learning_rate": 4.9745536047023324e-05,
158
+ "loss": 1.309,
159
+ "step": 25
160
+ },
161
+ {
162
+ "epoch": 1.46,
163
+ "learning_rate": 4.971054437113406e-05,
164
+ "loss": 1.3004,
165
+ "step": 26
166
+ },
167
+ {
168
+ "epoch": 1.52,
169
+ "learning_rate": 4.967331319611206e-05,
170
+ "loss": 1.3309,
171
+ "step": 27
172
+ },
173
+ {
174
+ "epoch": 1.58,
175
+ "learning_rate": 4.963384589619233e-05,
176
+ "loss": 1.4956,
177
+ "step": 28
178
+ },
179
+ {
180
+ "epoch": 1.63,
181
+ "learning_rate": 4.959214604826831e-05,
182
+ "loss": 1.2682,
183
+ "step": 29
184
+ },
185
+ {
186
+ "epoch": 1.69,
187
+ "learning_rate": 4.9548217431567665e-05,
188
+ "loss": 1.391,
189
+ "step": 30
190
+ },
191
+ {
192
+ "epoch": 1.75,
193
+ "learning_rate": 4.9502064027309836e-05,
194
+ "loss": 1.3102,
195
+ "step": 31
196
+ },
197
+ {
198
+ "epoch": 1.8,
199
+ "learning_rate": 4.9453690018345144e-05,
200
+ "loss": 1.3891,
201
+ "step": 32
202
+ },
203
+ {
204
+ "epoch": 1.86,
205
+ "learning_rate": 4.9403099788775754e-05,
206
+ "loss": 1.3538,
207
+ "step": 33
208
+ },
209
+ {
210
+ "epoch": 1.92,
211
+ "learning_rate": 4.935029792355834e-05,
212
+ "loss": 1.3443,
213
+ "step": 34
214
+ },
215
+ {
216
+ "epoch": 1.97,
217
+ "learning_rate": 4.929528920808854e-05,
218
+ "loss": 1.3384,
219
+ "step": 35
220
+ },
221
+ {
222
+ "epoch": 2.03,
223
+ "learning_rate": 4.923807862776728e-05,
224
+ "loss": 1.2968,
225
+ "step": 36
226
+ },
227
+ {
228
+ "epoch": 2.08,
229
+ "learning_rate": 4.917867136754893e-05,
230
+ "loss": 1.2624,
231
+ "step": 37
232
+ },
233
+ {
234
+ "epoch": 2.14,
235
+ "learning_rate": 4.91170728114714e-05,
236
+ "loss": 1.3547,
237
+ "step": 38
238
+ },
239
+ {
240
+ "epoch": 2.2,
241
+ "learning_rate": 4.9053288542168185e-05,
242
+ "loss": 1.3857,
243
+ "step": 39
244
+ },
245
+ {
246
+ "epoch": 2.25,
247
+ "learning_rate": 4.898732434036244e-05,
248
+ "loss": 1.2261,
249
+ "step": 40
250
+ },
251
+ {
252
+ "epoch": 2.31,
253
+ "learning_rate": 4.8919186184343046e-05,
254
+ "loss": 1.4072,
255
+ "step": 41
256
+ },
257
+ {
258
+ "epoch": 2.37,
259
+ "learning_rate": 4.8848880249422815e-05,
260
+ "loss": 1.3283,
261
+ "step": 42
262
+ },
263
+ {
264
+ "epoch": 2.42,
265
+ "learning_rate": 4.877641290737884e-05,
266
+ "loss": 1.2464,
267
+ "step": 43
268
+ },
269
+ {
270
+ "epoch": 2.48,
271
+ "learning_rate": 4.870179072587499e-05,
272
+ "loss": 1.3433,
273
+ "step": 44
274
+ },
275
+ {
276
+ "epoch": 2.54,
277
+ "learning_rate": 4.862502046786671e-05,
278
+ "loss": 1.2861,
279
+ "step": 45
280
+ },
281
+ {
282
+ "epoch": 2.59,
283
+ "learning_rate": 4.854610909098812e-05,
284
+ "loss": 1.1888,
285
+ "step": 46
286
+ },
287
+ {
288
+ "epoch": 2.65,
289
+ "learning_rate": 4.8465063746921395e-05,
290
+ "loss": 1.2509,
291
+ "step": 47
292
+ },
293
+ {
294
+ "epoch": 2.7,
295
+ "learning_rate": 4.838189178074867e-05,
296
+ "loss": 1.4449,
297
+ "step": 48
298
+ },
299
+ {
300
+ "epoch": 2.76,
301
+ "learning_rate": 4.829660073028631e-05,
302
+ "loss": 1.4645,
303
+ "step": 49
304
+ },
305
+ {
306
+ "epoch": 2.82,
307
+ "learning_rate": 4.8209198325401815e-05,
308
+ "loss": 1.3586,
309
+ "step": 50
310
+ },
311
+ {
312
+ "epoch": 2.87,
313
+ "learning_rate": 4.811969248731323e-05,
314
+ "loss": 1.3542,
315
+ "step": 51
316
+ },
317
+ {
318
+ "epoch": 2.93,
319
+ "learning_rate": 4.802809132787125e-05,
320
+ "loss": 1.3145,
321
+ "step": 52
322
+ },
323
+ {
324
+ "epoch": 2.99,
325
+ "learning_rate": 4.793440314882408e-05,
326
+ "loss": 1.2343,
327
+ "step": 53
328
+ },
329
+ {
330
+ "epoch": 3.04,
331
+ "learning_rate": 4.783863644106502e-05,
332
+ "loss": 1.2946,
333
+ "step": 54
334
+ },
335
+ {
336
+ "epoch": 3.1,
337
+ "learning_rate": 4.774079988386296e-05,
338
+ "loss": 1.2621,
339
+ "step": 55
340
+ },
341
+ {
342
+ "epoch": 3.15,
343
+ "learning_rate": 4.764090234407577e-05,
344
+ "loss": 1.3581,
345
+ "step": 56
346
+ },
347
+ {
348
+ "epoch": 3.21,
349
+ "learning_rate": 4.753895287534673e-05,
350
+ "loss": 1.3891,
351
+ "step": 57
352
+ },
353
+ {
354
+ "epoch": 3.27,
355
+ "learning_rate": 4.743496071728396e-05,
356
+ "loss": 1.334,
357
+ "step": 58
358
+ },
359
+ {
360
+ "epoch": 3.32,
361
+ "learning_rate": 4.73289352946231e-05,
362
+ "loss": 1.3105,
363
+ "step": 59
364
+ },
365
+ {
366
+ "epoch": 3.38,
367
+ "learning_rate": 4.722088621637309e-05,
368
+ "loss": 1.2921,
369
+ "step": 60
370
+ },
371
+ {
372
+ "epoch": 3.44,
373
+ "learning_rate": 4.711082327494536e-05,
374
+ "loss": 1.2406,
375
+ "step": 61
376
+ },
377
+ {
378
+ "epoch": 3.49,
379
+ "learning_rate": 4.6998756445266336e-05,
380
+ "loss": 1.4246,
381
+ "step": 62
382
+ },
383
+ {
384
+ "epoch": 3.55,
385
+ "learning_rate": 4.688469588387339e-05,
386
+ "loss": 1.2126,
387
+ "step": 63
388
+ },
389
+ {
390
+ "epoch": 3.61,
391
+ "learning_rate": 4.6768651927994434e-05,
392
+ "loss": 1.3662,
393
+ "step": 64
394
+ },
395
+ {
396
+ "epoch": 3.66,
397
+ "learning_rate": 4.665063509461097e-05,
398
+ "loss": 1.2044,
399
+ "step": 65
400
+ },
401
+ {
402
+ "epoch": 3.72,
403
+ "learning_rate": 4.653065607950502e-05,
404
+ "loss": 1.4344,
405
+ "step": 66
406
+ },
407
+ {
408
+ "epoch": 3.77,
409
+ "learning_rate": 4.640872575628973e-05,
410
+ "loss": 1.3652,
411
+ "step": 67
412
+ },
413
+ {
414
+ "epoch": 3.83,
415
+ "learning_rate": 4.628485517542392e-05,
416
+ "loss": 1.231,
417
+ "step": 68
418
+ },
419
+ {
420
+ "epoch": 3.89,
421
+ "learning_rate": 4.6159055563210604e-05,
422
+ "loss": 1.1948,
423
+ "step": 69
424
+ },
425
+ {
426
+ "epoch": 3.94,
427
+ "learning_rate": 4.6031338320779534e-05,
428
+ "loss": 1.2484,
429
+ "step": 70
430
+ },
431
+ {
432
+ "epoch": 4.0,
433
+ "learning_rate": 4.59017150230539e-05,
434
+ "loss": 1.2968,
435
+ "step": 71
436
+ },
437
+ {
438
+ "epoch": 4.06,
439
+ "learning_rate": 4.5770197417701365e-05,
440
+ "loss": 1.1691,
441
+ "step": 72
442
+ },
443
+ {
444
+ "epoch": 4.11,
445
+ "learning_rate": 4.563679742406935e-05,
446
+ "loss": 1.224,
447
+ "step": 73
448
+ },
449
+ {
450
+ "epoch": 4.17,
451
+ "learning_rate": 4.550152713210478e-05,
452
+ "loss": 1.2669,
453
+ "step": 74
454
+ },
455
+ {
456
+ "epoch": 4.23,
457
+ "learning_rate": 4.5364398801258396e-05,
458
+ "loss": 1.3218,
459
+ "step": 75
460
+ },
461
+ {
462
+ "epoch": 4.28,
463
+ "learning_rate": 4.522542485937369e-05,
464
+ "loss": 1.3309,
465
+ "step": 76
466
+ },
467
+ {
468
+ "epoch": 4.34,
469
+ "learning_rate": 4.508461790156056e-05,
470
+ "loss": 1.3356,
471
+ "step": 77
472
+ },
473
+ {
474
+ "epoch": 4.39,
475
+ "learning_rate": 4.4941990689053886e-05,
476
+ "loss": 1.3367,
477
+ "step": 78
478
+ },
479
+ {
480
+ "epoch": 4.45,
481
+ "learning_rate": 4.479755614805688e-05,
482
+ "loss": 1.3568,
483
+ "step": 79
484
+ },
485
+ {
486
+ "epoch": 4.51,
487
+ "learning_rate": 4.465132736856969e-05,
488
+ "loss": 1.29,
489
+ "step": 80
490
+ },
491
+ {
492
+ "epoch": 4.56,
493
+ "learning_rate": 4.450331760320302e-05,
494
+ "loss": 1.3522,
495
+ "step": 81
496
+ },
497
+ {
498
+ "epoch": 4.62,
499
+ "learning_rate": 4.4353540265977064e-05,
500
+ "loss": 1.3021,
501
+ "step": 82
502
+ },
503
+ {
504
+ "epoch": 4.68,
505
+ "learning_rate": 4.4202008931105795e-05,
506
+ "loss": 1.303,
507
+ "step": 83
508
+ },
509
+ {
510
+ "epoch": 4.73,
511
+ "learning_rate": 4.404873733176678e-05,
512
+ "loss": 1.313,
513
+ "step": 84
514
+ },
515
+ {
516
+ "epoch": 4.79,
517
+ "learning_rate": 4.389373935885646e-05,
518
+ "loss": 1.231,
519
+ "step": 85
520
+ },
521
+ {
522
+ "epoch": 4.85,
523
+ "learning_rate": 4.373702905973135e-05,
524
+ "loss": 1.2908,
525
+ "step": 86
526
+ },
527
+ {
528
+ "epoch": 4.9,
529
+ "learning_rate": 4.357862063693486e-05,
530
+ "loss": 1.2575,
531
+ "step": 87
532
+ },
533
+ {
534
+ "epoch": 4.96,
535
+ "learning_rate": 4.341852844691012e-05,
536
+ "loss": 1.1942,
537
+ "step": 88
538
+ },
539
+ {
540
+ "epoch": 5.01,
541
+ "learning_rate": 4.3256766998698936e-05,
542
+ "loss": 1.1607,
543
+ "step": 89
544
+ },
545
+ {
546
+ "epoch": 5.07,
547
+ "learning_rate": 4.309335095262676e-05,
548
+ "loss": 1.3386,
549
+ "step": 90
550
+ },
551
+ {
552
+ "epoch": 5.13,
553
+ "learning_rate": 4.292829511897409e-05,
554
+ "loss": 1.371,
555
+ "step": 91
556
+ },
557
+ {
558
+ "epoch": 5.18,
559
+ "learning_rate": 4.276161445663423e-05,
560
+ "loss": 1.236,
561
+ "step": 92
562
+ },
563
+ {
564
+ "epoch": 5.24,
565
+ "learning_rate": 4.259332407175751e-05,
566
+ "loss": 1.2531,
567
+ "step": 93
568
+ },
569
+ {
570
+ "epoch": 5.3,
571
+ "learning_rate": 4.242343921638234e-05,
572
+ "loss": 1.2593,
573
+ "step": 94
574
+ },
575
+ {
576
+ "epoch": 5.35,
577
+ "learning_rate": 4.2251975287052804e-05,
578
+ "loss": 1.1245,
579
+ "step": 95
580
+ },
581
+ {
582
+ "epoch": 5.41,
583
+ "learning_rate": 4.2078947823423364e-05,
584
+ "loss": 1.2538,
585
+ "step": 96
586
+ },
587
+ {
588
+ "epoch": 5.46,
589
+ "learning_rate": 4.1904372506850484e-05,
590
+ "loss": 1.2847,
591
+ "step": 97
592
+ },
593
+ {
594
+ "epoch": 5.52,
595
+ "learning_rate": 4.172826515897146e-05,
596
+ "loss": 1.2071,
597
+ "step": 98
598
+ },
599
+ {
600
+ "epoch": 5.58,
601
+ "learning_rate": 4.155064174027047e-05,
602
+ "loss": 1.2664,
603
+ "step": 99
604
+ },
605
+ {
606
+ "epoch": 5.63,
607
+ "learning_rate": 4.137151834863213e-05,
608
+ "loss": 1.1919,
609
+ "step": 100
610
+ },
611
+ {
612
+ "epoch": 5.63,
613
+ "eval_loss": 1.2261024713516235,
614
+ "eval_runtime": 1.4353,
615
+ "eval_samples_per_second": 6.967,
616
+ "eval_steps_per_second": 3.484,
617
+ "step": 100
618
+ },
619
+ {
620
+ "epoch": 5.69,
621
+ "learning_rate": 4.119091121788256e-05,
622
+ "loss": 1.1509,
623
+ "step": 101
624
+ },
625
+ {
626
+ "epoch": 5.75,
627
+ "learning_rate": 4.100883671631806e-05,
628
+ "loss": 1.3013,
629
+ "step": 102
630
+ },
631
+ {
632
+ "epoch": 5.8,
633
+ "learning_rate": 4.082531134522176e-05,
634
+ "loss": 1.313,
635
+ "step": 103
636
+ },
637
+ {
638
+ "epoch": 5.86,
639
+ "learning_rate": 4.064035173736804e-05,
640
+ "loss": 1.2176,
641
+ "step": 104
642
+ },
643
+ {
644
+ "epoch": 5.92,
645
+ "learning_rate": 4.045397465551513e-05,
646
+ "loss": 1.1004,
647
+ "step": 105
648
+ },
649
+ {
650
+ "epoch": 5.97,
651
+ "learning_rate": 4.0266196990885955e-05,
652
+ "loss": 1.3575,
653
+ "step": 106
654
+ },
655
+ {
656
+ "epoch": 6.03,
657
+ "learning_rate": 4.007703576163724e-05,
658
+ "loss": 1.2923,
659
+ "step": 107
660
+ },
661
+ {
662
+ "epoch": 6.08,
663
+ "learning_rate": 3.98865081113172e-05,
664
+ "loss": 1.2083,
665
+ "step": 108
666
+ },
667
+ {
668
+ "epoch": 6.14,
669
+ "learning_rate": 3.969463130731183e-05,
670
+ "loss": 1.2141,
671
+ "step": 109
672
+ },
673
+ {
674
+ "epoch": 6.2,
675
+ "learning_rate": 3.9501422739279956e-05,
676
+ "loss": 1.3297,
677
+ "step": 110
678
+ },
679
+ {
680
+ "epoch": 6.25,
681
+ "learning_rate": 3.9306899917577245e-05,
682
+ "loss": 1.0901,
683
+ "step": 111
684
+ },
685
+ {
686
+ "epoch": 6.31,
687
+ "learning_rate": 3.911108047166924e-05,
688
+ "loss": 1.2235,
689
+ "step": 112
690
+ },
691
+ {
692
+ "epoch": 6.37,
693
+ "learning_rate": 3.89139821485336e-05,
694
+ "loss": 1.341,
695
+ "step": 113
696
+ },
697
+ {
698
+ "epoch": 6.42,
699
+ "learning_rate": 3.871562281105175e-05,
700
+ "loss": 1.362,
701
+ "step": 114
702
+ },
703
+ {
704
+ "epoch": 6.48,
705
+ "learning_rate": 3.851602043638994e-05,
706
+ "loss": 1.2738,
707
+ "step": 115
708
+ },
709
+ {
710
+ "epoch": 6.54,
711
+ "learning_rate": 3.8315193114369996e-05,
712
+ "loss": 1.2264,
713
+ "step": 116
714
+ },
715
+ {
716
+ "epoch": 6.59,
717
+ "learning_rate": 3.8113159045829864e-05,
718
+ "loss": 1.171,
719
+ "step": 117
720
+ },
721
+ {
722
+ "epoch": 6.65,
723
+ "learning_rate": 3.790993654097405e-05,
724
+ "loss": 1.1848,
725
+ "step": 118
726
+ },
727
+ {
728
+ "epoch": 6.7,
729
+ "learning_rate": 3.770554401771423e-05,
730
+ "loss": 1.1545,
731
+ "step": 119
732
+ },
733
+ {
734
+ "epoch": 6.76,
735
+ "learning_rate": 3.7500000000000003e-05,
736
+ "loss": 1.2799,
737
+ "step": 120
738
+ },
739
+ {
740
+ "epoch": 6.82,
741
+ "learning_rate": 3.72933231161401e-05,
742
+ "loss": 1.2278,
743
+ "step": 121
744
+ },
745
+ {
746
+ "epoch": 6.87,
747
+ "learning_rate": 3.708553209711409e-05,
748
+ "loss": 1.1959,
749
+ "step": 122
750
+ },
751
+ {
752
+ "epoch": 6.93,
753
+ "learning_rate": 3.6876645774874876e-05,
754
+ "loss": 1.1151,
755
+ "step": 123
756
+ },
757
+ {
758
+ "epoch": 6.99,
759
+ "learning_rate": 3.6666683080641846e-05,
760
+ "loss": 1.2178,
761
+ "step": 124
762
+ },
763
+ {
764
+ "epoch": 7.04,
765
+ "learning_rate": 3.645566304318526e-05,
766
+ "loss": 1.1584,
767
+ "step": 125
768
+ },
769
+ {
770
+ "epoch": 7.1,
771
+ "learning_rate": 3.624360478710165e-05,
772
+ "loss": 1.0959,
773
+ "step": 126
774
+ },
775
+ {
776
+ "epoch": 7.15,
777
+ "learning_rate": 3.603052753108053e-05,
778
+ "loss": 1.3026,
779
+ "step": 127
780
+ },
781
+ {
782
+ "epoch": 7.21,
783
+ "learning_rate": 3.581645058616271e-05,
784
+ "loss": 1.2007,
785
+ "step": 128
786
+ },
787
+ {
788
+ "epoch": 7.27,
789
+ "learning_rate": 3.5601393353990046e-05,
790
+ "loss": 1.2242,
791
+ "step": 129
792
+ },
793
+ {
794
+ "epoch": 7.32,
795
+ "learning_rate": 3.5385375325047166e-05,
796
+ "loss": 1.2737,
797
+ "step": 130
798
+ },
799
+ {
800
+ "epoch": 7.38,
801
+ "learning_rate": 3.516841607689501e-05,
802
+ "loss": 1.2724,
803
+ "step": 131
804
+ },
805
+ {
806
+ "epoch": 7.44,
807
+ "learning_rate": 3.495053527239656e-05,
808
+ "loss": 1.1916,
809
+ "step": 132
810
+ },
811
+ {
812
+ "epoch": 7.49,
813
+ "learning_rate": 3.4731752657934794e-05,
814
+ "loss": 1.2107,
815
+ "step": 133
816
+ },
817
+ {
818
+ "epoch": 7.55,
819
+ "learning_rate": 3.4512088061623075e-05,
820
+ "loss": 1.246,
821
+ "step": 134
822
+ },
823
+ {
824
+ "epoch": 7.61,
825
+ "learning_rate": 3.4291561391508185e-05,
826
+ "loss": 1.1841,
827
+ "step": 135
828
+ },
829
+ {
830
+ "epoch": 7.66,
831
+ "learning_rate": 3.4070192633766025e-05,
832
+ "loss": 1.1898,
833
+ "step": 136
834
+ },
835
+ {
836
+ "epoch": 7.72,
837
+ "learning_rate": 3.3848001850890296e-05,
838
+ "loss": 1.1195,
839
+ "step": 137
840
+ },
841
+ {
842
+ "epoch": 7.77,
843
+ "learning_rate": 3.3625009179874265e-05,
844
+ "loss": 1.1685,
845
+ "step": 138
846
+ },
847
+ {
848
+ "epoch": 7.83,
849
+ "learning_rate": 3.3401234830385756e-05,
850
+ "loss": 1.2748,
851
+ "step": 139
852
+ },
853
+ {
854
+ "epoch": 7.89,
855
+ "learning_rate": 3.3176699082935545e-05,
856
+ "loss": 1.1697,
857
+ "step": 140
858
+ },
859
+ {
860
+ "epoch": 7.94,
861
+ "learning_rate": 3.295142228703938e-05,
862
+ "loss": 1.1624,
863
+ "step": 141
864
+ },
865
+ {
866
+ "epoch": 8.0,
867
+ "learning_rate": 3.272542485937369e-05,
868
+ "loss": 1.1679,
869
+ "step": 142
870
+ },
871
+ {
872
+ "epoch": 8.06,
873
+ "learning_rate": 3.249872728192527e-05,
874
+ "loss": 1.1163,
875
+ "step": 143
876
+ },
877
+ {
878
+ "epoch": 8.11,
879
+ "learning_rate": 3.2271350100134975e-05,
880
+ "loss": 1.1319,
881
+ "step": 144
882
+ },
883
+ {
884
+ "epoch": 8.17,
885
+ "learning_rate": 3.2043313921035743e-05,
886
+ "loss": 1.2151,
887
+ "step": 145
888
+ },
889
+ {
890
+ "epoch": 8.23,
891
+ "learning_rate": 3.181463941138495e-05,
892
+ "loss": 1.1339,
893
+ "step": 146
894
+ },
895
+ {
896
+ "epoch": 8.28,
897
+ "learning_rate": 3.158534729579142e-05,
898
+ "loss": 1.1503,
899
+ "step": 147
900
+ },
901
+ {
902
+ "epoch": 8.34,
903
+ "learning_rate": 3.135545835483718e-05,
904
+ "loss": 1.2326,
905
+ "step": 148
906
+ },
907
+ {
908
+ "epoch": 8.39,
909
+ "learning_rate": 3.11249934231941e-05,
910
+ "loss": 1.2938,
911
+ "step": 149
912
+ },
913
+ {
914
+ "epoch": 8.45,
915
+ "learning_rate": 3.0893973387735687e-05,
916
+ "loss": 1.1454,
917
+ "step": 150
918
+ },
919
+ {
920
+ "epoch": 8.51,
921
+ "learning_rate": 3.0662419185644115e-05,
922
+ "loss": 1.2504,
923
+ "step": 151
924
+ },
925
+ {
926
+ "epoch": 8.56,
927
+ "learning_rate": 3.0430351802512698e-05,
928
+ "loss": 0.987,
929
+ "step": 152
930
+ },
931
+ {
932
+ "epoch": 8.62,
933
+ "learning_rate": 3.0197792270443982e-05,
934
+ "loss": 1.1387,
935
+ "step": 153
936
+ },
937
+ {
938
+ "epoch": 8.68,
939
+ "learning_rate": 2.996476166614364e-05,
940
+ "loss": 1.1142,
941
+ "step": 154
942
+ },
943
+ {
944
+ "epoch": 8.73,
945
+ "learning_rate": 2.9731281109010256e-05,
946
+ "loss": 1.1714,
947
+ "step": 155
948
+ },
949
+ {
950
+ "epoch": 8.79,
951
+ "learning_rate": 2.9497371759221347e-05,
952
+ "loss": 1.1858,
953
+ "step": 156
954
+ },
955
+ {
956
+ "epoch": 8.85,
957
+ "learning_rate": 2.92630548158156e-05,
958
+ "loss": 1.2169,
959
+ "step": 157
960
+ },
961
+ {
962
+ "epoch": 8.9,
963
+ "learning_rate": 2.9028351514771606e-05,
964
+ "loss": 1.2204,
965
+ "step": 158
966
+ },
967
+ {
968
+ "epoch": 8.96,
969
+ "learning_rate": 2.8793283127083292e-05,
970
+ "loss": 1.0313,
971
+ "step": 159
972
+ },
973
+ {
974
+ "epoch": 9.01,
975
+ "learning_rate": 2.8557870956832132e-05,
976
+ "loss": 1.2432,
977
+ "step": 160
978
+ },
979
+ {
980
+ "epoch": 9.07,
981
+ "learning_rate": 2.8322136339256356e-05,
982
+ "loss": 1.0614,
983
+ "step": 161
984
+ },
985
+ {
986
+ "epoch": 9.13,
987
+ "learning_rate": 2.808610063881737e-05,
988
+ "loss": 1.1682,
989
+ "step": 162
990
+ },
991
+ {
992
+ "epoch": 9.18,
993
+ "learning_rate": 2.7849785247263515e-05,
994
+ "loss": 1.0238,
995
+ "step": 163
996
+ },
997
+ {
998
+ "epoch": 9.24,
999
+ "learning_rate": 2.761321158169134e-05,
1000
+ "loss": 1.166,
1001
+ "step": 164
1002
+ },
1003
+ {
1004
+ "epoch": 9.3,
1005
+ "learning_rate": 2.7376401082604564e-05,
1006
+ "loss": 1.0965,
1007
+ "step": 165
1008
+ },
1009
+ {
1010
+ "epoch": 9.35,
1011
+ "learning_rate": 2.7139375211970996e-05,
1012
+ "loss": 1.096,
1013
+ "step": 166
1014
+ },
1015
+ {
1016
+ "epoch": 9.41,
1017
+ "learning_rate": 2.6902155451277377e-05,
1018
+ "loss": 1.1161,
1019
+ "step": 167
1020
+ },
1021
+ {
1022
+ "epoch": 9.46,
1023
+ "learning_rate": 2.6664763299582602e-05,
1024
+ "loss": 1.1328,
1025
+ "step": 168
1026
+ },
1027
+ {
1028
+ "epoch": 9.52,
1029
+ "learning_rate": 2.6427220271569203e-05,
1030
+ "loss": 1.0747,
1031
+ "step": 169
1032
+ },
1033
+ {
1034
+ "epoch": 9.58,
1035
+ "learning_rate": 2.6189547895593562e-05,
1036
+ "loss": 1.1092,
1037
+ "step": 170
1038
+ },
1039
+ {
1040
+ "epoch": 9.63,
1041
+ "learning_rate": 2.5951767711734753e-05,
1042
+ "loss": 1.1624,
1043
+ "step": 171
1044
+ },
1045
+ {
1046
+ "epoch": 9.69,
1047
+ "learning_rate": 2.5713901269842404e-05,
1048
+ "loss": 1.1434,
1049
+ "step": 172
1050
+ },
1051
+ {
1052
+ "epoch": 9.75,
1053
+ "learning_rate": 2.5475970127583666e-05,
1054
+ "loss": 1.1167,
1055
+ "step": 173
1056
+ },
1057
+ {
1058
+ "epoch": 9.8,
1059
+ "learning_rate": 2.523799584848942e-05,
1060
+ "loss": 1.0509,
1061
+ "step": 174
1062
+ },
1063
+ {
1064
+ "epoch": 9.86,
1065
+ "learning_rate": 2.5e-05,
1066
+ "loss": 1.1449,
1067
+ "step": 175
1068
+ },
1069
+ {
1070
+ "epoch": 9.92,
1071
+ "learning_rate": 2.4762004151510584e-05,
1072
+ "loss": 1.109,
1073
+ "step": 176
1074
+ },
1075
+ {
1076
+ "epoch": 9.97,
1077
+ "learning_rate": 2.4524029872416333e-05,
1078
+ "loss": 1.0272,
1079
+ "step": 177
1080
+ },
1081
+ {
1082
+ "epoch": 10.03,
1083
+ "learning_rate": 2.42860987301576e-05,
1084
+ "loss": 1.1489,
1085
+ "step": 178
1086
+ },
1087
+ {
1088
+ "epoch": 10.08,
1089
+ "learning_rate": 2.4048232288265253e-05,
1090
+ "loss": 1.0263,
1091
+ "step": 179
1092
+ },
1093
+ {
1094
+ "epoch": 10.14,
1095
+ "learning_rate": 2.3810452104406444e-05,
1096
+ "loss": 1.0305,
1097
+ "step": 180
1098
+ },
1099
+ {
1100
+ "epoch": 10.2,
1101
+ "learning_rate": 2.35727797284308e-05,
1102
+ "loss": 0.9612,
1103
+ "step": 181
1104
+ },
1105
+ {
1106
+ "epoch": 10.25,
1107
+ "learning_rate": 2.3335236700417404e-05,
1108
+ "loss": 1.1883,
1109
+ "step": 182
1110
+ },
1111
+ {
1112
+ "epoch": 10.31,
1113
+ "learning_rate": 2.309784454872262e-05,
1114
+ "loss": 1.192,
1115
+ "step": 183
1116
+ },
1117
+ {
1118
+ "epoch": 10.37,
1119
+ "learning_rate": 2.2860624788029013e-05,
1120
+ "loss": 1.0919,
1121
+ "step": 184
1122
+ },
1123
+ {
1124
+ "epoch": 10.42,
1125
+ "learning_rate": 2.2623598917395438e-05,
1126
+ "loss": 1.0974,
1127
+ "step": 185
1128
+ },
1129
+ {
1130
+ "epoch": 10.48,
1131
+ "learning_rate": 2.238678841830867e-05,
1132
+ "loss": 1.0938,
1133
+ "step": 186
1134
+ },
1135
+ {
1136
+ "epoch": 10.54,
1137
+ "learning_rate": 2.2150214752736488e-05,
1138
+ "loss": 0.9879,
1139
+ "step": 187
1140
+ },
1141
+ {
1142
+ "epoch": 10.59,
1143
+ "learning_rate": 2.1913899361182632e-05,
1144
+ "loss": 1.0284,
1145
+ "step": 188
1146
+ },
1147
+ {
1148
+ "epoch": 10.65,
1149
+ "learning_rate": 2.167786366074365e-05,
1150
+ "loss": 1.0184,
1151
+ "step": 189
1152
+ },
1153
+ {
1154
+ "epoch": 10.7,
1155
+ "learning_rate": 2.1442129043167874e-05,
1156
+ "loss": 1.0419,
1157
+ "step": 190
1158
+ },
1159
+ {
1160
+ "epoch": 10.76,
1161
+ "learning_rate": 2.1206716872916714e-05,
1162
+ "loss": 1.0631,
1163
+ "step": 191
1164
+ },
1165
+ {
1166
+ "epoch": 10.82,
1167
+ "learning_rate": 2.09716484852284e-05,
1168
+ "loss": 1.0492,
1169
+ "step": 192
1170
+ },
1171
+ {
1172
+ "epoch": 10.87,
1173
+ "learning_rate": 2.0736945184184405e-05,
1174
+ "loss": 1.0924,
1175
+ "step": 193
1176
+ },
1177
+ {
1178
+ "epoch": 10.93,
1179
+ "learning_rate": 2.0502628240778655e-05,
1180
+ "loss": 0.9832,
1181
+ "step": 194
1182
+ },
1183
+ {
1184
+ "epoch": 10.99,
1185
+ "learning_rate": 2.0268718890989753e-05,
1186
+ "loss": 1.1266,
1187
+ "step": 195
1188
+ },
1189
+ {
1190
+ "epoch": 11.04,
1191
+ "learning_rate": 2.003523833385637e-05,
1192
+ "loss": 1.0762,
1193
+ "step": 196
1194
+ },
1195
+ {
1196
+ "epoch": 11.1,
1197
+ "learning_rate": 1.980220772955602e-05,
1198
+ "loss": 1.0011,
1199
+ "step": 197
1200
+ },
1201
+ {
1202
+ "epoch": 11.15,
1203
+ "learning_rate": 1.956964819748731e-05,
1204
+ "loss": 1.0195,
1205
+ "step": 198
1206
+ },
1207
+ {
1208
+ "epoch": 11.21,
1209
+ "learning_rate": 1.9337580814355888e-05,
1210
+ "loss": 1.0442,
1211
+ "step": 199
1212
+ },
1213
+ {
1214
+ "epoch": 11.27,
1215
+ "learning_rate": 1.9106026612264316e-05,
1216
+ "loss": 1.0285,
1217
+ "step": 200
1218
+ },
1219
+ {
1220
+ "epoch": 11.27,
1221
+ "eval_loss": 1.3292417526245117,
1222
+ "eval_runtime": 1.4356,
1223
+ "eval_samples_per_second": 6.966,
1224
+ "eval_steps_per_second": 3.483,
1225
+ "step": 200
1226
+ },
1227
+ {
1228
+ "epoch": 11.32,
1229
+ "learning_rate": 1.8875006576805914e-05,
1230
+ "loss": 0.9805,
1231
+ "step": 201
1232
+ },
1233
+ {
1234
+ "epoch": 11.38,
1235
+ "learning_rate": 1.8644541645162834e-05,
1236
+ "loss": 0.9562,
1237
+ "step": 202
1238
+ },
1239
+ {
1240
+ "epoch": 11.44,
1241
+ "learning_rate": 1.8414652704208583e-05,
1242
+ "loss": 1.0457,
1243
+ "step": 203
1244
+ },
1245
+ {
1246
+ "epoch": 11.49,
1247
+ "learning_rate": 1.8185360588615058e-05,
1248
+ "loss": 1.0217,
1249
+ "step": 204
1250
+ },
1251
+ {
1252
+ "epoch": 11.55,
1253
+ "learning_rate": 1.795668607896426e-05,
1254
+ "loss": 1.1281,
1255
+ "step": 205
1256
+ },
1257
+ {
1258
+ "epoch": 11.61,
1259
+ "learning_rate": 1.7728649899865024e-05,
1260
+ "loss": 0.9654,
1261
+ "step": 206
1262
+ },
1263
+ {
1264
+ "epoch": 11.66,
1265
+ "learning_rate": 1.7501272718074737e-05,
1266
+ "loss": 0.9011,
1267
+ "step": 207
1268
+ },
1269
+ {
1270
+ "epoch": 11.72,
1271
+ "learning_rate": 1.7274575140626318e-05,
1272
+ "loss": 1.097,
1273
+ "step": 208
1274
+ },
1275
+ {
1276
+ "epoch": 11.77,
1277
+ "learning_rate": 1.7048577712960627e-05,
1278
+ "loss": 1.0652,
1279
+ "step": 209
1280
+ },
1281
+ {
1282
+ "epoch": 11.83,
1283
+ "learning_rate": 1.682330091706446e-05,
1284
+ "loss": 1.011,
1285
+ "step": 210
1286
+ },
1287
+ {
1288
+ "epoch": 11.89,
1289
+ "learning_rate": 1.6598765169614243e-05,
1290
+ "loss": 1.1001,
1291
+ "step": 211
1292
+ },
1293
+ {
1294
+ "epoch": 11.94,
1295
+ "learning_rate": 1.637499082012574e-05,
1296
+ "loss": 1.1224,
1297
+ "step": 212
1298
+ },
1299
+ {
1300
+ "epoch": 12.0,
1301
+ "learning_rate": 1.615199814910971e-05,
1302
+ "loss": 0.967,
1303
+ "step": 213
1304
+ },
1305
+ {
1306
+ "epoch": 12.06,
1307
+ "learning_rate": 1.5929807366233977e-05,
1308
+ "loss": 0.9923,
1309
+ "step": 214
1310
+ },
1311
+ {
1312
+ "epoch": 12.11,
1313
+ "learning_rate": 1.5708438608491814e-05,
1314
+ "loss": 1.0722,
1315
+ "step": 215
1316
+ },
1317
+ {
1318
+ "epoch": 12.17,
1319
+ "learning_rate": 1.5487911938376924e-05,
1320
+ "loss": 0.9666,
1321
+ "step": 216
1322
+ },
1323
+ {
1324
+ "epoch": 12.23,
1325
+ "learning_rate": 1.5268247342065215e-05,
1326
+ "loss": 0.9136,
1327
+ "step": 217
1328
+ },
1329
+ {
1330
+ "epoch": 12.28,
1331
+ "learning_rate": 1.5049464727603452e-05,
1332
+ "loss": 0.9274,
1333
+ "step": 218
1334
+ },
1335
+ {
1336
+ "epoch": 12.34,
1337
+ "learning_rate": 1.4831583923104999e-05,
1338
+ "loss": 1.0955,
1339
+ "step": 219
1340
+ },
1341
+ {
1342
+ "epoch": 12.39,
1343
+ "learning_rate": 1.4614624674952842e-05,
1344
+ "loss": 0.9058,
1345
+ "step": 220
1346
+ },
1347
+ {
1348
+ "epoch": 12.45,
1349
+ "learning_rate": 1.4398606646009963e-05,
1350
+ "loss": 0.8351,
1351
+ "step": 221
1352
+ },
1353
+ {
1354
+ "epoch": 12.51,
1355
+ "learning_rate": 1.4183549413837289e-05,
1356
+ "loss": 1.119,
1357
+ "step": 222
1358
+ },
1359
+ {
1360
+ "epoch": 12.56,
1361
+ "learning_rate": 1.3969472468919461e-05,
1362
+ "loss": 1.0418,
1363
+ "step": 223
1364
+ },
1365
+ {
1366
+ "epoch": 12.62,
1367
+ "learning_rate": 1.3756395212898359e-05,
1368
+ "loss": 0.9102,
1369
+ "step": 224
1370
+ },
1371
+ {
1372
+ "epoch": 12.68,
1373
+ "learning_rate": 1.354433695681474e-05,
1374
+ "loss": 0.997,
1375
+ "step": 225
1376
+ },
1377
+ {
1378
+ "epoch": 12.73,
1379
+ "learning_rate": 1.3333316919358157e-05,
1380
+ "loss": 1.1737,
1381
+ "step": 226
1382
+ },
1383
+ {
1384
+ "epoch": 12.79,
1385
+ "learning_rate": 1.3123354225125128e-05,
1386
+ "loss": 0.9887,
1387
+ "step": 227
1388
+ },
1389
+ {
1390
+ "epoch": 12.85,
1391
+ "learning_rate": 1.2914467902885901e-05,
1392
+ "loss": 0.9984,
1393
+ "step": 228
1394
+ },
1395
+ {
1396
+ "epoch": 12.9,
1397
+ "learning_rate": 1.2706676883859903e-05,
1398
+ "loss": 0.9507,
1399
+ "step": 229
1400
+ },
1401
+ {
1402
+ "epoch": 12.96,
1403
+ "learning_rate": 1.2500000000000006e-05,
1404
+ "loss": 0.9395,
1405
+ "step": 230
1406
+ },
1407
+ {
1408
+ "epoch": 13.01,
1409
+ "learning_rate": 1.2294455982285777e-05,
1410
+ "loss": 1.1445,
1411
+ "step": 231
1412
+ },
1413
+ {
1414
+ "epoch": 13.07,
1415
+ "learning_rate": 1.2090063459025955e-05,
1416
+ "loss": 0.8469,
1417
+ "step": 232
1418
+ },
1419
+ {
1420
+ "epoch": 13.13,
1421
+ "learning_rate": 1.1886840954170142e-05,
1422
+ "loss": 0.9936,
1423
+ "step": 233
1424
+ },
1425
+ {
1426
+ "epoch": 13.18,
1427
+ "learning_rate": 1.1684806885630004e-05,
1428
+ "loss": 0.8797,
1429
+ "step": 234
1430
+ },
1431
+ {
1432
+ "epoch": 13.24,
1433
+ "learning_rate": 1.148397956361007e-05,
1434
+ "loss": 1.1331,
1435
+ "step": 235
1436
+ },
1437
+ {
1438
+ "epoch": 13.3,
1439
+ "learning_rate": 1.1284377188948258e-05,
1440
+ "loss": 0.9613,
1441
+ "step": 236
1442
+ },
1443
+ {
1444
+ "epoch": 13.35,
1445
+ "learning_rate": 1.1086017851466413e-05,
1446
+ "loss": 0.9343,
1447
+ "step": 237
1448
+ },
1449
+ {
1450
+ "epoch": 13.41,
1451
+ "learning_rate": 1.0888919528330777e-05,
1452
+ "loss": 1.0098,
1453
+ "step": 238
1454
+ },
1455
+ {
1456
+ "epoch": 13.46,
1457
+ "learning_rate": 1.0693100082422763e-05,
1458
+ "loss": 0.9669,
1459
+ "step": 239
1460
+ },
1461
+ {
1462
+ "epoch": 13.52,
1463
+ "learning_rate": 1.049857726072005e-05,
1464
+ "loss": 0.9686,
1465
+ "step": 240
1466
+ },
1467
+ {
1468
+ "epoch": 13.58,
1469
+ "learning_rate": 1.0305368692688174e-05,
1470
+ "loss": 0.8916,
1471
+ "step": 241
1472
+ },
1473
+ {
1474
+ "epoch": 13.63,
1475
+ "learning_rate": 1.0113491888682802e-05,
1476
+ "loss": 0.9274,
1477
+ "step": 242
1478
+ },
1479
+ {
1480
+ "epoch": 13.69,
1481
+ "learning_rate": 9.922964238362762e-06,
1482
+ "loss": 0.9307,
1483
+ "step": 243
1484
+ },
1485
+ {
1486
+ "epoch": 13.75,
1487
+ "learning_rate": 9.733803009114045e-06,
1488
+ "loss": 1.0112,
1489
+ "step": 244
1490
+ },
1491
+ {
1492
+ "epoch": 13.8,
1493
+ "learning_rate": 9.546025344484869e-06,
1494
+ "loss": 1.0465,
1495
+ "step": 245
1496
+ },
1497
+ {
1498
+ "epoch": 13.86,
1499
+ "learning_rate": 9.359648262631961e-06,
1500
+ "loss": 1.0822,
1501
+ "step": 246
1502
+ },
1503
+ {
1504
+ "epoch": 13.92,
1505
+ "learning_rate": 9.174688654778243e-06,
1506
+ "loss": 0.8992,
1507
+ "step": 247
1508
+ },
1509
+ {
1510
+ "epoch": 13.97,
1511
+ "learning_rate": 8.991163283681944e-06,
1512
+ "loss": 0.8778,
1513
+ "step": 248
1514
+ },
1515
+ {
1516
+ "epoch": 14.03,
1517
+ "learning_rate": 8.809088782117452e-06,
1518
+ "loss": 1.0079,
1519
+ "step": 249
1520
+ },
1521
+ {
1522
+ "epoch": 14.08,
1523
+ "learning_rate": 8.628481651367876e-06,
1524
+ "loss": 0.9508,
1525
+ "step": 250
1526
+ },
1527
+ {
1528
+ "epoch": 14.14,
1529
+ "learning_rate": 8.449358259729537e-06,
1530
+ "loss": 0.997,
1531
+ "step": 251
1532
+ },
1533
+ {
1534
+ "epoch": 14.2,
1535
+ "learning_rate": 8.271734841028553e-06,
1536
+ "loss": 0.8669,
1537
+ "step": 252
1538
+ },
1539
+ {
1540
+ "epoch": 14.25,
1541
+ "learning_rate": 8.09562749314952e-06,
1542
+ "loss": 0.9721,
1543
+ "step": 253
1544
+ },
1545
+ {
1546
+ "epoch": 14.31,
1547
+ "learning_rate": 7.921052176576644e-06,
1548
+ "loss": 1.0777,
1549
+ "step": 254
1550
+ },
1551
+ {
1552
+ "epoch": 14.37,
1553
+ "learning_rate": 7.748024712947205e-06,
1554
+ "loss": 0.8644,
1555
+ "step": 255
1556
+ },
1557
+ {
1558
+ "epoch": 14.42,
1559
+ "learning_rate": 7.576560783617668e-06,
1560
+ "loss": 0.9119,
1561
+ "step": 256
1562
+ },
1563
+ {
1564
+ "epoch": 14.48,
1565
+ "learning_rate": 7.40667592824249e-06,
1566
+ "loss": 0.8645,
1567
+ "step": 257
1568
+ },
1569
+ {
1570
+ "epoch": 14.54,
1571
+ "learning_rate": 7.238385543365783e-06,
1572
+ "loss": 0.8786,
1573
+ "step": 258
1574
+ },
1575
+ {
1576
+ "epoch": 14.59,
1577
+ "learning_rate": 7.071704881025915e-06,
1578
+ "loss": 0.9539,
1579
+ "step": 259
1580
+ },
1581
+ {
1582
+ "epoch": 14.65,
1583
+ "learning_rate": 6.906649047373246e-06,
1584
+ "loss": 0.9521,
1585
+ "step": 260
1586
+ },
1587
+ {
1588
+ "epoch": 14.7,
1589
+ "learning_rate": 6.743233001301067e-06,
1590
+ "loss": 1.0708,
1591
+ "step": 261
1592
+ },
1593
+ {
1594
+ "epoch": 14.76,
1595
+ "learning_rate": 6.5814715530898745e-06,
1596
+ "loss": 1.033,
1597
+ "step": 262
1598
+ },
1599
+ {
1600
+ "epoch": 14.82,
1601
+ "learning_rate": 6.421379363065142e-06,
1602
+ "loss": 0.9344,
1603
+ "step": 263
1604
+ },
1605
+ {
1606
+ "epoch": 14.87,
1607
+ "learning_rate": 6.2629709402686535e-06,
1608
+ "loss": 0.9491,
1609
+ "step": 264
1610
+ },
1611
+ {
1612
+ "epoch": 14.93,
1613
+ "learning_rate": 6.106260641143546e-06,
1614
+ "loss": 0.8894,
1615
+ "step": 265
1616
+ },
1617
+ {
1618
+ "epoch": 14.99,
1619
+ "learning_rate": 5.951262668233232e-06,
1620
+ "loss": 0.8825,
1621
+ "step": 266
1622
+ },
1623
+ {
1624
+ "epoch": 15.04,
1625
+ "learning_rate": 5.797991068894201e-06,
1626
+ "loss": 0.925,
1627
+ "step": 267
1628
+ },
1629
+ {
1630
+ "epoch": 15.1,
1631
+ "learning_rate": 5.646459734022938e-06,
1632
+ "loss": 0.8431,
1633
+ "step": 268
1634
+ },
1635
+ {
1636
+ "epoch": 15.15,
1637
+ "learning_rate": 5.49668239679699e-06,
1638
+ "loss": 1.036,
1639
+ "step": 269
1640
+ },
1641
+ {
1642
+ "epoch": 15.21,
1643
+ "learning_rate": 5.348672631430318e-06,
1644
+ "loss": 0.9463,
1645
+ "step": 270
1646
+ },
1647
+ {
1648
+ "epoch": 15.27,
1649
+ "learning_rate": 5.202443851943126e-06,
1650
+ "loss": 0.9942,
1651
+ "step": 271
1652
+ },
1653
+ {
1654
+ "epoch": 15.32,
1655
+ "learning_rate": 5.058009310946119e-06,
1656
+ "loss": 0.9657,
1657
+ "step": 272
1658
+ },
1659
+ {
1660
+ "epoch": 15.38,
1661
+ "learning_rate": 4.915382098439436e-06,
1662
+ "loss": 0.9375,
1663
+ "step": 273
1664
+ },
1665
+ {
1666
+ "epoch": 15.44,
1667
+ "learning_rate": 4.7745751406263165e-06,
1668
+ "loss": 0.9347,
1669
+ "step": 274
1670
+ },
1671
+ {
1672
+ "epoch": 15.49,
1673
+ "learning_rate": 4.6356011987416075e-06,
1674
+ "loss": 0.9796,
1675
+ "step": 275
1676
+ },
1677
+ {
1678
+ "epoch": 15.55,
1679
+ "learning_rate": 4.498472867895223e-06,
1680
+ "loss": 0.987,
1681
+ "step": 276
1682
+ },
1683
+ {
1684
+ "epoch": 15.61,
1685
+ "learning_rate": 4.36320257593065e-06,
1686
+ "loss": 0.8946,
1687
+ "step": 277
1688
+ },
1689
+ {
1690
+ "epoch": 15.66,
1691
+ "learning_rate": 4.229802582298633e-06,
1692
+ "loss": 0.9034,
1693
+ "step": 278
1694
+ },
1695
+ {
1696
+ "epoch": 15.72,
1697
+ "learning_rate": 4.098284976946101e-06,
1698
+ "loss": 1.0294,
1699
+ "step": 279
1700
+ },
1701
+ {
1702
+ "epoch": 15.77,
1703
+ "learning_rate": 3.968661679220468e-06,
1704
+ "loss": 0.9631,
1705
+ "step": 280
1706
+ },
1707
+ {
1708
+ "epoch": 15.83,
1709
+ "learning_rate": 3.840944436789393e-06,
1710
+ "loss": 0.8831,
1711
+ "step": 281
1712
+ },
1713
+ {
1714
+ "epoch": 15.89,
1715
+ "learning_rate": 3.715144824576078e-06,
1716
+ "loss": 0.9274,
1717
+ "step": 282
1718
+ },
1719
+ {
1720
+ "epoch": 15.94,
1721
+ "learning_rate": 3.591274243710277e-06,
1722
+ "loss": 0.8896,
1723
+ "step": 283
1724
+ },
1725
+ {
1726
+ "epoch": 16.0,
1727
+ "learning_rate": 3.4693439204949858e-06,
1728
+ "loss": 0.8672,
1729
+ "step": 284
1730
+ },
1731
+ {
1732
+ "epoch": 16.06,
1733
+ "learning_rate": 3.3493649053890326e-06,
1734
+ "loss": 0.9043,
1735
+ "step": 285
1736
+ },
1737
+ {
1738
+ "epoch": 16.11,
1739
+ "learning_rate": 3.2313480720055745e-06,
1740
+ "loss": 1.1059,
1741
+ "step": 286
1742
+ },
1743
+ {
1744
+ "epoch": 16.17,
1745
+ "learning_rate": 3.1153041161266134e-06,
1746
+ "loss": 0.9241,
1747
+ "step": 287
1748
+ },
1749
+ {
1750
+ "epoch": 16.23,
1751
+ "learning_rate": 3.0012435547336737e-06,
1752
+ "loss": 0.7761,
1753
+ "step": 288
1754
+ },
1755
+ {
1756
+ "epoch": 16.28,
1757
+ "learning_rate": 2.889176725054643e-06,
1758
+ "loss": 1.0323,
1759
+ "step": 289
1760
+ },
1761
+ {
1762
+ "epoch": 16.34,
1763
+ "learning_rate": 2.779113783626916e-06,
1764
+ "loss": 0.8534,
1765
+ "step": 290
1766
+ },
1767
+ {
1768
+ "epoch": 16.39,
1769
+ "learning_rate": 2.671064705376905e-06,
1770
+ "loss": 0.7669,
1771
+ "step": 291
1772
+ },
1773
+ {
1774
+ "epoch": 16.45,
1775
+ "learning_rate": 2.565039282716045e-06,
1776
+ "loss": 0.984,
1777
+ "step": 292
1778
+ },
1779
+ {
1780
+ "epoch": 16.51,
1781
+ "learning_rate": 2.461047124653279e-06,
1782
+ "loss": 0.9542,
1783
+ "step": 293
1784
+ },
1785
+ {
1786
+ "epoch": 16.56,
1787
+ "learning_rate": 2.3590976559242278e-06,
1788
+ "loss": 0.9439,
1789
+ "step": 294
1790
+ },
1791
+ {
1792
+ "epoch": 16.62,
1793
+ "learning_rate": 2.2592001161370392e-06,
1794
+ "loss": 0.8776,
1795
+ "step": 295
1796
+ },
1797
+ {
1798
+ "epoch": 16.68,
1799
+ "learning_rate": 2.1613635589349756e-06,
1800
+ "loss": 0.9211,
1801
+ "step": 296
1802
+ },
1803
+ {
1804
+ "epoch": 16.73,
1805
+ "learning_rate": 2.0655968511759134e-06,
1806
+ "loss": 0.9145,
1807
+ "step": 297
1808
+ },
1809
+ {
1810
+ "epoch": 16.79,
1811
+ "learning_rate": 1.97190867212875e-06,
1812
+ "loss": 0.9481,
1813
+ "step": 298
1814
+ },
1815
+ {
1816
+ "epoch": 16.85,
1817
+ "learning_rate": 1.8803075126867715e-06,
1818
+ "loss": 0.9805,
1819
+ "step": 299
1820
+ },
1821
+ {
1822
+ "epoch": 16.9,
1823
+ "learning_rate": 1.790801674598186e-06,
1824
+ "loss": 0.8387,
1825
+ "step": 300
1826
+ },
1827
+ {
1828
+ "epoch": 16.9,
1829
+ "eval_loss": 1.3825798034667969,
1830
+ "eval_runtime": 1.4357,
1831
+ "eval_samples_per_second": 6.965,
1832
+ "eval_steps_per_second": 3.483,
1833
+ "step": 300
1834
+ },
1835
+ {
1836
+ "epoch": 16.96,
1837
+ "learning_rate": 1.703399269713693e-06,
1838
+ "loss": 1.005,
1839
+ "step": 301
1840
+ }
1841
+ ],
1842
+ "logging_steps": 1,
1843
+ "max_steps": 340,
1844
+ "num_train_epochs": 20,
1845
+ "save_steps": 500,
1846
+ "total_flos": 3.775247564931072e+17,
1847
+ "trial_name": null,
1848
+ "trial_params": null
1849
+ }
checkpoint-301/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af7bb882cac5b4d33f676aa38a935fcc9173c63dc3da717a69a225490ea5e0f7
3
+ size 4155
checkpoint-319/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0.dev0
checkpoint-319/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
4
+ "bias": "none",
5
+ "fan_in_fan_out": null,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 32,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "k_proj",
18
+ "gate_proj",
19
+ "v_proj",
20
+ "o_proj",
21
+ "q_proj",
22
+ "down_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
checkpoint-319/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ecd695972132ccf97b4e9a520f70d714fdfeaed1241c1d42fd7f5f347ae3f63
3
+ size 500897101
checkpoint-319/adapter_model/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0.dev0
checkpoint-319/adapter_model/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
4
+ "bias": "none",
5
+ "fan_in_fan_out": null,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 32,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "k_proj",
18
+ "gate_proj",
19
+ "v_proj",
20
+ "o_proj",
21
+ "q_proj",
22
+ "down_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
checkpoint-319/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ecd695972132ccf97b4e9a520f70d714fdfeaed1241c1d42fd7f5f347ae3f63
3
+ size 500897101
checkpoint-319/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6e425bfc563ced4cabe7ef81224acf2900c9c47cfd68d90956f91bfa68cc886b
3
+ size 1001737021
checkpoint-319/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1825a8ae85cbc524e0ec795578def9c0694d9915d32179190db8e409224f9651
3
+ size 14575
checkpoint-319/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:858293b1a43f90845c2289a0b15b110d176e35531d0c7d32285b12d3372bc510
3
+ size 627
checkpoint-319/trainer_state.json ADDED
@@ -0,0 +1,1957 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 17.971830985915492,
5
+ "eval_steps": 100,
6
+ "global_step": 319,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.06,
13
+ "learning_rate": 5e-06,
14
+ "loss": 1.5896,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.11,
19
+ "learning_rate": 1e-05,
20
+ "loss": 1.6521,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.17,
25
+ "learning_rate": 1.5e-05,
26
+ "loss": 1.3654,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.23,
31
+ "learning_rate": 2e-05,
32
+ "loss": 1.3569,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.28,
37
+ "learning_rate": 2.5e-05,
38
+ "loss": 1.4352,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.34,
43
+ "learning_rate": 3e-05,
44
+ "loss": 1.425,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.39,
49
+ "learning_rate": 3.5e-05,
50
+ "loss": 1.3529,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.45,
55
+ "learning_rate": 4e-05,
56
+ "loss": 1.384,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.51,
61
+ "learning_rate": 4.5e-05,
62
+ "loss": 1.3881,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.56,
67
+ "learning_rate": 5e-05,
68
+ "loss": 1.5593,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.62,
73
+ "learning_rate": 4.999886713385432e-05,
74
+ "loss": 1.3423,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.68,
79
+ "learning_rate": 4.999546863808815e-05,
80
+ "loss": 1.4725,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.73,
85
+ "learning_rate": 4.9989804820704735e-05,
86
+ "loss": 1.4342,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.79,
91
+ "learning_rate": 4.9981876195011844e-05,
92
+ "loss": 1.3769,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.85,
97
+ "learning_rate": 4.99716834795752e-05,
98
+ "loss": 1.3976,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.9,
103
+ "learning_rate": 4.995922759815339e-05,
104
+ "loss": 1.328,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.96,
109
+ "learning_rate": 4.994450967961413e-05,
110
+ "loss": 1.266,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 1.01,
115
+ "learning_rate": 4.992753105783194e-05,
116
+ "loss": 1.4671,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 1.07,
121
+ "learning_rate": 4.9908293271567286e-05,
122
+ "loss": 1.3973,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 1.13,
127
+ "learning_rate": 4.988679806432712e-05,
128
+ "loss": 1.3835,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 1.18,
133
+ "learning_rate": 4.9863047384206835e-05,
134
+ "loss": 1.2203,
135
+ "step": 21
136
+ },
137
+ {
138
+ "epoch": 1.24,
139
+ "learning_rate": 4.9837043383713753e-05,
140
+ "loss": 1.3752,
141
+ "step": 22
142
+ },
143
+ {
144
+ "epoch": 1.3,
145
+ "learning_rate": 4.980878841957203e-05,
146
+ "loss": 1.4142,
147
+ "step": 23
148
+ },
149
+ {
150
+ "epoch": 1.35,
151
+ "learning_rate": 4.977828505250903e-05,
152
+ "loss": 1.3657,
153
+ "step": 24
154
+ },
155
+ {
156
+ "epoch": 1.41,
157
+ "learning_rate": 4.9745536047023324e-05,
158
+ "loss": 1.309,
159
+ "step": 25
160
+ },
161
+ {
162
+ "epoch": 1.46,
163
+ "learning_rate": 4.971054437113406e-05,
164
+ "loss": 1.3004,
165
+ "step": 26
166
+ },
167
+ {
168
+ "epoch": 1.52,
169
+ "learning_rate": 4.967331319611206e-05,
170
+ "loss": 1.3309,
171
+ "step": 27
172
+ },
173
+ {
174
+ "epoch": 1.58,
175
+ "learning_rate": 4.963384589619233e-05,
176
+ "loss": 1.4956,
177
+ "step": 28
178
+ },
179
+ {
180
+ "epoch": 1.63,
181
+ "learning_rate": 4.959214604826831e-05,
182
+ "loss": 1.2682,
183
+ "step": 29
184
+ },
185
+ {
186
+ "epoch": 1.69,
187
+ "learning_rate": 4.9548217431567665e-05,
188
+ "loss": 1.391,
189
+ "step": 30
190
+ },
191
+ {
192
+ "epoch": 1.75,
193
+ "learning_rate": 4.9502064027309836e-05,
194
+ "loss": 1.3102,
195
+ "step": 31
196
+ },
197
+ {
198
+ "epoch": 1.8,
199
+ "learning_rate": 4.9453690018345144e-05,
200
+ "loss": 1.3891,
201
+ "step": 32
202
+ },
203
+ {
204
+ "epoch": 1.86,
205
+ "learning_rate": 4.9403099788775754e-05,
206
+ "loss": 1.3538,
207
+ "step": 33
208
+ },
209
+ {
210
+ "epoch": 1.92,
211
+ "learning_rate": 4.935029792355834e-05,
212
+ "loss": 1.3443,
213
+ "step": 34
214
+ },
215
+ {
216
+ "epoch": 1.97,
217
+ "learning_rate": 4.929528920808854e-05,
218
+ "loss": 1.3384,
219
+ "step": 35
220
+ },
221
+ {
222
+ "epoch": 2.03,
223
+ "learning_rate": 4.923807862776728e-05,
224
+ "loss": 1.2968,
225
+ "step": 36
226
+ },
227
+ {
228
+ "epoch": 2.08,
229
+ "learning_rate": 4.917867136754893e-05,
230
+ "loss": 1.2624,
231
+ "step": 37
232
+ },
233
+ {
234
+ "epoch": 2.14,
235
+ "learning_rate": 4.91170728114714e-05,
236
+ "loss": 1.3547,
237
+ "step": 38
238
+ },
239
+ {
240
+ "epoch": 2.2,
241
+ "learning_rate": 4.9053288542168185e-05,
242
+ "loss": 1.3857,
243
+ "step": 39
244
+ },
245
+ {
246
+ "epoch": 2.25,
247
+ "learning_rate": 4.898732434036244e-05,
248
+ "loss": 1.2261,
249
+ "step": 40
250
+ },
251
+ {
252
+ "epoch": 2.31,
253
+ "learning_rate": 4.8919186184343046e-05,
254
+ "loss": 1.4072,
255
+ "step": 41
256
+ },
257
+ {
258
+ "epoch": 2.37,
259
+ "learning_rate": 4.8848880249422815e-05,
260
+ "loss": 1.3283,
261
+ "step": 42
262
+ },
263
+ {
264
+ "epoch": 2.42,
265
+ "learning_rate": 4.877641290737884e-05,
266
+ "loss": 1.2464,
267
+ "step": 43
268
+ },
269
+ {
270
+ "epoch": 2.48,
271
+ "learning_rate": 4.870179072587499e-05,
272
+ "loss": 1.3433,
273
+ "step": 44
274
+ },
275
+ {
276
+ "epoch": 2.54,
277
+ "learning_rate": 4.862502046786671e-05,
278
+ "loss": 1.2861,
279
+ "step": 45
280
+ },
281
+ {
282
+ "epoch": 2.59,
283
+ "learning_rate": 4.854610909098812e-05,
284
+ "loss": 1.1888,
285
+ "step": 46
286
+ },
287
+ {
288
+ "epoch": 2.65,
289
+ "learning_rate": 4.8465063746921395e-05,
290
+ "loss": 1.2509,
291
+ "step": 47
292
+ },
293
+ {
294
+ "epoch": 2.7,
295
+ "learning_rate": 4.838189178074867e-05,
296
+ "loss": 1.4449,
297
+ "step": 48
298
+ },
299
+ {
300
+ "epoch": 2.76,
301
+ "learning_rate": 4.829660073028631e-05,
302
+ "loss": 1.4645,
303
+ "step": 49
304
+ },
305
+ {
306
+ "epoch": 2.82,
307
+ "learning_rate": 4.8209198325401815e-05,
308
+ "loss": 1.3586,
309
+ "step": 50
310
+ },
311
+ {
312
+ "epoch": 2.87,
313
+ "learning_rate": 4.811969248731323e-05,
314
+ "loss": 1.3542,
315
+ "step": 51
316
+ },
317
+ {
318
+ "epoch": 2.93,
319
+ "learning_rate": 4.802809132787125e-05,
320
+ "loss": 1.3145,
321
+ "step": 52
322
+ },
323
+ {
324
+ "epoch": 2.99,
325
+ "learning_rate": 4.793440314882408e-05,
326
+ "loss": 1.2343,
327
+ "step": 53
328
+ },
329
+ {
330
+ "epoch": 3.04,
331
+ "learning_rate": 4.783863644106502e-05,
332
+ "loss": 1.2946,
333
+ "step": 54
334
+ },
335
+ {
336
+ "epoch": 3.1,
337
+ "learning_rate": 4.774079988386296e-05,
338
+ "loss": 1.2621,
339
+ "step": 55
340
+ },
341
+ {
342
+ "epoch": 3.15,
343
+ "learning_rate": 4.764090234407577e-05,
344
+ "loss": 1.3581,
345
+ "step": 56
346
+ },
347
+ {
348
+ "epoch": 3.21,
349
+ "learning_rate": 4.753895287534673e-05,
350
+ "loss": 1.3891,
351
+ "step": 57
352
+ },
353
+ {
354
+ "epoch": 3.27,
355
+ "learning_rate": 4.743496071728396e-05,
356
+ "loss": 1.334,
357
+ "step": 58
358
+ },
359
+ {
360
+ "epoch": 3.32,
361
+ "learning_rate": 4.73289352946231e-05,
362
+ "loss": 1.3105,
363
+ "step": 59
364
+ },
365
+ {
366
+ "epoch": 3.38,
367
+ "learning_rate": 4.722088621637309e-05,
368
+ "loss": 1.2921,
369
+ "step": 60
370
+ },
371
+ {
372
+ "epoch": 3.44,
373
+ "learning_rate": 4.711082327494536e-05,
374
+ "loss": 1.2406,
375
+ "step": 61
376
+ },
377
+ {
378
+ "epoch": 3.49,
379
+ "learning_rate": 4.6998756445266336e-05,
380
+ "loss": 1.4246,
381
+ "step": 62
382
+ },
383
+ {
384
+ "epoch": 3.55,
385
+ "learning_rate": 4.688469588387339e-05,
386
+ "loss": 1.2126,
387
+ "step": 63
388
+ },
389
+ {
390
+ "epoch": 3.61,
391
+ "learning_rate": 4.6768651927994434e-05,
392
+ "loss": 1.3662,
393
+ "step": 64
394
+ },
395
+ {
396
+ "epoch": 3.66,
397
+ "learning_rate": 4.665063509461097e-05,
398
+ "loss": 1.2044,
399
+ "step": 65
400
+ },
401
+ {
402
+ "epoch": 3.72,
403
+ "learning_rate": 4.653065607950502e-05,
404
+ "loss": 1.4344,
405
+ "step": 66
406
+ },
407
+ {
408
+ "epoch": 3.77,
409
+ "learning_rate": 4.640872575628973e-05,
410
+ "loss": 1.3652,
411
+ "step": 67
412
+ },
413
+ {
414
+ "epoch": 3.83,
415
+ "learning_rate": 4.628485517542392e-05,
416
+ "loss": 1.231,
417
+ "step": 68
418
+ },
419
+ {
420
+ "epoch": 3.89,
421
+ "learning_rate": 4.6159055563210604e-05,
422
+ "loss": 1.1948,
423
+ "step": 69
424
+ },
425
+ {
426
+ "epoch": 3.94,
427
+ "learning_rate": 4.6031338320779534e-05,
428
+ "loss": 1.2484,
429
+ "step": 70
430
+ },
431
+ {
432
+ "epoch": 4.0,
433
+ "learning_rate": 4.59017150230539e-05,
434
+ "loss": 1.2968,
435
+ "step": 71
436
+ },
437
+ {
438
+ "epoch": 4.06,
439
+ "learning_rate": 4.5770197417701365e-05,
440
+ "loss": 1.1691,
441
+ "step": 72
442
+ },
443
+ {
444
+ "epoch": 4.11,
445
+ "learning_rate": 4.563679742406935e-05,
446
+ "loss": 1.224,
447
+ "step": 73
448
+ },
449
+ {
450
+ "epoch": 4.17,
451
+ "learning_rate": 4.550152713210478e-05,
452
+ "loss": 1.2669,
453
+ "step": 74
454
+ },
455
+ {
456
+ "epoch": 4.23,
457
+ "learning_rate": 4.5364398801258396e-05,
458
+ "loss": 1.3218,
459
+ "step": 75
460
+ },
461
+ {
462
+ "epoch": 4.28,
463
+ "learning_rate": 4.522542485937369e-05,
464
+ "loss": 1.3309,
465
+ "step": 76
466
+ },
467
+ {
468
+ "epoch": 4.34,
469
+ "learning_rate": 4.508461790156056e-05,
470
+ "loss": 1.3356,
471
+ "step": 77
472
+ },
473
+ {
474
+ "epoch": 4.39,
475
+ "learning_rate": 4.4941990689053886e-05,
476
+ "loss": 1.3367,
477
+ "step": 78
478
+ },
479
+ {
480
+ "epoch": 4.45,
481
+ "learning_rate": 4.479755614805688e-05,
482
+ "loss": 1.3568,
483
+ "step": 79
484
+ },
485
+ {
486
+ "epoch": 4.51,
487
+ "learning_rate": 4.465132736856969e-05,
488
+ "loss": 1.29,
489
+ "step": 80
490
+ },
491
+ {
492
+ "epoch": 4.56,
493
+ "learning_rate": 4.450331760320302e-05,
494
+ "loss": 1.3522,
495
+ "step": 81
496
+ },
497
+ {
498
+ "epoch": 4.62,
499
+ "learning_rate": 4.4353540265977064e-05,
500
+ "loss": 1.3021,
501
+ "step": 82
502
+ },
503
+ {
504
+ "epoch": 4.68,
505
+ "learning_rate": 4.4202008931105795e-05,
506
+ "loss": 1.303,
507
+ "step": 83
508
+ },
509
+ {
510
+ "epoch": 4.73,
511
+ "learning_rate": 4.404873733176678e-05,
512
+ "loss": 1.313,
513
+ "step": 84
514
+ },
515
+ {
516
+ "epoch": 4.79,
517
+ "learning_rate": 4.389373935885646e-05,
518
+ "loss": 1.231,
519
+ "step": 85
520
+ },
521
+ {
522
+ "epoch": 4.85,
523
+ "learning_rate": 4.373702905973135e-05,
524
+ "loss": 1.2908,
525
+ "step": 86
526
+ },
527
+ {
528
+ "epoch": 4.9,
529
+ "learning_rate": 4.357862063693486e-05,
530
+ "loss": 1.2575,
531
+ "step": 87
532
+ },
533
+ {
534
+ "epoch": 4.96,
535
+ "learning_rate": 4.341852844691012e-05,
536
+ "loss": 1.1942,
537
+ "step": 88
538
+ },
539
+ {
540
+ "epoch": 5.01,
541
+ "learning_rate": 4.3256766998698936e-05,
542
+ "loss": 1.1607,
543
+ "step": 89
544
+ },
545
+ {
546
+ "epoch": 5.07,
547
+ "learning_rate": 4.309335095262676e-05,
548
+ "loss": 1.3386,
549
+ "step": 90
550
+ },
551
+ {
552
+ "epoch": 5.13,
553
+ "learning_rate": 4.292829511897409e-05,
554
+ "loss": 1.371,
555
+ "step": 91
556
+ },
557
+ {
558
+ "epoch": 5.18,
559
+ "learning_rate": 4.276161445663423e-05,
560
+ "loss": 1.236,
561
+ "step": 92
562
+ },
563
+ {
564
+ "epoch": 5.24,
565
+ "learning_rate": 4.259332407175751e-05,
566
+ "loss": 1.2531,
567
+ "step": 93
568
+ },
569
+ {
570
+ "epoch": 5.3,
571
+ "learning_rate": 4.242343921638234e-05,
572
+ "loss": 1.2593,
573
+ "step": 94
574
+ },
575
+ {
576
+ "epoch": 5.35,
577
+ "learning_rate": 4.2251975287052804e-05,
578
+ "loss": 1.1245,
579
+ "step": 95
580
+ },
581
+ {
582
+ "epoch": 5.41,
583
+ "learning_rate": 4.2078947823423364e-05,
584
+ "loss": 1.2538,
585
+ "step": 96
586
+ },
587
+ {
588
+ "epoch": 5.46,
589
+ "learning_rate": 4.1904372506850484e-05,
590
+ "loss": 1.2847,
591
+ "step": 97
592
+ },
593
+ {
594
+ "epoch": 5.52,
595
+ "learning_rate": 4.172826515897146e-05,
596
+ "loss": 1.2071,
597
+ "step": 98
598
+ },
599
+ {
600
+ "epoch": 5.58,
601
+ "learning_rate": 4.155064174027047e-05,
602
+ "loss": 1.2664,
603
+ "step": 99
604
+ },
605
+ {
606
+ "epoch": 5.63,
607
+ "learning_rate": 4.137151834863213e-05,
608
+ "loss": 1.1919,
609
+ "step": 100
610
+ },
611
+ {
612
+ "epoch": 5.63,
613
+ "eval_loss": 1.2261024713516235,
614
+ "eval_runtime": 1.4353,
615
+ "eval_samples_per_second": 6.967,
616
+ "eval_steps_per_second": 3.484,
617
+ "step": 100
618
+ },
619
+ {
620
+ "epoch": 5.69,
621
+ "learning_rate": 4.119091121788256e-05,
622
+ "loss": 1.1509,
623
+ "step": 101
624
+ },
625
+ {
626
+ "epoch": 5.75,
627
+ "learning_rate": 4.100883671631806e-05,
628
+ "loss": 1.3013,
629
+ "step": 102
630
+ },
631
+ {
632
+ "epoch": 5.8,
633
+ "learning_rate": 4.082531134522176e-05,
634
+ "loss": 1.313,
635
+ "step": 103
636
+ },
637
+ {
638
+ "epoch": 5.86,
639
+ "learning_rate": 4.064035173736804e-05,
640
+ "loss": 1.2176,
641
+ "step": 104
642
+ },
643
+ {
644
+ "epoch": 5.92,
645
+ "learning_rate": 4.045397465551513e-05,
646
+ "loss": 1.1004,
647
+ "step": 105
648
+ },
649
+ {
650
+ "epoch": 5.97,
651
+ "learning_rate": 4.0266196990885955e-05,
652
+ "loss": 1.3575,
653
+ "step": 106
654
+ },
655
+ {
656
+ "epoch": 6.03,
657
+ "learning_rate": 4.007703576163724e-05,
658
+ "loss": 1.2923,
659
+ "step": 107
660
+ },
661
+ {
662
+ "epoch": 6.08,
663
+ "learning_rate": 3.98865081113172e-05,
664
+ "loss": 1.2083,
665
+ "step": 108
666
+ },
667
+ {
668
+ "epoch": 6.14,
669
+ "learning_rate": 3.969463130731183e-05,
670
+ "loss": 1.2141,
671
+ "step": 109
672
+ },
673
+ {
674
+ "epoch": 6.2,
675
+ "learning_rate": 3.9501422739279956e-05,
676
+ "loss": 1.3297,
677
+ "step": 110
678
+ },
679
+ {
680
+ "epoch": 6.25,
681
+ "learning_rate": 3.9306899917577245e-05,
682
+ "loss": 1.0901,
683
+ "step": 111
684
+ },
685
+ {
686
+ "epoch": 6.31,
687
+ "learning_rate": 3.911108047166924e-05,
688
+ "loss": 1.2235,
689
+ "step": 112
690
+ },
691
+ {
692
+ "epoch": 6.37,
693
+ "learning_rate": 3.89139821485336e-05,
694
+ "loss": 1.341,
695
+ "step": 113
696
+ },
697
+ {
698
+ "epoch": 6.42,
699
+ "learning_rate": 3.871562281105175e-05,
700
+ "loss": 1.362,
701
+ "step": 114
702
+ },
703
+ {
704
+ "epoch": 6.48,
705
+ "learning_rate": 3.851602043638994e-05,
706
+ "loss": 1.2738,
707
+ "step": 115
708
+ },
709
+ {
710
+ "epoch": 6.54,
711
+ "learning_rate": 3.8315193114369996e-05,
712
+ "loss": 1.2264,
713
+ "step": 116
714
+ },
715
+ {
716
+ "epoch": 6.59,
717
+ "learning_rate": 3.8113159045829864e-05,
718
+ "loss": 1.171,
719
+ "step": 117
720
+ },
721
+ {
722
+ "epoch": 6.65,
723
+ "learning_rate": 3.790993654097405e-05,
724
+ "loss": 1.1848,
725
+ "step": 118
726
+ },
727
+ {
728
+ "epoch": 6.7,
729
+ "learning_rate": 3.770554401771423e-05,
730
+ "loss": 1.1545,
731
+ "step": 119
732
+ },
733
+ {
734
+ "epoch": 6.76,
735
+ "learning_rate": 3.7500000000000003e-05,
736
+ "loss": 1.2799,
737
+ "step": 120
738
+ },
739
+ {
740
+ "epoch": 6.82,
741
+ "learning_rate": 3.72933231161401e-05,
742
+ "loss": 1.2278,
743
+ "step": 121
744
+ },
745
+ {
746
+ "epoch": 6.87,
747
+ "learning_rate": 3.708553209711409e-05,
748
+ "loss": 1.1959,
749
+ "step": 122
750
+ },
751
+ {
752
+ "epoch": 6.93,
753
+ "learning_rate": 3.6876645774874876e-05,
754
+ "loss": 1.1151,
755
+ "step": 123
756
+ },
757
+ {
758
+ "epoch": 6.99,
759
+ "learning_rate": 3.6666683080641846e-05,
760
+ "loss": 1.2178,
761
+ "step": 124
762
+ },
763
+ {
764
+ "epoch": 7.04,
765
+ "learning_rate": 3.645566304318526e-05,
766
+ "loss": 1.1584,
767
+ "step": 125
768
+ },
769
+ {
770
+ "epoch": 7.1,
771
+ "learning_rate": 3.624360478710165e-05,
772
+ "loss": 1.0959,
773
+ "step": 126
774
+ },
775
+ {
776
+ "epoch": 7.15,
777
+ "learning_rate": 3.603052753108053e-05,
778
+ "loss": 1.3026,
779
+ "step": 127
780
+ },
781
+ {
782
+ "epoch": 7.21,
783
+ "learning_rate": 3.581645058616271e-05,
784
+ "loss": 1.2007,
785
+ "step": 128
786
+ },
787
+ {
788
+ "epoch": 7.27,
789
+ "learning_rate": 3.5601393353990046e-05,
790
+ "loss": 1.2242,
791
+ "step": 129
792
+ },
793
+ {
794
+ "epoch": 7.32,
795
+ "learning_rate": 3.5385375325047166e-05,
796
+ "loss": 1.2737,
797
+ "step": 130
798
+ },
799
+ {
800
+ "epoch": 7.38,
801
+ "learning_rate": 3.516841607689501e-05,
802
+ "loss": 1.2724,
803
+ "step": 131
804
+ },
805
+ {
806
+ "epoch": 7.44,
807
+ "learning_rate": 3.495053527239656e-05,
808
+ "loss": 1.1916,
809
+ "step": 132
810
+ },
811
+ {
812
+ "epoch": 7.49,
813
+ "learning_rate": 3.4731752657934794e-05,
814
+ "loss": 1.2107,
815
+ "step": 133
816
+ },
817
+ {
818
+ "epoch": 7.55,
819
+ "learning_rate": 3.4512088061623075e-05,
820
+ "loss": 1.246,
821
+ "step": 134
822
+ },
823
+ {
824
+ "epoch": 7.61,
825
+ "learning_rate": 3.4291561391508185e-05,
826
+ "loss": 1.1841,
827
+ "step": 135
828
+ },
829
+ {
830
+ "epoch": 7.66,
831
+ "learning_rate": 3.4070192633766025e-05,
832
+ "loss": 1.1898,
833
+ "step": 136
834
+ },
835
+ {
836
+ "epoch": 7.72,
837
+ "learning_rate": 3.3848001850890296e-05,
838
+ "loss": 1.1195,
839
+ "step": 137
840
+ },
841
+ {
842
+ "epoch": 7.77,
843
+ "learning_rate": 3.3625009179874265e-05,
844
+ "loss": 1.1685,
845
+ "step": 138
846
+ },
847
+ {
848
+ "epoch": 7.83,
849
+ "learning_rate": 3.3401234830385756e-05,
850
+ "loss": 1.2748,
851
+ "step": 139
852
+ },
853
+ {
854
+ "epoch": 7.89,
855
+ "learning_rate": 3.3176699082935545e-05,
856
+ "loss": 1.1697,
857
+ "step": 140
858
+ },
859
+ {
860
+ "epoch": 7.94,
861
+ "learning_rate": 3.295142228703938e-05,
862
+ "loss": 1.1624,
863
+ "step": 141
864
+ },
865
+ {
866
+ "epoch": 8.0,
867
+ "learning_rate": 3.272542485937369e-05,
868
+ "loss": 1.1679,
869
+ "step": 142
870
+ },
871
+ {
872
+ "epoch": 8.06,
873
+ "learning_rate": 3.249872728192527e-05,
874
+ "loss": 1.1163,
875
+ "step": 143
876
+ },
877
+ {
878
+ "epoch": 8.11,
879
+ "learning_rate": 3.2271350100134975e-05,
880
+ "loss": 1.1319,
881
+ "step": 144
882
+ },
883
+ {
884
+ "epoch": 8.17,
885
+ "learning_rate": 3.2043313921035743e-05,
886
+ "loss": 1.2151,
887
+ "step": 145
888
+ },
889
+ {
890
+ "epoch": 8.23,
891
+ "learning_rate": 3.181463941138495e-05,
892
+ "loss": 1.1339,
893
+ "step": 146
894
+ },
895
+ {
896
+ "epoch": 8.28,
897
+ "learning_rate": 3.158534729579142e-05,
898
+ "loss": 1.1503,
899
+ "step": 147
900
+ },
901
+ {
902
+ "epoch": 8.34,
903
+ "learning_rate": 3.135545835483718e-05,
904
+ "loss": 1.2326,
905
+ "step": 148
906
+ },
907
+ {
908
+ "epoch": 8.39,
909
+ "learning_rate": 3.11249934231941e-05,
910
+ "loss": 1.2938,
911
+ "step": 149
912
+ },
913
+ {
914
+ "epoch": 8.45,
915
+ "learning_rate": 3.0893973387735687e-05,
916
+ "loss": 1.1454,
917
+ "step": 150
918
+ },
919
+ {
920
+ "epoch": 8.51,
921
+ "learning_rate": 3.0662419185644115e-05,
922
+ "loss": 1.2504,
923
+ "step": 151
924
+ },
925
+ {
926
+ "epoch": 8.56,
927
+ "learning_rate": 3.0430351802512698e-05,
928
+ "loss": 0.987,
929
+ "step": 152
930
+ },
931
+ {
932
+ "epoch": 8.62,
933
+ "learning_rate": 3.0197792270443982e-05,
934
+ "loss": 1.1387,
935
+ "step": 153
936
+ },
937
+ {
938
+ "epoch": 8.68,
939
+ "learning_rate": 2.996476166614364e-05,
940
+ "loss": 1.1142,
941
+ "step": 154
942
+ },
943
+ {
944
+ "epoch": 8.73,
945
+ "learning_rate": 2.9731281109010256e-05,
946
+ "loss": 1.1714,
947
+ "step": 155
948
+ },
949
+ {
950
+ "epoch": 8.79,
951
+ "learning_rate": 2.9497371759221347e-05,
952
+ "loss": 1.1858,
953
+ "step": 156
954
+ },
955
+ {
956
+ "epoch": 8.85,
957
+ "learning_rate": 2.92630548158156e-05,
958
+ "loss": 1.2169,
959
+ "step": 157
960
+ },
961
+ {
962
+ "epoch": 8.9,
963
+ "learning_rate": 2.9028351514771606e-05,
964
+ "loss": 1.2204,
965
+ "step": 158
966
+ },
967
+ {
968
+ "epoch": 8.96,
969
+ "learning_rate": 2.8793283127083292e-05,
970
+ "loss": 1.0313,
971
+ "step": 159
972
+ },
973
+ {
974
+ "epoch": 9.01,
975
+ "learning_rate": 2.8557870956832132e-05,
976
+ "loss": 1.2432,
977
+ "step": 160
978
+ },
979
+ {
980
+ "epoch": 9.07,
981
+ "learning_rate": 2.8322136339256356e-05,
982
+ "loss": 1.0614,
983
+ "step": 161
984
+ },
985
+ {
986
+ "epoch": 9.13,
987
+ "learning_rate": 2.808610063881737e-05,
988
+ "loss": 1.1682,
989
+ "step": 162
990
+ },
991
+ {
992
+ "epoch": 9.18,
993
+ "learning_rate": 2.7849785247263515e-05,
994
+ "loss": 1.0238,
995
+ "step": 163
996
+ },
997
+ {
998
+ "epoch": 9.24,
999
+ "learning_rate": 2.761321158169134e-05,
1000
+ "loss": 1.166,
1001
+ "step": 164
1002
+ },
1003
+ {
1004
+ "epoch": 9.3,
1005
+ "learning_rate": 2.7376401082604564e-05,
1006
+ "loss": 1.0965,
1007
+ "step": 165
1008
+ },
1009
+ {
1010
+ "epoch": 9.35,
1011
+ "learning_rate": 2.7139375211970996e-05,
1012
+ "loss": 1.096,
1013
+ "step": 166
1014
+ },
1015
+ {
1016
+ "epoch": 9.41,
1017
+ "learning_rate": 2.6902155451277377e-05,
1018
+ "loss": 1.1161,
1019
+ "step": 167
1020
+ },
1021
+ {
1022
+ "epoch": 9.46,
1023
+ "learning_rate": 2.6664763299582602e-05,
1024
+ "loss": 1.1328,
1025
+ "step": 168
1026
+ },
1027
+ {
1028
+ "epoch": 9.52,
1029
+ "learning_rate": 2.6427220271569203e-05,
1030
+ "loss": 1.0747,
1031
+ "step": 169
1032
+ },
1033
+ {
1034
+ "epoch": 9.58,
1035
+ "learning_rate": 2.6189547895593562e-05,
1036
+ "loss": 1.1092,
1037
+ "step": 170
1038
+ },
1039
+ {
1040
+ "epoch": 9.63,
1041
+ "learning_rate": 2.5951767711734753e-05,
1042
+ "loss": 1.1624,
1043
+ "step": 171
1044
+ },
1045
+ {
1046
+ "epoch": 9.69,
1047
+ "learning_rate": 2.5713901269842404e-05,
1048
+ "loss": 1.1434,
1049
+ "step": 172
1050
+ },
1051
+ {
1052
+ "epoch": 9.75,
1053
+ "learning_rate": 2.5475970127583666e-05,
1054
+ "loss": 1.1167,
1055
+ "step": 173
1056
+ },
1057
+ {
1058
+ "epoch": 9.8,
1059
+ "learning_rate": 2.523799584848942e-05,
1060
+ "loss": 1.0509,
1061
+ "step": 174
1062
+ },
1063
+ {
1064
+ "epoch": 9.86,
1065
+ "learning_rate": 2.5e-05,
1066
+ "loss": 1.1449,
1067
+ "step": 175
1068
+ },
1069
+ {
1070
+ "epoch": 9.92,
1071
+ "learning_rate": 2.4762004151510584e-05,
1072
+ "loss": 1.109,
1073
+ "step": 176
1074
+ },
1075
+ {
1076
+ "epoch": 9.97,
1077
+ "learning_rate": 2.4524029872416333e-05,
1078
+ "loss": 1.0272,
1079
+ "step": 177
1080
+ },
1081
+ {
1082
+ "epoch": 10.03,
1083
+ "learning_rate": 2.42860987301576e-05,
1084
+ "loss": 1.1489,
1085
+ "step": 178
1086
+ },
1087
+ {
1088
+ "epoch": 10.08,
1089
+ "learning_rate": 2.4048232288265253e-05,
1090
+ "loss": 1.0263,
1091
+ "step": 179
1092
+ },
1093
+ {
1094
+ "epoch": 10.14,
1095
+ "learning_rate": 2.3810452104406444e-05,
1096
+ "loss": 1.0305,
1097
+ "step": 180
1098
+ },
1099
+ {
1100
+ "epoch": 10.2,
1101
+ "learning_rate": 2.35727797284308e-05,
1102
+ "loss": 0.9612,
1103
+ "step": 181
1104
+ },
1105
+ {
1106
+ "epoch": 10.25,
1107
+ "learning_rate": 2.3335236700417404e-05,
1108
+ "loss": 1.1883,
1109
+ "step": 182
1110
+ },
1111
+ {
1112
+ "epoch": 10.31,
1113
+ "learning_rate": 2.309784454872262e-05,
1114
+ "loss": 1.192,
1115
+ "step": 183
1116
+ },
1117
+ {
1118
+ "epoch": 10.37,
1119
+ "learning_rate": 2.2860624788029013e-05,
1120
+ "loss": 1.0919,
1121
+ "step": 184
1122
+ },
1123
+ {
1124
+ "epoch": 10.42,
1125
+ "learning_rate": 2.2623598917395438e-05,
1126
+ "loss": 1.0974,
1127
+ "step": 185
1128
+ },
1129
+ {
1130
+ "epoch": 10.48,
1131
+ "learning_rate": 2.238678841830867e-05,
1132
+ "loss": 1.0938,
1133
+ "step": 186
1134
+ },
1135
+ {
1136
+ "epoch": 10.54,
1137
+ "learning_rate": 2.2150214752736488e-05,
1138
+ "loss": 0.9879,
1139
+ "step": 187
1140
+ },
1141
+ {
1142
+ "epoch": 10.59,
1143
+ "learning_rate": 2.1913899361182632e-05,
1144
+ "loss": 1.0284,
1145
+ "step": 188
1146
+ },
1147
+ {
1148
+ "epoch": 10.65,
1149
+ "learning_rate": 2.167786366074365e-05,
1150
+ "loss": 1.0184,
1151
+ "step": 189
1152
+ },
1153
+ {
1154
+ "epoch": 10.7,
1155
+ "learning_rate": 2.1442129043167874e-05,
1156
+ "loss": 1.0419,
1157
+ "step": 190
1158
+ },
1159
+ {
1160
+ "epoch": 10.76,
1161
+ "learning_rate": 2.1206716872916714e-05,
1162
+ "loss": 1.0631,
1163
+ "step": 191
1164
+ },
1165
+ {
1166
+ "epoch": 10.82,
1167
+ "learning_rate": 2.09716484852284e-05,
1168
+ "loss": 1.0492,
1169
+ "step": 192
1170
+ },
1171
+ {
1172
+ "epoch": 10.87,
1173
+ "learning_rate": 2.0736945184184405e-05,
1174
+ "loss": 1.0924,
1175
+ "step": 193
1176
+ },
1177
+ {
1178
+ "epoch": 10.93,
1179
+ "learning_rate": 2.0502628240778655e-05,
1180
+ "loss": 0.9832,
1181
+ "step": 194
1182
+ },
1183
+ {
1184
+ "epoch": 10.99,
1185
+ "learning_rate": 2.0268718890989753e-05,
1186
+ "loss": 1.1266,
1187
+ "step": 195
1188
+ },
1189
+ {
1190
+ "epoch": 11.04,
1191
+ "learning_rate": 2.003523833385637e-05,
1192
+ "loss": 1.0762,
1193
+ "step": 196
1194
+ },
1195
+ {
1196
+ "epoch": 11.1,
1197
+ "learning_rate": 1.980220772955602e-05,
1198
+ "loss": 1.0011,
1199
+ "step": 197
1200
+ },
1201
+ {
1202
+ "epoch": 11.15,
1203
+ "learning_rate": 1.956964819748731e-05,
1204
+ "loss": 1.0195,
1205
+ "step": 198
1206
+ },
1207
+ {
1208
+ "epoch": 11.21,
1209
+ "learning_rate": 1.9337580814355888e-05,
1210
+ "loss": 1.0442,
1211
+ "step": 199
1212
+ },
1213
+ {
1214
+ "epoch": 11.27,
1215
+ "learning_rate": 1.9106026612264316e-05,
1216
+ "loss": 1.0285,
1217
+ "step": 200
1218
+ },
1219
+ {
1220
+ "epoch": 11.27,
1221
+ "eval_loss": 1.3292417526245117,
1222
+ "eval_runtime": 1.4356,
1223
+ "eval_samples_per_second": 6.966,
1224
+ "eval_steps_per_second": 3.483,
1225
+ "step": 200
1226
+ },
1227
+ {
1228
+ "epoch": 11.32,
1229
+ "learning_rate": 1.8875006576805914e-05,
1230
+ "loss": 0.9805,
1231
+ "step": 201
1232
+ },
1233
+ {
1234
+ "epoch": 11.38,
1235
+ "learning_rate": 1.8644541645162834e-05,
1236
+ "loss": 0.9562,
1237
+ "step": 202
1238
+ },
1239
+ {
1240
+ "epoch": 11.44,
1241
+ "learning_rate": 1.8414652704208583e-05,
1242
+ "loss": 1.0457,
1243
+ "step": 203
1244
+ },
1245
+ {
1246
+ "epoch": 11.49,
1247
+ "learning_rate": 1.8185360588615058e-05,
1248
+ "loss": 1.0217,
1249
+ "step": 204
1250
+ },
1251
+ {
1252
+ "epoch": 11.55,
1253
+ "learning_rate": 1.795668607896426e-05,
1254
+ "loss": 1.1281,
1255
+ "step": 205
1256
+ },
1257
+ {
1258
+ "epoch": 11.61,
1259
+ "learning_rate": 1.7728649899865024e-05,
1260
+ "loss": 0.9654,
1261
+ "step": 206
1262
+ },
1263
+ {
1264
+ "epoch": 11.66,
1265
+ "learning_rate": 1.7501272718074737e-05,
1266
+ "loss": 0.9011,
1267
+ "step": 207
1268
+ },
1269
+ {
1270
+ "epoch": 11.72,
1271
+ "learning_rate": 1.7274575140626318e-05,
1272
+ "loss": 1.097,
1273
+ "step": 208
1274
+ },
1275
+ {
1276
+ "epoch": 11.77,
1277
+ "learning_rate": 1.7048577712960627e-05,
1278
+ "loss": 1.0652,
1279
+ "step": 209
1280
+ },
1281
+ {
1282
+ "epoch": 11.83,
1283
+ "learning_rate": 1.682330091706446e-05,
1284
+ "loss": 1.011,
1285
+ "step": 210
1286
+ },
1287
+ {
1288
+ "epoch": 11.89,
1289
+ "learning_rate": 1.6598765169614243e-05,
1290
+ "loss": 1.1001,
1291
+ "step": 211
1292
+ },
1293
+ {
1294
+ "epoch": 11.94,
1295
+ "learning_rate": 1.637499082012574e-05,
1296
+ "loss": 1.1224,
1297
+ "step": 212
1298
+ },
1299
+ {
1300
+ "epoch": 12.0,
1301
+ "learning_rate": 1.615199814910971e-05,
1302
+ "loss": 0.967,
1303
+ "step": 213
1304
+ },
1305
+ {
1306
+ "epoch": 12.06,
1307
+ "learning_rate": 1.5929807366233977e-05,
1308
+ "loss": 0.9923,
1309
+ "step": 214
1310
+ },
1311
+ {
1312
+ "epoch": 12.11,
1313
+ "learning_rate": 1.5708438608491814e-05,
1314
+ "loss": 1.0722,
1315
+ "step": 215
1316
+ },
1317
+ {
1318
+ "epoch": 12.17,
1319
+ "learning_rate": 1.5487911938376924e-05,
1320
+ "loss": 0.9666,
1321
+ "step": 216
1322
+ },
1323
+ {
1324
+ "epoch": 12.23,
1325
+ "learning_rate": 1.5268247342065215e-05,
1326
+ "loss": 0.9136,
1327
+ "step": 217
1328
+ },
1329
+ {
1330
+ "epoch": 12.28,
1331
+ "learning_rate": 1.5049464727603452e-05,
1332
+ "loss": 0.9274,
1333
+ "step": 218
1334
+ },
1335
+ {
1336
+ "epoch": 12.34,
1337
+ "learning_rate": 1.4831583923104999e-05,
1338
+ "loss": 1.0955,
1339
+ "step": 219
1340
+ },
1341
+ {
1342
+ "epoch": 12.39,
1343
+ "learning_rate": 1.4614624674952842e-05,
1344
+ "loss": 0.9058,
1345
+ "step": 220
1346
+ },
1347
+ {
1348
+ "epoch": 12.45,
1349
+ "learning_rate": 1.4398606646009963e-05,
1350
+ "loss": 0.8351,
1351
+ "step": 221
1352
+ },
1353
+ {
1354
+ "epoch": 12.51,
1355
+ "learning_rate": 1.4183549413837289e-05,
1356
+ "loss": 1.119,
1357
+ "step": 222
1358
+ },
1359
+ {
1360
+ "epoch": 12.56,
1361
+ "learning_rate": 1.3969472468919461e-05,
1362
+ "loss": 1.0418,
1363
+ "step": 223
1364
+ },
1365
+ {
1366
+ "epoch": 12.62,
1367
+ "learning_rate": 1.3756395212898359e-05,
1368
+ "loss": 0.9102,
1369
+ "step": 224
1370
+ },
1371
+ {
1372
+ "epoch": 12.68,
1373
+ "learning_rate": 1.354433695681474e-05,
1374
+ "loss": 0.997,
1375
+ "step": 225
1376
+ },
1377
+ {
1378
+ "epoch": 12.73,
1379
+ "learning_rate": 1.3333316919358157e-05,
1380
+ "loss": 1.1737,
1381
+ "step": 226
1382
+ },
1383
+ {
1384
+ "epoch": 12.79,
1385
+ "learning_rate": 1.3123354225125128e-05,
1386
+ "loss": 0.9887,
1387
+ "step": 227
1388
+ },
1389
+ {
1390
+ "epoch": 12.85,
1391
+ "learning_rate": 1.2914467902885901e-05,
1392
+ "loss": 0.9984,
1393
+ "step": 228
1394
+ },
1395
+ {
1396
+ "epoch": 12.9,
1397
+ "learning_rate": 1.2706676883859903e-05,
1398
+ "loss": 0.9507,
1399
+ "step": 229
1400
+ },
1401
+ {
1402
+ "epoch": 12.96,
1403
+ "learning_rate": 1.2500000000000006e-05,
1404
+ "loss": 0.9395,
1405
+ "step": 230
1406
+ },
1407
+ {
1408
+ "epoch": 13.01,
1409
+ "learning_rate": 1.2294455982285777e-05,
1410
+ "loss": 1.1445,
1411
+ "step": 231
1412
+ },
1413
+ {
1414
+ "epoch": 13.07,
1415
+ "learning_rate": 1.2090063459025955e-05,
1416
+ "loss": 0.8469,
1417
+ "step": 232
1418
+ },
1419
+ {
1420
+ "epoch": 13.13,
1421
+ "learning_rate": 1.1886840954170142e-05,
1422
+ "loss": 0.9936,
1423
+ "step": 233
1424
+ },
1425
+ {
1426
+ "epoch": 13.18,
1427
+ "learning_rate": 1.1684806885630004e-05,
1428
+ "loss": 0.8797,
1429
+ "step": 234
1430
+ },
1431
+ {
1432
+ "epoch": 13.24,
1433
+ "learning_rate": 1.148397956361007e-05,
1434
+ "loss": 1.1331,
1435
+ "step": 235
1436
+ },
1437
+ {
1438
+ "epoch": 13.3,
1439
+ "learning_rate": 1.1284377188948258e-05,
1440
+ "loss": 0.9613,
1441
+ "step": 236
1442
+ },
1443
+ {
1444
+ "epoch": 13.35,
1445
+ "learning_rate": 1.1086017851466413e-05,
1446
+ "loss": 0.9343,
1447
+ "step": 237
1448
+ },
1449
+ {
1450
+ "epoch": 13.41,
1451
+ "learning_rate": 1.0888919528330777e-05,
1452
+ "loss": 1.0098,
1453
+ "step": 238
1454
+ },
1455
+ {
1456
+ "epoch": 13.46,
1457
+ "learning_rate": 1.0693100082422763e-05,
1458
+ "loss": 0.9669,
1459
+ "step": 239
1460
+ },
1461
+ {
1462
+ "epoch": 13.52,
1463
+ "learning_rate": 1.049857726072005e-05,
1464
+ "loss": 0.9686,
1465
+ "step": 240
1466
+ },
1467
+ {
1468
+ "epoch": 13.58,
1469
+ "learning_rate": 1.0305368692688174e-05,
1470
+ "loss": 0.8916,
1471
+ "step": 241
1472
+ },
1473
+ {
1474
+ "epoch": 13.63,
1475
+ "learning_rate": 1.0113491888682802e-05,
1476
+ "loss": 0.9274,
1477
+ "step": 242
1478
+ },
1479
+ {
1480
+ "epoch": 13.69,
1481
+ "learning_rate": 9.922964238362762e-06,
1482
+ "loss": 0.9307,
1483
+ "step": 243
1484
+ },
1485
+ {
1486
+ "epoch": 13.75,
1487
+ "learning_rate": 9.733803009114045e-06,
1488
+ "loss": 1.0112,
1489
+ "step": 244
1490
+ },
1491
+ {
1492
+ "epoch": 13.8,
1493
+ "learning_rate": 9.546025344484869e-06,
1494
+ "loss": 1.0465,
1495
+ "step": 245
1496
+ },
1497
+ {
1498
+ "epoch": 13.86,
1499
+ "learning_rate": 9.359648262631961e-06,
1500
+ "loss": 1.0822,
1501
+ "step": 246
1502
+ },
1503
+ {
1504
+ "epoch": 13.92,
1505
+ "learning_rate": 9.174688654778243e-06,
1506
+ "loss": 0.8992,
1507
+ "step": 247
1508
+ },
1509
+ {
1510
+ "epoch": 13.97,
1511
+ "learning_rate": 8.991163283681944e-06,
1512
+ "loss": 0.8778,
1513
+ "step": 248
1514
+ },
1515
+ {
1516
+ "epoch": 14.03,
1517
+ "learning_rate": 8.809088782117452e-06,
1518
+ "loss": 1.0079,
1519
+ "step": 249
1520
+ },
1521
+ {
1522
+ "epoch": 14.08,
1523
+ "learning_rate": 8.628481651367876e-06,
1524
+ "loss": 0.9508,
1525
+ "step": 250
1526
+ },
1527
+ {
1528
+ "epoch": 14.14,
1529
+ "learning_rate": 8.449358259729537e-06,
1530
+ "loss": 0.997,
1531
+ "step": 251
1532
+ },
1533
+ {
1534
+ "epoch": 14.2,
1535
+ "learning_rate": 8.271734841028553e-06,
1536
+ "loss": 0.8669,
1537
+ "step": 252
1538
+ },
1539
+ {
1540
+ "epoch": 14.25,
1541
+ "learning_rate": 8.09562749314952e-06,
1542
+ "loss": 0.9721,
1543
+ "step": 253
1544
+ },
1545
+ {
1546
+ "epoch": 14.31,
1547
+ "learning_rate": 7.921052176576644e-06,
1548
+ "loss": 1.0777,
1549
+ "step": 254
1550
+ },
1551
+ {
1552
+ "epoch": 14.37,
1553
+ "learning_rate": 7.748024712947205e-06,
1554
+ "loss": 0.8644,
1555
+ "step": 255
1556
+ },
1557
+ {
1558
+ "epoch": 14.42,
1559
+ "learning_rate": 7.576560783617668e-06,
1560
+ "loss": 0.9119,
1561
+ "step": 256
1562
+ },
1563
+ {
1564
+ "epoch": 14.48,
1565
+ "learning_rate": 7.40667592824249e-06,
1566
+ "loss": 0.8645,
1567
+ "step": 257
1568
+ },
1569
+ {
1570
+ "epoch": 14.54,
1571
+ "learning_rate": 7.238385543365783e-06,
1572
+ "loss": 0.8786,
1573
+ "step": 258
1574
+ },
1575
+ {
1576
+ "epoch": 14.59,
1577
+ "learning_rate": 7.071704881025915e-06,
1578
+ "loss": 0.9539,
1579
+ "step": 259
1580
+ },
1581
+ {
1582
+ "epoch": 14.65,
1583
+ "learning_rate": 6.906649047373246e-06,
1584
+ "loss": 0.9521,
1585
+ "step": 260
1586
+ },
1587
+ {
1588
+ "epoch": 14.7,
1589
+ "learning_rate": 6.743233001301067e-06,
1590
+ "loss": 1.0708,
1591
+ "step": 261
1592
+ },
1593
+ {
1594
+ "epoch": 14.76,
1595
+ "learning_rate": 6.5814715530898745e-06,
1596
+ "loss": 1.033,
1597
+ "step": 262
1598
+ },
1599
+ {
1600
+ "epoch": 14.82,
1601
+ "learning_rate": 6.421379363065142e-06,
1602
+ "loss": 0.9344,
1603
+ "step": 263
1604
+ },
1605
+ {
1606
+ "epoch": 14.87,
1607
+ "learning_rate": 6.2629709402686535e-06,
1608
+ "loss": 0.9491,
1609
+ "step": 264
1610
+ },
1611
+ {
1612
+ "epoch": 14.93,
1613
+ "learning_rate": 6.106260641143546e-06,
1614
+ "loss": 0.8894,
1615
+ "step": 265
1616
+ },
1617
+ {
1618
+ "epoch": 14.99,
1619
+ "learning_rate": 5.951262668233232e-06,
1620
+ "loss": 0.8825,
1621
+ "step": 266
1622
+ },
1623
+ {
1624
+ "epoch": 15.04,
1625
+ "learning_rate": 5.797991068894201e-06,
1626
+ "loss": 0.925,
1627
+ "step": 267
1628
+ },
1629
+ {
1630
+ "epoch": 15.1,
1631
+ "learning_rate": 5.646459734022938e-06,
1632
+ "loss": 0.8431,
1633
+ "step": 268
1634
+ },
1635
+ {
1636
+ "epoch": 15.15,
1637
+ "learning_rate": 5.49668239679699e-06,
1638
+ "loss": 1.036,
1639
+ "step": 269
1640
+ },
1641
+ {
1642
+ "epoch": 15.21,
1643
+ "learning_rate": 5.348672631430318e-06,
1644
+ "loss": 0.9463,
1645
+ "step": 270
1646
+ },
1647
+ {
1648
+ "epoch": 15.27,
1649
+ "learning_rate": 5.202443851943126e-06,
1650
+ "loss": 0.9942,
1651
+ "step": 271
1652
+ },
1653
+ {
1654
+ "epoch": 15.32,
1655
+ "learning_rate": 5.058009310946119e-06,
1656
+ "loss": 0.9657,
1657
+ "step": 272
1658
+ },
1659
+ {
1660
+ "epoch": 15.38,
1661
+ "learning_rate": 4.915382098439436e-06,
1662
+ "loss": 0.9375,
1663
+ "step": 273
1664
+ },
1665
+ {
1666
+ "epoch": 15.44,
1667
+ "learning_rate": 4.7745751406263165e-06,
1668
+ "loss": 0.9347,
1669
+ "step": 274
1670
+ },
1671
+ {
1672
+ "epoch": 15.49,
1673
+ "learning_rate": 4.6356011987416075e-06,
1674
+ "loss": 0.9796,
1675
+ "step": 275
1676
+ },
1677
+ {
1678
+ "epoch": 15.55,
1679
+ "learning_rate": 4.498472867895223e-06,
1680
+ "loss": 0.987,
1681
+ "step": 276
1682
+ },
1683
+ {
1684
+ "epoch": 15.61,
1685
+ "learning_rate": 4.36320257593065e-06,
1686
+ "loss": 0.8946,
1687
+ "step": 277
1688
+ },
1689
+ {
1690
+ "epoch": 15.66,
1691
+ "learning_rate": 4.229802582298633e-06,
1692
+ "loss": 0.9034,
1693
+ "step": 278
1694
+ },
1695
+ {
1696
+ "epoch": 15.72,
1697
+ "learning_rate": 4.098284976946101e-06,
1698
+ "loss": 1.0294,
1699
+ "step": 279
1700
+ },
1701
+ {
1702
+ "epoch": 15.77,
1703
+ "learning_rate": 3.968661679220468e-06,
1704
+ "loss": 0.9631,
1705
+ "step": 280
1706
+ },
1707
+ {
1708
+ "epoch": 15.83,
1709
+ "learning_rate": 3.840944436789393e-06,
1710
+ "loss": 0.8831,
1711
+ "step": 281
1712
+ },
1713
+ {
1714
+ "epoch": 15.89,
1715
+ "learning_rate": 3.715144824576078e-06,
1716
+ "loss": 0.9274,
1717
+ "step": 282
1718
+ },
1719
+ {
1720
+ "epoch": 15.94,
1721
+ "learning_rate": 3.591274243710277e-06,
1722
+ "loss": 0.8896,
1723
+ "step": 283
1724
+ },
1725
+ {
1726
+ "epoch": 16.0,
1727
+ "learning_rate": 3.4693439204949858e-06,
1728
+ "loss": 0.8672,
1729
+ "step": 284
1730
+ },
1731
+ {
1732
+ "epoch": 16.06,
1733
+ "learning_rate": 3.3493649053890326e-06,
1734
+ "loss": 0.9043,
1735
+ "step": 285
1736
+ },
1737
+ {
1738
+ "epoch": 16.11,
1739
+ "learning_rate": 3.2313480720055745e-06,
1740
+ "loss": 1.1059,
1741
+ "step": 286
1742
+ },
1743
+ {
1744
+ "epoch": 16.17,
1745
+ "learning_rate": 3.1153041161266134e-06,
1746
+ "loss": 0.9241,
1747
+ "step": 287
1748
+ },
1749
+ {
1750
+ "epoch": 16.23,
1751
+ "learning_rate": 3.0012435547336737e-06,
1752
+ "loss": 0.7761,
1753
+ "step": 288
1754
+ },
1755
+ {
1756
+ "epoch": 16.28,
1757
+ "learning_rate": 2.889176725054643e-06,
1758
+ "loss": 1.0323,
1759
+ "step": 289
1760
+ },
1761
+ {
1762
+ "epoch": 16.34,
1763
+ "learning_rate": 2.779113783626916e-06,
1764
+ "loss": 0.8534,
1765
+ "step": 290
1766
+ },
1767
+ {
1768
+ "epoch": 16.39,
1769
+ "learning_rate": 2.671064705376905e-06,
1770
+ "loss": 0.7669,
1771
+ "step": 291
1772
+ },
1773
+ {
1774
+ "epoch": 16.45,
1775
+ "learning_rate": 2.565039282716045e-06,
1776
+ "loss": 0.984,
1777
+ "step": 292
1778
+ },
1779
+ {
1780
+ "epoch": 16.51,
1781
+ "learning_rate": 2.461047124653279e-06,
1782
+ "loss": 0.9542,
1783
+ "step": 293
1784
+ },
1785
+ {
1786
+ "epoch": 16.56,
1787
+ "learning_rate": 2.3590976559242278e-06,
1788
+ "loss": 0.9439,
1789
+ "step": 294
1790
+ },
1791
+ {
1792
+ "epoch": 16.62,
1793
+ "learning_rate": 2.2592001161370392e-06,
1794
+ "loss": 0.8776,
1795
+ "step": 295
1796
+ },
1797
+ {
1798
+ "epoch": 16.68,
1799
+ "learning_rate": 2.1613635589349756e-06,
1800
+ "loss": 0.9211,
1801
+ "step": 296
1802
+ },
1803
+ {
1804
+ "epoch": 16.73,
1805
+ "learning_rate": 2.0655968511759134e-06,
1806
+ "loss": 0.9145,
1807
+ "step": 297
1808
+ },
1809
+ {
1810
+ "epoch": 16.79,
1811
+ "learning_rate": 1.97190867212875e-06,
1812
+ "loss": 0.9481,
1813
+ "step": 298
1814
+ },
1815
+ {
1816
+ "epoch": 16.85,
1817
+ "learning_rate": 1.8803075126867715e-06,
1818
+ "loss": 0.9805,
1819
+ "step": 299
1820
+ },
1821
+ {
1822
+ "epoch": 16.9,
1823
+ "learning_rate": 1.790801674598186e-06,
1824
+ "loss": 0.8387,
1825
+ "step": 300
1826
+ },
1827
+ {
1828
+ "epoch": 16.9,
1829
+ "eval_loss": 1.3825798034667969,
1830
+ "eval_runtime": 1.4357,
1831
+ "eval_samples_per_second": 6.965,
1832
+ "eval_steps_per_second": 3.483,
1833
+ "step": 300
1834
+ },
1835
+ {
1836
+ "epoch": 16.96,
1837
+ "learning_rate": 1.703399269713693e-06,
1838
+ "loss": 1.005,
1839
+ "step": 301
1840
+ },
1841
+ {
1842
+ "epoch": 17.01,
1843
+ "learning_rate": 1.6181082192513353e-06,
1844
+ "loss": 0.9032,
1845
+ "step": 302
1846
+ },
1847
+ {
1848
+ "epoch": 17.07,
1849
+ "learning_rate": 1.534936253078606e-06,
1850
+ "loss": 1.0064,
1851
+ "step": 303
1852
+ },
1853
+ {
1854
+ "epoch": 17.13,
1855
+ "learning_rate": 1.4538909090118846e-06,
1856
+ "loss": 0.8834,
1857
+ "step": 304
1858
+ },
1859
+ {
1860
+ "epoch": 17.18,
1861
+ "learning_rate": 1.3749795321332887e-06,
1862
+ "loss": 0.93,
1863
+ "step": 305
1864
+ },
1865
+ {
1866
+ "epoch": 17.24,
1867
+ "learning_rate": 1.2982092741250145e-06,
1868
+ "loss": 0.9181,
1869
+ "step": 306
1870
+ },
1871
+ {
1872
+ "epoch": 17.3,
1873
+ "learning_rate": 1.2235870926211619e-06,
1874
+ "loss": 0.8043,
1875
+ "step": 307
1876
+ },
1877
+ {
1878
+ "epoch": 17.35,
1879
+ "learning_rate": 1.1511197505771843e-06,
1880
+ "loss": 0.8552,
1881
+ "step": 308
1882
+ },
1883
+ {
1884
+ "epoch": 17.41,
1885
+ "learning_rate": 1.0808138156569614e-06,
1886
+ "loss": 0.8864,
1887
+ "step": 309
1888
+ },
1889
+ {
1890
+ "epoch": 17.46,
1891
+ "learning_rate": 1.0126756596375686e-06,
1892
+ "loss": 0.8856,
1893
+ "step": 310
1894
+ },
1895
+ {
1896
+ "epoch": 17.52,
1897
+ "learning_rate": 9.467114578318226e-07,
1898
+ "loss": 0.9447,
1899
+ "step": 311
1900
+ },
1901
+ {
1902
+ "epoch": 17.58,
1903
+ "learning_rate": 8.829271885286094e-07,
1904
+ "loss": 0.8091,
1905
+ "step": 312
1906
+ },
1907
+ {
1908
+ "epoch": 17.63,
1909
+ "learning_rate": 8.213286324510738e-07,
1910
+ "loss": 0.9956,
1911
+ "step": 313
1912
+ },
1913
+ {
1914
+ "epoch": 17.69,
1915
+ "learning_rate": 7.619213722327185e-07,
1916
+ "loss": 0.88,
1917
+ "step": 314
1918
+ },
1919
+ {
1920
+ "epoch": 17.75,
1921
+ "learning_rate": 7.047107919114588e-07,
1922
+ "loss": 0.9142,
1923
+ "step": 315
1924
+ },
1925
+ {
1926
+ "epoch": 17.8,
1927
+ "learning_rate": 6.497020764416633e-07,
1928
+ "loss": 0.9984,
1929
+ "step": 316
1930
+ },
1931
+ {
1932
+ "epoch": 17.86,
1933
+ "learning_rate": 5.969002112242494e-07,
1934
+ "loss": 0.8869,
1935
+ "step": 317
1936
+ },
1937
+ {
1938
+ "epoch": 17.92,
1939
+ "learning_rate": 5.463099816548579e-07,
1940
+ "loss": 0.9627,
1941
+ "step": 318
1942
+ },
1943
+ {
1944
+ "epoch": 17.97,
1945
+ "learning_rate": 4.979359726901639e-07,
1946
+ "loss": 0.9783,
1947
+ "step": 319
1948
+ }
1949
+ ],
1950
+ "logging_steps": 1,
1951
+ "max_steps": 340,
1952
+ "num_train_epochs": 20,
1953
+ "save_steps": 500,
1954
+ "total_flos": 3.997210073235456e+17,
1955
+ "trial_name": null,
1956
+ "trial_params": null
1957
+ }
checkpoint-319/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af7bb882cac5b4d33f676aa38a935fcc9173c63dc3da717a69a225490ea5e0f7
3
+ size 4155
checkpoint-337/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0.dev0
checkpoint-337/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
4
+ "bias": "none",
5
+ "fan_in_fan_out": null,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 32,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "k_proj",
18
+ "gate_proj",
19
+ "v_proj",
20
+ "o_proj",
21
+ "q_proj",
22
+ "down_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
checkpoint-337/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c20bcb5f4d6501e4e80dce150a08d03f4bf4213b17c1679c63d34539666854c4
3
+ size 500897101
checkpoint-337/adapter_model/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0.dev0
checkpoint-337/adapter_model/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
4
+ "bias": "none",
5
+ "fan_in_fan_out": null,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 32,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "k_proj",
18
+ "gate_proj",
19
+ "v_proj",
20
+ "o_proj",
21
+ "q_proj",
22
+ "down_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
checkpoint-337/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c20bcb5f4d6501e4e80dce150a08d03f4bf4213b17c1679c63d34539666854c4
3
+ size 500897101
checkpoint-337/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c996595fbd1ae30ef3b827127704e5751c9906776a1e473ba8cb98cab999084
3
+ size 1001737021
checkpoint-337/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c625ea4f2d62b26a085c62ff27dcfc90a47b8871ad5f185890c820cbaa720a5e
3
+ size 14575
checkpoint-337/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b95821819818de2a7391efb6614fc82b972e4a685f392f0abfd1090a329088f9
3
+ size 627
checkpoint-337/trainer_state.json ADDED
@@ -0,0 +1,2065 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 18.985915492957748,
5
+ "eval_steps": 100,
6
+ "global_step": 337,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.06,
13
+ "learning_rate": 5e-06,
14
+ "loss": 1.5896,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.11,
19
+ "learning_rate": 1e-05,
20
+ "loss": 1.6521,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.17,
25
+ "learning_rate": 1.5e-05,
26
+ "loss": 1.3654,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.23,
31
+ "learning_rate": 2e-05,
32
+ "loss": 1.3569,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.28,
37
+ "learning_rate": 2.5e-05,
38
+ "loss": 1.4352,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.34,
43
+ "learning_rate": 3e-05,
44
+ "loss": 1.425,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.39,
49
+ "learning_rate": 3.5e-05,
50
+ "loss": 1.3529,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.45,
55
+ "learning_rate": 4e-05,
56
+ "loss": 1.384,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.51,
61
+ "learning_rate": 4.5e-05,
62
+ "loss": 1.3881,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.56,
67
+ "learning_rate": 5e-05,
68
+ "loss": 1.5593,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.62,
73
+ "learning_rate": 4.999886713385432e-05,
74
+ "loss": 1.3423,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.68,
79
+ "learning_rate": 4.999546863808815e-05,
80
+ "loss": 1.4725,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.73,
85
+ "learning_rate": 4.9989804820704735e-05,
86
+ "loss": 1.4342,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.79,
91
+ "learning_rate": 4.9981876195011844e-05,
92
+ "loss": 1.3769,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.85,
97
+ "learning_rate": 4.99716834795752e-05,
98
+ "loss": 1.3976,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.9,
103
+ "learning_rate": 4.995922759815339e-05,
104
+ "loss": 1.328,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.96,
109
+ "learning_rate": 4.994450967961413e-05,
110
+ "loss": 1.266,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 1.01,
115
+ "learning_rate": 4.992753105783194e-05,
116
+ "loss": 1.4671,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 1.07,
121
+ "learning_rate": 4.9908293271567286e-05,
122
+ "loss": 1.3973,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 1.13,
127
+ "learning_rate": 4.988679806432712e-05,
128
+ "loss": 1.3835,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 1.18,
133
+ "learning_rate": 4.9863047384206835e-05,
134
+ "loss": 1.2203,
135
+ "step": 21
136
+ },
137
+ {
138
+ "epoch": 1.24,
139
+ "learning_rate": 4.9837043383713753e-05,
140
+ "loss": 1.3752,
141
+ "step": 22
142
+ },
143
+ {
144
+ "epoch": 1.3,
145
+ "learning_rate": 4.980878841957203e-05,
146
+ "loss": 1.4142,
147
+ "step": 23
148
+ },
149
+ {
150
+ "epoch": 1.35,
151
+ "learning_rate": 4.977828505250903e-05,
152
+ "loss": 1.3657,
153
+ "step": 24
154
+ },
155
+ {
156
+ "epoch": 1.41,
157
+ "learning_rate": 4.9745536047023324e-05,
158
+ "loss": 1.309,
159
+ "step": 25
160
+ },
161
+ {
162
+ "epoch": 1.46,
163
+ "learning_rate": 4.971054437113406e-05,
164
+ "loss": 1.3004,
165
+ "step": 26
166
+ },
167
+ {
168
+ "epoch": 1.52,
169
+ "learning_rate": 4.967331319611206e-05,
170
+ "loss": 1.3309,
171
+ "step": 27
172
+ },
173
+ {
174
+ "epoch": 1.58,
175
+ "learning_rate": 4.963384589619233e-05,
176
+ "loss": 1.4956,
177
+ "step": 28
178
+ },
179
+ {
180
+ "epoch": 1.63,
181
+ "learning_rate": 4.959214604826831e-05,
182
+ "loss": 1.2682,
183
+ "step": 29
184
+ },
185
+ {
186
+ "epoch": 1.69,
187
+ "learning_rate": 4.9548217431567665e-05,
188
+ "loss": 1.391,
189
+ "step": 30
190
+ },
191
+ {
192
+ "epoch": 1.75,
193
+ "learning_rate": 4.9502064027309836e-05,
194
+ "loss": 1.3102,
195
+ "step": 31
196
+ },
197
+ {
198
+ "epoch": 1.8,
199
+ "learning_rate": 4.9453690018345144e-05,
200
+ "loss": 1.3891,
201
+ "step": 32
202
+ },
203
+ {
204
+ "epoch": 1.86,
205
+ "learning_rate": 4.9403099788775754e-05,
206
+ "loss": 1.3538,
207
+ "step": 33
208
+ },
209
+ {
210
+ "epoch": 1.92,
211
+ "learning_rate": 4.935029792355834e-05,
212
+ "loss": 1.3443,
213
+ "step": 34
214
+ },
215
+ {
216
+ "epoch": 1.97,
217
+ "learning_rate": 4.929528920808854e-05,
218
+ "loss": 1.3384,
219
+ "step": 35
220
+ },
221
+ {
222
+ "epoch": 2.03,
223
+ "learning_rate": 4.923807862776728e-05,
224
+ "loss": 1.2968,
225
+ "step": 36
226
+ },
227
+ {
228
+ "epoch": 2.08,
229
+ "learning_rate": 4.917867136754893e-05,
230
+ "loss": 1.2624,
231
+ "step": 37
232
+ },
233
+ {
234
+ "epoch": 2.14,
235
+ "learning_rate": 4.91170728114714e-05,
236
+ "loss": 1.3547,
237
+ "step": 38
238
+ },
239
+ {
240
+ "epoch": 2.2,
241
+ "learning_rate": 4.9053288542168185e-05,
242
+ "loss": 1.3857,
243
+ "step": 39
244
+ },
245
+ {
246
+ "epoch": 2.25,
247
+ "learning_rate": 4.898732434036244e-05,
248
+ "loss": 1.2261,
249
+ "step": 40
250
+ },
251
+ {
252
+ "epoch": 2.31,
253
+ "learning_rate": 4.8919186184343046e-05,
254
+ "loss": 1.4072,
255
+ "step": 41
256
+ },
257
+ {
258
+ "epoch": 2.37,
259
+ "learning_rate": 4.8848880249422815e-05,
260
+ "loss": 1.3283,
261
+ "step": 42
262
+ },
263
+ {
264
+ "epoch": 2.42,
265
+ "learning_rate": 4.877641290737884e-05,
266
+ "loss": 1.2464,
267
+ "step": 43
268
+ },
269
+ {
270
+ "epoch": 2.48,
271
+ "learning_rate": 4.870179072587499e-05,
272
+ "loss": 1.3433,
273
+ "step": 44
274
+ },
275
+ {
276
+ "epoch": 2.54,
277
+ "learning_rate": 4.862502046786671e-05,
278
+ "loss": 1.2861,
279
+ "step": 45
280
+ },
281
+ {
282
+ "epoch": 2.59,
283
+ "learning_rate": 4.854610909098812e-05,
284
+ "loss": 1.1888,
285
+ "step": 46
286
+ },
287
+ {
288
+ "epoch": 2.65,
289
+ "learning_rate": 4.8465063746921395e-05,
290
+ "loss": 1.2509,
291
+ "step": 47
292
+ },
293
+ {
294
+ "epoch": 2.7,
295
+ "learning_rate": 4.838189178074867e-05,
296
+ "loss": 1.4449,
297
+ "step": 48
298
+ },
299
+ {
300
+ "epoch": 2.76,
301
+ "learning_rate": 4.829660073028631e-05,
302
+ "loss": 1.4645,
303
+ "step": 49
304
+ },
305
+ {
306
+ "epoch": 2.82,
307
+ "learning_rate": 4.8209198325401815e-05,
308
+ "loss": 1.3586,
309
+ "step": 50
310
+ },
311
+ {
312
+ "epoch": 2.87,
313
+ "learning_rate": 4.811969248731323e-05,
314
+ "loss": 1.3542,
315
+ "step": 51
316
+ },
317
+ {
318
+ "epoch": 2.93,
319
+ "learning_rate": 4.802809132787125e-05,
320
+ "loss": 1.3145,
321
+ "step": 52
322
+ },
323
+ {
324
+ "epoch": 2.99,
325
+ "learning_rate": 4.793440314882408e-05,
326
+ "loss": 1.2343,
327
+ "step": 53
328
+ },
329
+ {
330
+ "epoch": 3.04,
331
+ "learning_rate": 4.783863644106502e-05,
332
+ "loss": 1.2946,
333
+ "step": 54
334
+ },
335
+ {
336
+ "epoch": 3.1,
337
+ "learning_rate": 4.774079988386296e-05,
338
+ "loss": 1.2621,
339
+ "step": 55
340
+ },
341
+ {
342
+ "epoch": 3.15,
343
+ "learning_rate": 4.764090234407577e-05,
344
+ "loss": 1.3581,
345
+ "step": 56
346
+ },
347
+ {
348
+ "epoch": 3.21,
349
+ "learning_rate": 4.753895287534673e-05,
350
+ "loss": 1.3891,
351
+ "step": 57
352
+ },
353
+ {
354
+ "epoch": 3.27,
355
+ "learning_rate": 4.743496071728396e-05,
356
+ "loss": 1.334,
357
+ "step": 58
358
+ },
359
+ {
360
+ "epoch": 3.32,
361
+ "learning_rate": 4.73289352946231e-05,
362
+ "loss": 1.3105,
363
+ "step": 59
364
+ },
365
+ {
366
+ "epoch": 3.38,
367
+ "learning_rate": 4.722088621637309e-05,
368
+ "loss": 1.2921,
369
+ "step": 60
370
+ },
371
+ {
372
+ "epoch": 3.44,
373
+ "learning_rate": 4.711082327494536e-05,
374
+ "loss": 1.2406,
375
+ "step": 61
376
+ },
377
+ {
378
+ "epoch": 3.49,
379
+ "learning_rate": 4.6998756445266336e-05,
380
+ "loss": 1.4246,
381
+ "step": 62
382
+ },
383
+ {
384
+ "epoch": 3.55,
385
+ "learning_rate": 4.688469588387339e-05,
386
+ "loss": 1.2126,
387
+ "step": 63
388
+ },
389
+ {
390
+ "epoch": 3.61,
391
+ "learning_rate": 4.6768651927994434e-05,
392
+ "loss": 1.3662,
393
+ "step": 64
394
+ },
395
+ {
396
+ "epoch": 3.66,
397
+ "learning_rate": 4.665063509461097e-05,
398
+ "loss": 1.2044,
399
+ "step": 65
400
+ },
401
+ {
402
+ "epoch": 3.72,
403
+ "learning_rate": 4.653065607950502e-05,
404
+ "loss": 1.4344,
405
+ "step": 66
406
+ },
407
+ {
408
+ "epoch": 3.77,
409
+ "learning_rate": 4.640872575628973e-05,
410
+ "loss": 1.3652,
411
+ "step": 67
412
+ },
413
+ {
414
+ "epoch": 3.83,
415
+ "learning_rate": 4.628485517542392e-05,
416
+ "loss": 1.231,
417
+ "step": 68
418
+ },
419
+ {
420
+ "epoch": 3.89,
421
+ "learning_rate": 4.6159055563210604e-05,
422
+ "loss": 1.1948,
423
+ "step": 69
424
+ },
425
+ {
426
+ "epoch": 3.94,
427
+ "learning_rate": 4.6031338320779534e-05,
428
+ "loss": 1.2484,
429
+ "step": 70
430
+ },
431
+ {
432
+ "epoch": 4.0,
433
+ "learning_rate": 4.59017150230539e-05,
434
+ "loss": 1.2968,
435
+ "step": 71
436
+ },
437
+ {
438
+ "epoch": 4.06,
439
+ "learning_rate": 4.5770197417701365e-05,
440
+ "loss": 1.1691,
441
+ "step": 72
442
+ },
443
+ {
444
+ "epoch": 4.11,
445
+ "learning_rate": 4.563679742406935e-05,
446
+ "loss": 1.224,
447
+ "step": 73
448
+ },
449
+ {
450
+ "epoch": 4.17,
451
+ "learning_rate": 4.550152713210478e-05,
452
+ "loss": 1.2669,
453
+ "step": 74
454
+ },
455
+ {
456
+ "epoch": 4.23,
457
+ "learning_rate": 4.5364398801258396e-05,
458
+ "loss": 1.3218,
459
+ "step": 75
460
+ },
461
+ {
462
+ "epoch": 4.28,
463
+ "learning_rate": 4.522542485937369e-05,
464
+ "loss": 1.3309,
465
+ "step": 76
466
+ },
467
+ {
468
+ "epoch": 4.34,
469
+ "learning_rate": 4.508461790156056e-05,
470
+ "loss": 1.3356,
471
+ "step": 77
472
+ },
473
+ {
474
+ "epoch": 4.39,
475
+ "learning_rate": 4.4941990689053886e-05,
476
+ "loss": 1.3367,
477
+ "step": 78
478
+ },
479
+ {
480
+ "epoch": 4.45,
481
+ "learning_rate": 4.479755614805688e-05,
482
+ "loss": 1.3568,
483
+ "step": 79
484
+ },
485
+ {
486
+ "epoch": 4.51,
487
+ "learning_rate": 4.465132736856969e-05,
488
+ "loss": 1.29,
489
+ "step": 80
490
+ },
491
+ {
492
+ "epoch": 4.56,
493
+ "learning_rate": 4.450331760320302e-05,
494
+ "loss": 1.3522,
495
+ "step": 81
496
+ },
497
+ {
498
+ "epoch": 4.62,
499
+ "learning_rate": 4.4353540265977064e-05,
500
+ "loss": 1.3021,
501
+ "step": 82
502
+ },
503
+ {
504
+ "epoch": 4.68,
505
+ "learning_rate": 4.4202008931105795e-05,
506
+ "loss": 1.303,
507
+ "step": 83
508
+ },
509
+ {
510
+ "epoch": 4.73,
511
+ "learning_rate": 4.404873733176678e-05,
512
+ "loss": 1.313,
513
+ "step": 84
514
+ },
515
+ {
516
+ "epoch": 4.79,
517
+ "learning_rate": 4.389373935885646e-05,
518
+ "loss": 1.231,
519
+ "step": 85
520
+ },
521
+ {
522
+ "epoch": 4.85,
523
+ "learning_rate": 4.373702905973135e-05,
524
+ "loss": 1.2908,
525
+ "step": 86
526
+ },
527
+ {
528
+ "epoch": 4.9,
529
+ "learning_rate": 4.357862063693486e-05,
530
+ "loss": 1.2575,
531
+ "step": 87
532
+ },
533
+ {
534
+ "epoch": 4.96,
535
+ "learning_rate": 4.341852844691012e-05,
536
+ "loss": 1.1942,
537
+ "step": 88
538
+ },
539
+ {
540
+ "epoch": 5.01,
541
+ "learning_rate": 4.3256766998698936e-05,
542
+ "loss": 1.1607,
543
+ "step": 89
544
+ },
545
+ {
546
+ "epoch": 5.07,
547
+ "learning_rate": 4.309335095262676e-05,
548
+ "loss": 1.3386,
549
+ "step": 90
550
+ },
551
+ {
552
+ "epoch": 5.13,
553
+ "learning_rate": 4.292829511897409e-05,
554
+ "loss": 1.371,
555
+ "step": 91
556
+ },
557
+ {
558
+ "epoch": 5.18,
559
+ "learning_rate": 4.276161445663423e-05,
560
+ "loss": 1.236,
561
+ "step": 92
562
+ },
563
+ {
564
+ "epoch": 5.24,
565
+ "learning_rate": 4.259332407175751e-05,
566
+ "loss": 1.2531,
567
+ "step": 93
568
+ },
569
+ {
570
+ "epoch": 5.3,
571
+ "learning_rate": 4.242343921638234e-05,
572
+ "loss": 1.2593,
573
+ "step": 94
574
+ },
575
+ {
576
+ "epoch": 5.35,
577
+ "learning_rate": 4.2251975287052804e-05,
578
+ "loss": 1.1245,
579
+ "step": 95
580
+ },
581
+ {
582
+ "epoch": 5.41,
583
+ "learning_rate": 4.2078947823423364e-05,
584
+ "loss": 1.2538,
585
+ "step": 96
586
+ },
587
+ {
588
+ "epoch": 5.46,
589
+ "learning_rate": 4.1904372506850484e-05,
590
+ "loss": 1.2847,
591
+ "step": 97
592
+ },
593
+ {
594
+ "epoch": 5.52,
595
+ "learning_rate": 4.172826515897146e-05,
596
+ "loss": 1.2071,
597
+ "step": 98
598
+ },
599
+ {
600
+ "epoch": 5.58,
601
+ "learning_rate": 4.155064174027047e-05,
602
+ "loss": 1.2664,
603
+ "step": 99
604
+ },
605
+ {
606
+ "epoch": 5.63,
607
+ "learning_rate": 4.137151834863213e-05,
608
+ "loss": 1.1919,
609
+ "step": 100
610
+ },
611
+ {
612
+ "epoch": 5.63,
613
+ "eval_loss": 1.2261024713516235,
614
+ "eval_runtime": 1.4353,
615
+ "eval_samples_per_second": 6.967,
616
+ "eval_steps_per_second": 3.484,
617
+ "step": 100
618
+ },
619
+ {
620
+ "epoch": 5.69,
621
+ "learning_rate": 4.119091121788256e-05,
622
+ "loss": 1.1509,
623
+ "step": 101
624
+ },
625
+ {
626
+ "epoch": 5.75,
627
+ "learning_rate": 4.100883671631806e-05,
628
+ "loss": 1.3013,
629
+ "step": 102
630
+ },
631
+ {
632
+ "epoch": 5.8,
633
+ "learning_rate": 4.082531134522176e-05,
634
+ "loss": 1.313,
635
+ "step": 103
636
+ },
637
+ {
638
+ "epoch": 5.86,
639
+ "learning_rate": 4.064035173736804e-05,
640
+ "loss": 1.2176,
641
+ "step": 104
642
+ },
643
+ {
644
+ "epoch": 5.92,
645
+ "learning_rate": 4.045397465551513e-05,
646
+ "loss": 1.1004,
647
+ "step": 105
648
+ },
649
+ {
650
+ "epoch": 5.97,
651
+ "learning_rate": 4.0266196990885955e-05,
652
+ "loss": 1.3575,
653
+ "step": 106
654
+ },
655
+ {
656
+ "epoch": 6.03,
657
+ "learning_rate": 4.007703576163724e-05,
658
+ "loss": 1.2923,
659
+ "step": 107
660
+ },
661
+ {
662
+ "epoch": 6.08,
663
+ "learning_rate": 3.98865081113172e-05,
664
+ "loss": 1.2083,
665
+ "step": 108
666
+ },
667
+ {
668
+ "epoch": 6.14,
669
+ "learning_rate": 3.969463130731183e-05,
670
+ "loss": 1.2141,
671
+ "step": 109
672
+ },
673
+ {
674
+ "epoch": 6.2,
675
+ "learning_rate": 3.9501422739279956e-05,
676
+ "loss": 1.3297,
677
+ "step": 110
678
+ },
679
+ {
680
+ "epoch": 6.25,
681
+ "learning_rate": 3.9306899917577245e-05,
682
+ "loss": 1.0901,
683
+ "step": 111
684
+ },
685
+ {
686
+ "epoch": 6.31,
687
+ "learning_rate": 3.911108047166924e-05,
688
+ "loss": 1.2235,
689
+ "step": 112
690
+ },
691
+ {
692
+ "epoch": 6.37,
693
+ "learning_rate": 3.89139821485336e-05,
694
+ "loss": 1.341,
695
+ "step": 113
696
+ },
697
+ {
698
+ "epoch": 6.42,
699
+ "learning_rate": 3.871562281105175e-05,
700
+ "loss": 1.362,
701
+ "step": 114
702
+ },
703
+ {
704
+ "epoch": 6.48,
705
+ "learning_rate": 3.851602043638994e-05,
706
+ "loss": 1.2738,
707
+ "step": 115
708
+ },
709
+ {
710
+ "epoch": 6.54,
711
+ "learning_rate": 3.8315193114369996e-05,
712
+ "loss": 1.2264,
713
+ "step": 116
714
+ },
715
+ {
716
+ "epoch": 6.59,
717
+ "learning_rate": 3.8113159045829864e-05,
718
+ "loss": 1.171,
719
+ "step": 117
720
+ },
721
+ {
722
+ "epoch": 6.65,
723
+ "learning_rate": 3.790993654097405e-05,
724
+ "loss": 1.1848,
725
+ "step": 118
726
+ },
727
+ {
728
+ "epoch": 6.7,
729
+ "learning_rate": 3.770554401771423e-05,
730
+ "loss": 1.1545,
731
+ "step": 119
732
+ },
733
+ {
734
+ "epoch": 6.76,
735
+ "learning_rate": 3.7500000000000003e-05,
736
+ "loss": 1.2799,
737
+ "step": 120
738
+ },
739
+ {
740
+ "epoch": 6.82,
741
+ "learning_rate": 3.72933231161401e-05,
742
+ "loss": 1.2278,
743
+ "step": 121
744
+ },
745
+ {
746
+ "epoch": 6.87,
747
+ "learning_rate": 3.708553209711409e-05,
748
+ "loss": 1.1959,
749
+ "step": 122
750
+ },
751
+ {
752
+ "epoch": 6.93,
753
+ "learning_rate": 3.6876645774874876e-05,
754
+ "loss": 1.1151,
755
+ "step": 123
756
+ },
757
+ {
758
+ "epoch": 6.99,
759
+ "learning_rate": 3.6666683080641846e-05,
760
+ "loss": 1.2178,
761
+ "step": 124
762
+ },
763
+ {
764
+ "epoch": 7.04,
765
+ "learning_rate": 3.645566304318526e-05,
766
+ "loss": 1.1584,
767
+ "step": 125
768
+ },
769
+ {
770
+ "epoch": 7.1,
771
+ "learning_rate": 3.624360478710165e-05,
772
+ "loss": 1.0959,
773
+ "step": 126
774
+ },
775
+ {
776
+ "epoch": 7.15,
777
+ "learning_rate": 3.603052753108053e-05,
778
+ "loss": 1.3026,
779
+ "step": 127
780
+ },
781
+ {
782
+ "epoch": 7.21,
783
+ "learning_rate": 3.581645058616271e-05,
784
+ "loss": 1.2007,
785
+ "step": 128
786
+ },
787
+ {
788
+ "epoch": 7.27,
789
+ "learning_rate": 3.5601393353990046e-05,
790
+ "loss": 1.2242,
791
+ "step": 129
792
+ },
793
+ {
794
+ "epoch": 7.32,
795
+ "learning_rate": 3.5385375325047166e-05,
796
+ "loss": 1.2737,
797
+ "step": 130
798
+ },
799
+ {
800
+ "epoch": 7.38,
801
+ "learning_rate": 3.516841607689501e-05,
802
+ "loss": 1.2724,
803
+ "step": 131
804
+ },
805
+ {
806
+ "epoch": 7.44,
807
+ "learning_rate": 3.495053527239656e-05,
808
+ "loss": 1.1916,
809
+ "step": 132
810
+ },
811
+ {
812
+ "epoch": 7.49,
813
+ "learning_rate": 3.4731752657934794e-05,
814
+ "loss": 1.2107,
815
+ "step": 133
816
+ },
817
+ {
818
+ "epoch": 7.55,
819
+ "learning_rate": 3.4512088061623075e-05,
820
+ "loss": 1.246,
821
+ "step": 134
822
+ },
823
+ {
824
+ "epoch": 7.61,
825
+ "learning_rate": 3.4291561391508185e-05,
826
+ "loss": 1.1841,
827
+ "step": 135
828
+ },
829
+ {
830
+ "epoch": 7.66,
831
+ "learning_rate": 3.4070192633766025e-05,
832
+ "loss": 1.1898,
833
+ "step": 136
834
+ },
835
+ {
836
+ "epoch": 7.72,
837
+ "learning_rate": 3.3848001850890296e-05,
838
+ "loss": 1.1195,
839
+ "step": 137
840
+ },
841
+ {
842
+ "epoch": 7.77,
843
+ "learning_rate": 3.3625009179874265e-05,
844
+ "loss": 1.1685,
845
+ "step": 138
846
+ },
847
+ {
848
+ "epoch": 7.83,
849
+ "learning_rate": 3.3401234830385756e-05,
850
+ "loss": 1.2748,
851
+ "step": 139
852
+ },
853
+ {
854
+ "epoch": 7.89,
855
+ "learning_rate": 3.3176699082935545e-05,
856
+ "loss": 1.1697,
857
+ "step": 140
858
+ },
859
+ {
860
+ "epoch": 7.94,
861
+ "learning_rate": 3.295142228703938e-05,
862
+ "loss": 1.1624,
863
+ "step": 141
864
+ },
865
+ {
866
+ "epoch": 8.0,
867
+ "learning_rate": 3.272542485937369e-05,
868
+ "loss": 1.1679,
869
+ "step": 142
870
+ },
871
+ {
872
+ "epoch": 8.06,
873
+ "learning_rate": 3.249872728192527e-05,
874
+ "loss": 1.1163,
875
+ "step": 143
876
+ },
877
+ {
878
+ "epoch": 8.11,
879
+ "learning_rate": 3.2271350100134975e-05,
880
+ "loss": 1.1319,
881
+ "step": 144
882
+ },
883
+ {
884
+ "epoch": 8.17,
885
+ "learning_rate": 3.2043313921035743e-05,
886
+ "loss": 1.2151,
887
+ "step": 145
888
+ },
889
+ {
890
+ "epoch": 8.23,
891
+ "learning_rate": 3.181463941138495e-05,
892
+ "loss": 1.1339,
893
+ "step": 146
894
+ },
895
+ {
896
+ "epoch": 8.28,
897
+ "learning_rate": 3.158534729579142e-05,
898
+ "loss": 1.1503,
899
+ "step": 147
900
+ },
901
+ {
902
+ "epoch": 8.34,
903
+ "learning_rate": 3.135545835483718e-05,
904
+ "loss": 1.2326,
905
+ "step": 148
906
+ },
907
+ {
908
+ "epoch": 8.39,
909
+ "learning_rate": 3.11249934231941e-05,
910
+ "loss": 1.2938,
911
+ "step": 149
912
+ },
913
+ {
914
+ "epoch": 8.45,
915
+ "learning_rate": 3.0893973387735687e-05,
916
+ "loss": 1.1454,
917
+ "step": 150
918
+ },
919
+ {
920
+ "epoch": 8.51,
921
+ "learning_rate": 3.0662419185644115e-05,
922
+ "loss": 1.2504,
923
+ "step": 151
924
+ },
925
+ {
926
+ "epoch": 8.56,
927
+ "learning_rate": 3.0430351802512698e-05,
928
+ "loss": 0.987,
929
+ "step": 152
930
+ },
931
+ {
932
+ "epoch": 8.62,
933
+ "learning_rate": 3.0197792270443982e-05,
934
+ "loss": 1.1387,
935
+ "step": 153
936
+ },
937
+ {
938
+ "epoch": 8.68,
939
+ "learning_rate": 2.996476166614364e-05,
940
+ "loss": 1.1142,
941
+ "step": 154
942
+ },
943
+ {
944
+ "epoch": 8.73,
945
+ "learning_rate": 2.9731281109010256e-05,
946
+ "loss": 1.1714,
947
+ "step": 155
948
+ },
949
+ {
950
+ "epoch": 8.79,
951
+ "learning_rate": 2.9497371759221347e-05,
952
+ "loss": 1.1858,
953
+ "step": 156
954
+ },
955
+ {
956
+ "epoch": 8.85,
957
+ "learning_rate": 2.92630548158156e-05,
958
+ "loss": 1.2169,
959
+ "step": 157
960
+ },
961
+ {
962
+ "epoch": 8.9,
963
+ "learning_rate": 2.9028351514771606e-05,
964
+ "loss": 1.2204,
965
+ "step": 158
966
+ },
967
+ {
968
+ "epoch": 8.96,
969
+ "learning_rate": 2.8793283127083292e-05,
970
+ "loss": 1.0313,
971
+ "step": 159
972
+ },
973
+ {
974
+ "epoch": 9.01,
975
+ "learning_rate": 2.8557870956832132e-05,
976
+ "loss": 1.2432,
977
+ "step": 160
978
+ },
979
+ {
980
+ "epoch": 9.07,
981
+ "learning_rate": 2.8322136339256356e-05,
982
+ "loss": 1.0614,
983
+ "step": 161
984
+ },
985
+ {
986
+ "epoch": 9.13,
987
+ "learning_rate": 2.808610063881737e-05,
988
+ "loss": 1.1682,
989
+ "step": 162
990
+ },
991
+ {
992
+ "epoch": 9.18,
993
+ "learning_rate": 2.7849785247263515e-05,
994
+ "loss": 1.0238,
995
+ "step": 163
996
+ },
997
+ {
998
+ "epoch": 9.24,
999
+ "learning_rate": 2.761321158169134e-05,
1000
+ "loss": 1.166,
1001
+ "step": 164
1002
+ },
1003
+ {
1004
+ "epoch": 9.3,
1005
+ "learning_rate": 2.7376401082604564e-05,
1006
+ "loss": 1.0965,
1007
+ "step": 165
1008
+ },
1009
+ {
1010
+ "epoch": 9.35,
1011
+ "learning_rate": 2.7139375211970996e-05,
1012
+ "loss": 1.096,
1013
+ "step": 166
1014
+ },
1015
+ {
1016
+ "epoch": 9.41,
1017
+ "learning_rate": 2.6902155451277377e-05,
1018
+ "loss": 1.1161,
1019
+ "step": 167
1020
+ },
1021
+ {
1022
+ "epoch": 9.46,
1023
+ "learning_rate": 2.6664763299582602e-05,
1024
+ "loss": 1.1328,
1025
+ "step": 168
1026
+ },
1027
+ {
1028
+ "epoch": 9.52,
1029
+ "learning_rate": 2.6427220271569203e-05,
1030
+ "loss": 1.0747,
1031
+ "step": 169
1032
+ },
1033
+ {
1034
+ "epoch": 9.58,
1035
+ "learning_rate": 2.6189547895593562e-05,
1036
+ "loss": 1.1092,
1037
+ "step": 170
1038
+ },
1039
+ {
1040
+ "epoch": 9.63,
1041
+ "learning_rate": 2.5951767711734753e-05,
1042
+ "loss": 1.1624,
1043
+ "step": 171
1044
+ },
1045
+ {
1046
+ "epoch": 9.69,
1047
+ "learning_rate": 2.5713901269842404e-05,
1048
+ "loss": 1.1434,
1049
+ "step": 172
1050
+ },
1051
+ {
1052
+ "epoch": 9.75,
1053
+ "learning_rate": 2.5475970127583666e-05,
1054
+ "loss": 1.1167,
1055
+ "step": 173
1056
+ },
1057
+ {
1058
+ "epoch": 9.8,
1059
+ "learning_rate": 2.523799584848942e-05,
1060
+ "loss": 1.0509,
1061
+ "step": 174
1062
+ },
1063
+ {
1064
+ "epoch": 9.86,
1065
+ "learning_rate": 2.5e-05,
1066
+ "loss": 1.1449,
1067
+ "step": 175
1068
+ },
1069
+ {
1070
+ "epoch": 9.92,
1071
+ "learning_rate": 2.4762004151510584e-05,
1072
+ "loss": 1.109,
1073
+ "step": 176
1074
+ },
1075
+ {
1076
+ "epoch": 9.97,
1077
+ "learning_rate": 2.4524029872416333e-05,
1078
+ "loss": 1.0272,
1079
+ "step": 177
1080
+ },
1081
+ {
1082
+ "epoch": 10.03,
1083
+ "learning_rate": 2.42860987301576e-05,
1084
+ "loss": 1.1489,
1085
+ "step": 178
1086
+ },
1087
+ {
1088
+ "epoch": 10.08,
1089
+ "learning_rate": 2.4048232288265253e-05,
1090
+ "loss": 1.0263,
1091
+ "step": 179
1092
+ },
1093
+ {
1094
+ "epoch": 10.14,
1095
+ "learning_rate": 2.3810452104406444e-05,
1096
+ "loss": 1.0305,
1097
+ "step": 180
1098
+ },
1099
+ {
1100
+ "epoch": 10.2,
1101
+ "learning_rate": 2.35727797284308e-05,
1102
+ "loss": 0.9612,
1103
+ "step": 181
1104
+ },
1105
+ {
1106
+ "epoch": 10.25,
1107
+ "learning_rate": 2.3335236700417404e-05,
1108
+ "loss": 1.1883,
1109
+ "step": 182
1110
+ },
1111
+ {
1112
+ "epoch": 10.31,
1113
+ "learning_rate": 2.309784454872262e-05,
1114
+ "loss": 1.192,
1115
+ "step": 183
1116
+ },
1117
+ {
1118
+ "epoch": 10.37,
1119
+ "learning_rate": 2.2860624788029013e-05,
1120
+ "loss": 1.0919,
1121
+ "step": 184
1122
+ },
1123
+ {
1124
+ "epoch": 10.42,
1125
+ "learning_rate": 2.2623598917395438e-05,
1126
+ "loss": 1.0974,
1127
+ "step": 185
1128
+ },
1129
+ {
1130
+ "epoch": 10.48,
1131
+ "learning_rate": 2.238678841830867e-05,
1132
+ "loss": 1.0938,
1133
+ "step": 186
1134
+ },
1135
+ {
1136
+ "epoch": 10.54,
1137
+ "learning_rate": 2.2150214752736488e-05,
1138
+ "loss": 0.9879,
1139
+ "step": 187
1140
+ },
1141
+ {
1142
+ "epoch": 10.59,
1143
+ "learning_rate": 2.1913899361182632e-05,
1144
+ "loss": 1.0284,
1145
+ "step": 188
1146
+ },
1147
+ {
1148
+ "epoch": 10.65,
1149
+ "learning_rate": 2.167786366074365e-05,
1150
+ "loss": 1.0184,
1151
+ "step": 189
1152
+ },
1153
+ {
1154
+ "epoch": 10.7,
1155
+ "learning_rate": 2.1442129043167874e-05,
1156
+ "loss": 1.0419,
1157
+ "step": 190
1158
+ },
1159
+ {
1160
+ "epoch": 10.76,
1161
+ "learning_rate": 2.1206716872916714e-05,
1162
+ "loss": 1.0631,
1163
+ "step": 191
1164
+ },
1165
+ {
1166
+ "epoch": 10.82,
1167
+ "learning_rate": 2.09716484852284e-05,
1168
+ "loss": 1.0492,
1169
+ "step": 192
1170
+ },
1171
+ {
1172
+ "epoch": 10.87,
1173
+ "learning_rate": 2.0736945184184405e-05,
1174
+ "loss": 1.0924,
1175
+ "step": 193
1176
+ },
1177
+ {
1178
+ "epoch": 10.93,
1179
+ "learning_rate": 2.0502628240778655e-05,
1180
+ "loss": 0.9832,
1181
+ "step": 194
1182
+ },
1183
+ {
1184
+ "epoch": 10.99,
1185
+ "learning_rate": 2.0268718890989753e-05,
1186
+ "loss": 1.1266,
1187
+ "step": 195
1188
+ },
1189
+ {
1190
+ "epoch": 11.04,
1191
+ "learning_rate": 2.003523833385637e-05,
1192
+ "loss": 1.0762,
1193
+ "step": 196
1194
+ },
1195
+ {
1196
+ "epoch": 11.1,
1197
+ "learning_rate": 1.980220772955602e-05,
1198
+ "loss": 1.0011,
1199
+ "step": 197
1200
+ },
1201
+ {
1202
+ "epoch": 11.15,
1203
+ "learning_rate": 1.956964819748731e-05,
1204
+ "loss": 1.0195,
1205
+ "step": 198
1206
+ },
1207
+ {
1208
+ "epoch": 11.21,
1209
+ "learning_rate": 1.9337580814355888e-05,
1210
+ "loss": 1.0442,
1211
+ "step": 199
1212
+ },
1213
+ {
1214
+ "epoch": 11.27,
1215
+ "learning_rate": 1.9106026612264316e-05,
1216
+ "loss": 1.0285,
1217
+ "step": 200
1218
+ },
1219
+ {
1220
+ "epoch": 11.27,
1221
+ "eval_loss": 1.3292417526245117,
1222
+ "eval_runtime": 1.4356,
1223
+ "eval_samples_per_second": 6.966,
1224
+ "eval_steps_per_second": 3.483,
1225
+ "step": 200
1226
+ },
1227
+ {
1228
+ "epoch": 11.32,
1229
+ "learning_rate": 1.8875006576805914e-05,
1230
+ "loss": 0.9805,
1231
+ "step": 201
1232
+ },
1233
+ {
1234
+ "epoch": 11.38,
1235
+ "learning_rate": 1.8644541645162834e-05,
1236
+ "loss": 0.9562,
1237
+ "step": 202
1238
+ },
1239
+ {
1240
+ "epoch": 11.44,
1241
+ "learning_rate": 1.8414652704208583e-05,
1242
+ "loss": 1.0457,
1243
+ "step": 203
1244
+ },
1245
+ {
1246
+ "epoch": 11.49,
1247
+ "learning_rate": 1.8185360588615058e-05,
1248
+ "loss": 1.0217,
1249
+ "step": 204
1250
+ },
1251
+ {
1252
+ "epoch": 11.55,
1253
+ "learning_rate": 1.795668607896426e-05,
1254
+ "loss": 1.1281,
1255
+ "step": 205
1256
+ },
1257
+ {
1258
+ "epoch": 11.61,
1259
+ "learning_rate": 1.7728649899865024e-05,
1260
+ "loss": 0.9654,
1261
+ "step": 206
1262
+ },
1263
+ {
1264
+ "epoch": 11.66,
1265
+ "learning_rate": 1.7501272718074737e-05,
1266
+ "loss": 0.9011,
1267
+ "step": 207
1268
+ },
1269
+ {
1270
+ "epoch": 11.72,
1271
+ "learning_rate": 1.7274575140626318e-05,
1272
+ "loss": 1.097,
1273
+ "step": 208
1274
+ },
1275
+ {
1276
+ "epoch": 11.77,
1277
+ "learning_rate": 1.7048577712960627e-05,
1278
+ "loss": 1.0652,
1279
+ "step": 209
1280
+ },
1281
+ {
1282
+ "epoch": 11.83,
1283
+ "learning_rate": 1.682330091706446e-05,
1284
+ "loss": 1.011,
1285
+ "step": 210
1286
+ },
1287
+ {
1288
+ "epoch": 11.89,
1289
+ "learning_rate": 1.6598765169614243e-05,
1290
+ "loss": 1.1001,
1291
+ "step": 211
1292
+ },
1293
+ {
1294
+ "epoch": 11.94,
1295
+ "learning_rate": 1.637499082012574e-05,
1296
+ "loss": 1.1224,
1297
+ "step": 212
1298
+ },
1299
+ {
1300
+ "epoch": 12.0,
1301
+ "learning_rate": 1.615199814910971e-05,
1302
+ "loss": 0.967,
1303
+ "step": 213
1304
+ },
1305
+ {
1306
+ "epoch": 12.06,
1307
+ "learning_rate": 1.5929807366233977e-05,
1308
+ "loss": 0.9923,
1309
+ "step": 214
1310
+ },
1311
+ {
1312
+ "epoch": 12.11,
1313
+ "learning_rate": 1.5708438608491814e-05,
1314
+ "loss": 1.0722,
1315
+ "step": 215
1316
+ },
1317
+ {
1318
+ "epoch": 12.17,
1319
+ "learning_rate": 1.5487911938376924e-05,
1320
+ "loss": 0.9666,
1321
+ "step": 216
1322
+ },
1323
+ {
1324
+ "epoch": 12.23,
1325
+ "learning_rate": 1.5268247342065215e-05,
1326
+ "loss": 0.9136,
1327
+ "step": 217
1328
+ },
1329
+ {
1330
+ "epoch": 12.28,
1331
+ "learning_rate": 1.5049464727603452e-05,
1332
+ "loss": 0.9274,
1333
+ "step": 218
1334
+ },
1335
+ {
1336
+ "epoch": 12.34,
1337
+ "learning_rate": 1.4831583923104999e-05,
1338
+ "loss": 1.0955,
1339
+ "step": 219
1340
+ },
1341
+ {
1342
+ "epoch": 12.39,
1343
+ "learning_rate": 1.4614624674952842e-05,
1344
+ "loss": 0.9058,
1345
+ "step": 220
1346
+ },
1347
+ {
1348
+ "epoch": 12.45,
1349
+ "learning_rate": 1.4398606646009963e-05,
1350
+ "loss": 0.8351,
1351
+ "step": 221
1352
+ },
1353
+ {
1354
+ "epoch": 12.51,
1355
+ "learning_rate": 1.4183549413837289e-05,
1356
+ "loss": 1.119,
1357
+ "step": 222
1358
+ },
1359
+ {
1360
+ "epoch": 12.56,
1361
+ "learning_rate": 1.3969472468919461e-05,
1362
+ "loss": 1.0418,
1363
+ "step": 223
1364
+ },
1365
+ {
1366
+ "epoch": 12.62,
1367
+ "learning_rate": 1.3756395212898359e-05,
1368
+ "loss": 0.9102,
1369
+ "step": 224
1370
+ },
1371
+ {
1372
+ "epoch": 12.68,
1373
+ "learning_rate": 1.354433695681474e-05,
1374
+ "loss": 0.997,
1375
+ "step": 225
1376
+ },
1377
+ {
1378
+ "epoch": 12.73,
1379
+ "learning_rate": 1.3333316919358157e-05,
1380
+ "loss": 1.1737,
1381
+ "step": 226
1382
+ },
1383
+ {
1384
+ "epoch": 12.79,
1385
+ "learning_rate": 1.3123354225125128e-05,
1386
+ "loss": 0.9887,
1387
+ "step": 227
1388
+ },
1389
+ {
1390
+ "epoch": 12.85,
1391
+ "learning_rate": 1.2914467902885901e-05,
1392
+ "loss": 0.9984,
1393
+ "step": 228
1394
+ },
1395
+ {
1396
+ "epoch": 12.9,
1397
+ "learning_rate": 1.2706676883859903e-05,
1398
+ "loss": 0.9507,
1399
+ "step": 229
1400
+ },
1401
+ {
1402
+ "epoch": 12.96,
1403
+ "learning_rate": 1.2500000000000006e-05,
1404
+ "loss": 0.9395,
1405
+ "step": 230
1406
+ },
1407
+ {
1408
+ "epoch": 13.01,
1409
+ "learning_rate": 1.2294455982285777e-05,
1410
+ "loss": 1.1445,
1411
+ "step": 231
1412
+ },
1413
+ {
1414
+ "epoch": 13.07,
1415
+ "learning_rate": 1.2090063459025955e-05,
1416
+ "loss": 0.8469,
1417
+ "step": 232
1418
+ },
1419
+ {
1420
+ "epoch": 13.13,
1421
+ "learning_rate": 1.1886840954170142e-05,
1422
+ "loss": 0.9936,
1423
+ "step": 233
1424
+ },
1425
+ {
1426
+ "epoch": 13.18,
1427
+ "learning_rate": 1.1684806885630004e-05,
1428
+ "loss": 0.8797,
1429
+ "step": 234
1430
+ },
1431
+ {
1432
+ "epoch": 13.24,
1433
+ "learning_rate": 1.148397956361007e-05,
1434
+ "loss": 1.1331,
1435
+ "step": 235
1436
+ },
1437
+ {
1438
+ "epoch": 13.3,
1439
+ "learning_rate": 1.1284377188948258e-05,
1440
+ "loss": 0.9613,
1441
+ "step": 236
1442
+ },
1443
+ {
1444
+ "epoch": 13.35,
1445
+ "learning_rate": 1.1086017851466413e-05,
1446
+ "loss": 0.9343,
1447
+ "step": 237
1448
+ },
1449
+ {
1450
+ "epoch": 13.41,
1451
+ "learning_rate": 1.0888919528330777e-05,
1452
+ "loss": 1.0098,
1453
+ "step": 238
1454
+ },
1455
+ {
1456
+ "epoch": 13.46,
1457
+ "learning_rate": 1.0693100082422763e-05,
1458
+ "loss": 0.9669,
1459
+ "step": 239
1460
+ },
1461
+ {
1462
+ "epoch": 13.52,
1463
+ "learning_rate": 1.049857726072005e-05,
1464
+ "loss": 0.9686,
1465
+ "step": 240
1466
+ },
1467
+ {
1468
+ "epoch": 13.58,
1469
+ "learning_rate": 1.0305368692688174e-05,
1470
+ "loss": 0.8916,
1471
+ "step": 241
1472
+ },
1473
+ {
1474
+ "epoch": 13.63,
1475
+ "learning_rate": 1.0113491888682802e-05,
1476
+ "loss": 0.9274,
1477
+ "step": 242
1478
+ },
1479
+ {
1480
+ "epoch": 13.69,
1481
+ "learning_rate": 9.922964238362762e-06,
1482
+ "loss": 0.9307,
1483
+ "step": 243
1484
+ },
1485
+ {
1486
+ "epoch": 13.75,
1487
+ "learning_rate": 9.733803009114045e-06,
1488
+ "loss": 1.0112,
1489
+ "step": 244
1490
+ },
1491
+ {
1492
+ "epoch": 13.8,
1493
+ "learning_rate": 9.546025344484869e-06,
1494
+ "loss": 1.0465,
1495
+ "step": 245
1496
+ },
1497
+ {
1498
+ "epoch": 13.86,
1499
+ "learning_rate": 9.359648262631961e-06,
1500
+ "loss": 1.0822,
1501
+ "step": 246
1502
+ },
1503
+ {
1504
+ "epoch": 13.92,
1505
+ "learning_rate": 9.174688654778243e-06,
1506
+ "loss": 0.8992,
1507
+ "step": 247
1508
+ },
1509
+ {
1510
+ "epoch": 13.97,
1511
+ "learning_rate": 8.991163283681944e-06,
1512
+ "loss": 0.8778,
1513
+ "step": 248
1514
+ },
1515
+ {
1516
+ "epoch": 14.03,
1517
+ "learning_rate": 8.809088782117452e-06,
1518
+ "loss": 1.0079,
1519
+ "step": 249
1520
+ },
1521
+ {
1522
+ "epoch": 14.08,
1523
+ "learning_rate": 8.628481651367876e-06,
1524
+ "loss": 0.9508,
1525
+ "step": 250
1526
+ },
1527
+ {
1528
+ "epoch": 14.14,
1529
+ "learning_rate": 8.449358259729537e-06,
1530
+ "loss": 0.997,
1531
+ "step": 251
1532
+ },
1533
+ {
1534
+ "epoch": 14.2,
1535
+ "learning_rate": 8.271734841028553e-06,
1536
+ "loss": 0.8669,
1537
+ "step": 252
1538
+ },
1539
+ {
1540
+ "epoch": 14.25,
1541
+ "learning_rate": 8.09562749314952e-06,
1542
+ "loss": 0.9721,
1543
+ "step": 253
1544
+ },
1545
+ {
1546
+ "epoch": 14.31,
1547
+ "learning_rate": 7.921052176576644e-06,
1548
+ "loss": 1.0777,
1549
+ "step": 254
1550
+ },
1551
+ {
1552
+ "epoch": 14.37,
1553
+ "learning_rate": 7.748024712947205e-06,
1554
+ "loss": 0.8644,
1555
+ "step": 255
1556
+ },
1557
+ {
1558
+ "epoch": 14.42,
1559
+ "learning_rate": 7.576560783617668e-06,
1560
+ "loss": 0.9119,
1561
+ "step": 256
1562
+ },
1563
+ {
1564
+ "epoch": 14.48,
1565
+ "learning_rate": 7.40667592824249e-06,
1566
+ "loss": 0.8645,
1567
+ "step": 257
1568
+ },
1569
+ {
1570
+ "epoch": 14.54,
1571
+ "learning_rate": 7.238385543365783e-06,
1572
+ "loss": 0.8786,
1573
+ "step": 258
1574
+ },
1575
+ {
1576
+ "epoch": 14.59,
1577
+ "learning_rate": 7.071704881025915e-06,
1578
+ "loss": 0.9539,
1579
+ "step": 259
1580
+ },
1581
+ {
1582
+ "epoch": 14.65,
1583
+ "learning_rate": 6.906649047373246e-06,
1584
+ "loss": 0.9521,
1585
+ "step": 260
1586
+ },
1587
+ {
1588
+ "epoch": 14.7,
1589
+ "learning_rate": 6.743233001301067e-06,
1590
+ "loss": 1.0708,
1591
+ "step": 261
1592
+ },
1593
+ {
1594
+ "epoch": 14.76,
1595
+ "learning_rate": 6.5814715530898745e-06,
1596
+ "loss": 1.033,
1597
+ "step": 262
1598
+ },
1599
+ {
1600
+ "epoch": 14.82,
1601
+ "learning_rate": 6.421379363065142e-06,
1602
+ "loss": 0.9344,
1603
+ "step": 263
1604
+ },
1605
+ {
1606
+ "epoch": 14.87,
1607
+ "learning_rate": 6.2629709402686535e-06,
1608
+ "loss": 0.9491,
1609
+ "step": 264
1610
+ },
1611
+ {
1612
+ "epoch": 14.93,
1613
+ "learning_rate": 6.106260641143546e-06,
1614
+ "loss": 0.8894,
1615
+ "step": 265
1616
+ },
1617
+ {
1618
+ "epoch": 14.99,
1619
+ "learning_rate": 5.951262668233232e-06,
1620
+ "loss": 0.8825,
1621
+ "step": 266
1622
+ },
1623
+ {
1624
+ "epoch": 15.04,
1625
+ "learning_rate": 5.797991068894201e-06,
1626
+ "loss": 0.925,
1627
+ "step": 267
1628
+ },
1629
+ {
1630
+ "epoch": 15.1,
1631
+ "learning_rate": 5.646459734022938e-06,
1632
+ "loss": 0.8431,
1633
+ "step": 268
1634
+ },
1635
+ {
1636
+ "epoch": 15.15,
1637
+ "learning_rate": 5.49668239679699e-06,
1638
+ "loss": 1.036,
1639
+ "step": 269
1640
+ },
1641
+ {
1642
+ "epoch": 15.21,
1643
+ "learning_rate": 5.348672631430318e-06,
1644
+ "loss": 0.9463,
1645
+ "step": 270
1646
+ },
1647
+ {
1648
+ "epoch": 15.27,
1649
+ "learning_rate": 5.202443851943126e-06,
1650
+ "loss": 0.9942,
1651
+ "step": 271
1652
+ },
1653
+ {
1654
+ "epoch": 15.32,
1655
+ "learning_rate": 5.058009310946119e-06,
1656
+ "loss": 0.9657,
1657
+ "step": 272
1658
+ },
1659
+ {
1660
+ "epoch": 15.38,
1661
+ "learning_rate": 4.915382098439436e-06,
1662
+ "loss": 0.9375,
1663
+ "step": 273
1664
+ },
1665
+ {
1666
+ "epoch": 15.44,
1667
+ "learning_rate": 4.7745751406263165e-06,
1668
+ "loss": 0.9347,
1669
+ "step": 274
1670
+ },
1671
+ {
1672
+ "epoch": 15.49,
1673
+ "learning_rate": 4.6356011987416075e-06,
1674
+ "loss": 0.9796,
1675
+ "step": 275
1676
+ },
1677
+ {
1678
+ "epoch": 15.55,
1679
+ "learning_rate": 4.498472867895223e-06,
1680
+ "loss": 0.987,
1681
+ "step": 276
1682
+ },
1683
+ {
1684
+ "epoch": 15.61,
1685
+ "learning_rate": 4.36320257593065e-06,
1686
+ "loss": 0.8946,
1687
+ "step": 277
1688
+ },
1689
+ {
1690
+ "epoch": 15.66,
1691
+ "learning_rate": 4.229802582298633e-06,
1692
+ "loss": 0.9034,
1693
+ "step": 278
1694
+ },
1695
+ {
1696
+ "epoch": 15.72,
1697
+ "learning_rate": 4.098284976946101e-06,
1698
+ "loss": 1.0294,
1699
+ "step": 279
1700
+ },
1701
+ {
1702
+ "epoch": 15.77,
1703
+ "learning_rate": 3.968661679220468e-06,
1704
+ "loss": 0.9631,
1705
+ "step": 280
1706
+ },
1707
+ {
1708
+ "epoch": 15.83,
1709
+ "learning_rate": 3.840944436789393e-06,
1710
+ "loss": 0.8831,
1711
+ "step": 281
1712
+ },
1713
+ {
1714
+ "epoch": 15.89,
1715
+ "learning_rate": 3.715144824576078e-06,
1716
+ "loss": 0.9274,
1717
+ "step": 282
1718
+ },
1719
+ {
1720
+ "epoch": 15.94,
1721
+ "learning_rate": 3.591274243710277e-06,
1722
+ "loss": 0.8896,
1723
+ "step": 283
1724
+ },
1725
+ {
1726
+ "epoch": 16.0,
1727
+ "learning_rate": 3.4693439204949858e-06,
1728
+ "loss": 0.8672,
1729
+ "step": 284
1730
+ },
1731
+ {
1732
+ "epoch": 16.06,
1733
+ "learning_rate": 3.3493649053890326e-06,
1734
+ "loss": 0.9043,
1735
+ "step": 285
1736
+ },
1737
+ {
1738
+ "epoch": 16.11,
1739
+ "learning_rate": 3.2313480720055745e-06,
1740
+ "loss": 1.1059,
1741
+ "step": 286
1742
+ },
1743
+ {
1744
+ "epoch": 16.17,
1745
+ "learning_rate": 3.1153041161266134e-06,
1746
+ "loss": 0.9241,
1747
+ "step": 287
1748
+ },
1749
+ {
1750
+ "epoch": 16.23,
1751
+ "learning_rate": 3.0012435547336737e-06,
1752
+ "loss": 0.7761,
1753
+ "step": 288
1754
+ },
1755
+ {
1756
+ "epoch": 16.28,
1757
+ "learning_rate": 2.889176725054643e-06,
1758
+ "loss": 1.0323,
1759
+ "step": 289
1760
+ },
1761
+ {
1762
+ "epoch": 16.34,
1763
+ "learning_rate": 2.779113783626916e-06,
1764
+ "loss": 0.8534,
1765
+ "step": 290
1766
+ },
1767
+ {
1768
+ "epoch": 16.39,
1769
+ "learning_rate": 2.671064705376905e-06,
1770
+ "loss": 0.7669,
1771
+ "step": 291
1772
+ },
1773
+ {
1774
+ "epoch": 16.45,
1775
+ "learning_rate": 2.565039282716045e-06,
1776
+ "loss": 0.984,
1777
+ "step": 292
1778
+ },
1779
+ {
1780
+ "epoch": 16.51,
1781
+ "learning_rate": 2.461047124653279e-06,
1782
+ "loss": 0.9542,
1783
+ "step": 293
1784
+ },
1785
+ {
1786
+ "epoch": 16.56,
1787
+ "learning_rate": 2.3590976559242278e-06,
1788
+ "loss": 0.9439,
1789
+ "step": 294
1790
+ },
1791
+ {
1792
+ "epoch": 16.62,
1793
+ "learning_rate": 2.2592001161370392e-06,
1794
+ "loss": 0.8776,
1795
+ "step": 295
1796
+ },
1797
+ {
1798
+ "epoch": 16.68,
1799
+ "learning_rate": 2.1613635589349756e-06,
1800
+ "loss": 0.9211,
1801
+ "step": 296
1802
+ },
1803
+ {
1804
+ "epoch": 16.73,
1805
+ "learning_rate": 2.0655968511759134e-06,
1806
+ "loss": 0.9145,
1807
+ "step": 297
1808
+ },
1809
+ {
1810
+ "epoch": 16.79,
1811
+ "learning_rate": 1.97190867212875e-06,
1812
+ "loss": 0.9481,
1813
+ "step": 298
1814
+ },
1815
+ {
1816
+ "epoch": 16.85,
1817
+ "learning_rate": 1.8803075126867715e-06,
1818
+ "loss": 0.9805,
1819
+ "step": 299
1820
+ },
1821
+ {
1822
+ "epoch": 16.9,
1823
+ "learning_rate": 1.790801674598186e-06,
1824
+ "loss": 0.8387,
1825
+ "step": 300
1826
+ },
1827
+ {
1828
+ "epoch": 16.9,
1829
+ "eval_loss": 1.3825798034667969,
1830
+ "eval_runtime": 1.4357,
1831
+ "eval_samples_per_second": 6.965,
1832
+ "eval_steps_per_second": 3.483,
1833
+ "step": 300
1834
+ },
1835
+ {
1836
+ "epoch": 16.96,
1837
+ "learning_rate": 1.703399269713693e-06,
1838
+ "loss": 1.005,
1839
+ "step": 301
1840
+ },
1841
+ {
1842
+ "epoch": 17.01,
1843
+ "learning_rate": 1.6181082192513353e-06,
1844
+ "loss": 0.9032,
1845
+ "step": 302
1846
+ },
1847
+ {
1848
+ "epoch": 17.07,
1849
+ "learning_rate": 1.534936253078606e-06,
1850
+ "loss": 1.0064,
1851
+ "step": 303
1852
+ },
1853
+ {
1854
+ "epoch": 17.13,
1855
+ "learning_rate": 1.4538909090118846e-06,
1856
+ "loss": 0.8834,
1857
+ "step": 304
1858
+ },
1859
+ {
1860
+ "epoch": 17.18,
1861
+ "learning_rate": 1.3749795321332887e-06,
1862
+ "loss": 0.93,
1863
+ "step": 305
1864
+ },
1865
+ {
1866
+ "epoch": 17.24,
1867
+ "learning_rate": 1.2982092741250145e-06,
1868
+ "loss": 0.9181,
1869
+ "step": 306
1870
+ },
1871
+ {
1872
+ "epoch": 17.3,
1873
+ "learning_rate": 1.2235870926211619e-06,
1874
+ "loss": 0.8043,
1875
+ "step": 307
1876
+ },
1877
+ {
1878
+ "epoch": 17.35,
1879
+ "learning_rate": 1.1511197505771843e-06,
1880
+ "loss": 0.8552,
1881
+ "step": 308
1882
+ },
1883
+ {
1884
+ "epoch": 17.41,
1885
+ "learning_rate": 1.0808138156569614e-06,
1886
+ "loss": 0.8864,
1887
+ "step": 309
1888
+ },
1889
+ {
1890
+ "epoch": 17.46,
1891
+ "learning_rate": 1.0126756596375686e-06,
1892
+ "loss": 0.8856,
1893
+ "step": 310
1894
+ },
1895
+ {
1896
+ "epoch": 17.52,
1897
+ "learning_rate": 9.467114578318226e-07,
1898
+ "loss": 0.9447,
1899
+ "step": 311
1900
+ },
1901
+ {
1902
+ "epoch": 17.58,
1903
+ "learning_rate": 8.829271885286094e-07,
1904
+ "loss": 0.8091,
1905
+ "step": 312
1906
+ },
1907
+ {
1908
+ "epoch": 17.63,
1909
+ "learning_rate": 8.213286324510738e-07,
1910
+ "loss": 0.9956,
1911
+ "step": 313
1912
+ },
1913
+ {
1914
+ "epoch": 17.69,
1915
+ "learning_rate": 7.619213722327185e-07,
1916
+ "loss": 0.88,
1917
+ "step": 314
1918
+ },
1919
+ {
1920
+ "epoch": 17.75,
1921
+ "learning_rate": 7.047107919114588e-07,
1922
+ "loss": 0.9142,
1923
+ "step": 315
1924
+ },
1925
+ {
1926
+ "epoch": 17.8,
1927
+ "learning_rate": 6.497020764416633e-07,
1928
+ "loss": 0.9984,
1929
+ "step": 316
1930
+ },
1931
+ {
1932
+ "epoch": 17.86,
1933
+ "learning_rate": 5.969002112242494e-07,
1934
+ "loss": 0.8869,
1935
+ "step": 317
1936
+ },
1937
+ {
1938
+ "epoch": 17.92,
1939
+ "learning_rate": 5.463099816548579e-07,
1940
+ "loss": 0.9627,
1941
+ "step": 318
1942
+ },
1943
+ {
1944
+ "epoch": 17.97,
1945
+ "learning_rate": 4.979359726901639e-07,
1946
+ "loss": 0.9783,
1947
+ "step": 319
1948
+ },
1949
+ {
1950
+ "epoch": 18.03,
1951
+ "learning_rate": 4.517825684323324e-07,
1952
+ "loss": 1.0067,
1953
+ "step": 320
1954
+ },
1955
+ {
1956
+ "epoch": 18.08,
1957
+ "learning_rate": 4.078539517316987e-07,
1958
+ "loss": 0.9261,
1959
+ "step": 321
1960
+ },
1961
+ {
1962
+ "epoch": 18.14,
1963
+ "learning_rate": 3.6615410380767544e-07,
1964
+ "loss": 1.0504,
1965
+ "step": 322
1966
+ },
1967
+ {
1968
+ "epoch": 18.2,
1969
+ "learning_rate": 3.266868038879434e-07,
1970
+ "loss": 0.8995,
1971
+ "step": 323
1972
+ },
1973
+ {
1974
+ "epoch": 18.25,
1975
+ "learning_rate": 2.894556288659395e-07,
1976
+ "loss": 0.915,
1977
+ "step": 324
1978
+ },
1979
+ {
1980
+ "epoch": 18.31,
1981
+ "learning_rate": 2.544639529766829e-07,
1982
+ "loss": 0.8657,
1983
+ "step": 325
1984
+ },
1985
+ {
1986
+ "epoch": 18.37,
1987
+ "learning_rate": 2.2171494749097245e-07,
1988
+ "loss": 0.9171,
1989
+ "step": 326
1990
+ },
1991
+ {
1992
+ "epoch": 18.42,
1993
+ "learning_rate": 1.912115804279746e-07,
1994
+ "loss": 0.7903,
1995
+ "step": 327
1996
+ },
1997
+ {
1998
+ "epoch": 18.48,
1999
+ "learning_rate": 1.6295661628624447e-07,
2000
+ "loss": 0.9827,
2001
+ "step": 328
2002
+ },
2003
+ {
2004
+ "epoch": 18.54,
2005
+ "learning_rate": 1.3695261579316777e-07,
2006
+ "loss": 0.9373,
2007
+ "step": 329
2008
+ },
2009
+ {
2010
+ "epoch": 18.59,
2011
+ "learning_rate": 1.1320193567288529e-07,
2012
+ "loss": 0.915,
2013
+ "step": 330
2014
+ },
2015
+ {
2016
+ "epoch": 18.65,
2017
+ "learning_rate": 9.170672843271666e-08,
2018
+ "loss": 1.0085,
2019
+ "step": 331
2020
+ },
2021
+ {
2022
+ "epoch": 18.7,
2023
+ "learning_rate": 7.246894216806355e-08,
2024
+ "loss": 0.9681,
2025
+ "step": 332
2026
+ },
2027
+ {
2028
+ "epoch": 18.76,
2029
+ "learning_rate": 5.54903203858731e-08,
2030
+ "loss": 0.8392,
2031
+ "step": 333
2032
+ },
2033
+ {
2034
+ "epoch": 18.82,
2035
+ "learning_rate": 4.07724018466088e-08,
2036
+ "loss": 0.9473,
2037
+ "step": 334
2038
+ },
2039
+ {
2040
+ "epoch": 18.87,
2041
+ "learning_rate": 2.831652042480093e-08,
2042
+ "loss": 0.9534,
2043
+ "step": 335
2044
+ },
2045
+ {
2046
+ "epoch": 18.93,
2047
+ "learning_rate": 1.812380498815991e-08,
2048
+ "loss": 0.9152,
2049
+ "step": 336
2050
+ },
2051
+ {
2052
+ "epoch": 18.99,
2053
+ "learning_rate": 1.0195179295269252e-08,
2054
+ "loss": 0.8761,
2055
+ "step": 337
2056
+ }
2057
+ ],
2058
+ "logging_steps": 1,
2059
+ "max_steps": 340,
2060
+ "num_train_epochs": 20,
2061
+ "save_steps": 500,
2062
+ "total_flos": 4.219147109597184e+17,
2063
+ "trial_name": null,
2064
+ "trial_params": null
2065
+ }
checkpoint-337/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af7bb882cac5b4d33f676aa38a935fcc9173c63dc3da717a69a225490ea5e0f7
3
+ size 4155
checkpoint-340/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0.dev0
checkpoint-340/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
4
+ "bias": "none",
5
+ "fan_in_fan_out": null,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 32,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "k_proj",
18
+ "gate_proj",
19
+ "v_proj",
20
+ "o_proj",
21
+ "q_proj",
22
+ "down_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
checkpoint-340/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c36d31ee32ee247944f1ebfb0ae67945a2bc9ac80545e7536bf0a970a3c1455
3
+ size 500897101
checkpoint-340/adapter_model/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0.dev0
checkpoint-340/adapter_model/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "NousResearch/Llama-2-13b-hf",
4
+ "bias": "none",
5
+ "fan_in_fan_out": null,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 32,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "k_proj",
18
+ "gate_proj",
19
+ "v_proj",
20
+ "o_proj",
21
+ "q_proj",
22
+ "down_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
checkpoint-340/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c36d31ee32ee247944f1ebfb0ae67945a2bc9ac80545e7536bf0a970a3c1455
3
+ size 500897101
checkpoint-340/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21f19801f9797742329fb21864ab0b303d2ec7e391ab13d05006f5f2ba0d2b88
3
+ size 1001737021
checkpoint-340/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bb250ab75b2eb528aac5a23e727bb19acb7d54b9dfdff217b08b555706614ff8
3
+ size 14575
checkpoint-340/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4b6fb4780c24f9911d5783eac25a5398071fb05b3676c5a1d721cb2c2d73528f
3
+ size 627
checkpoint-340/trainer_state.json ADDED
@@ -0,0 +1,2083 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 19.154929577464788,
5
+ "eval_steps": 100,
6
+ "global_step": 340,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.06,
13
+ "learning_rate": 5e-06,
14
+ "loss": 1.5896,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.11,
19
+ "learning_rate": 1e-05,
20
+ "loss": 1.6521,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.17,
25
+ "learning_rate": 1.5e-05,
26
+ "loss": 1.3654,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.23,
31
+ "learning_rate": 2e-05,
32
+ "loss": 1.3569,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.28,
37
+ "learning_rate": 2.5e-05,
38
+ "loss": 1.4352,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.34,
43
+ "learning_rate": 3e-05,
44
+ "loss": 1.425,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.39,
49
+ "learning_rate": 3.5e-05,
50
+ "loss": 1.3529,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.45,
55
+ "learning_rate": 4e-05,
56
+ "loss": 1.384,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.51,
61
+ "learning_rate": 4.5e-05,
62
+ "loss": 1.3881,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.56,
67
+ "learning_rate": 5e-05,
68
+ "loss": 1.5593,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.62,
73
+ "learning_rate": 4.999886713385432e-05,
74
+ "loss": 1.3423,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.68,
79
+ "learning_rate": 4.999546863808815e-05,
80
+ "loss": 1.4725,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.73,
85
+ "learning_rate": 4.9989804820704735e-05,
86
+ "loss": 1.4342,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.79,
91
+ "learning_rate": 4.9981876195011844e-05,
92
+ "loss": 1.3769,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.85,
97
+ "learning_rate": 4.99716834795752e-05,
98
+ "loss": 1.3976,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.9,
103
+ "learning_rate": 4.995922759815339e-05,
104
+ "loss": 1.328,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.96,
109
+ "learning_rate": 4.994450967961413e-05,
110
+ "loss": 1.266,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 1.01,
115
+ "learning_rate": 4.992753105783194e-05,
116
+ "loss": 1.4671,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 1.07,
121
+ "learning_rate": 4.9908293271567286e-05,
122
+ "loss": 1.3973,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 1.13,
127
+ "learning_rate": 4.988679806432712e-05,
128
+ "loss": 1.3835,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 1.18,
133
+ "learning_rate": 4.9863047384206835e-05,
134
+ "loss": 1.2203,
135
+ "step": 21
136
+ },
137
+ {
138
+ "epoch": 1.24,
139
+ "learning_rate": 4.9837043383713753e-05,
140
+ "loss": 1.3752,
141
+ "step": 22
142
+ },
143
+ {
144
+ "epoch": 1.3,
145
+ "learning_rate": 4.980878841957203e-05,
146
+ "loss": 1.4142,
147
+ "step": 23
148
+ },
149
+ {
150
+ "epoch": 1.35,
151
+ "learning_rate": 4.977828505250903e-05,
152
+ "loss": 1.3657,
153
+ "step": 24
154
+ },
155
+ {
156
+ "epoch": 1.41,
157
+ "learning_rate": 4.9745536047023324e-05,
158
+ "loss": 1.309,
159
+ "step": 25
160
+ },
161
+ {
162
+ "epoch": 1.46,
163
+ "learning_rate": 4.971054437113406e-05,
164
+ "loss": 1.3004,
165
+ "step": 26
166
+ },
167
+ {
168
+ "epoch": 1.52,
169
+ "learning_rate": 4.967331319611206e-05,
170
+ "loss": 1.3309,
171
+ "step": 27
172
+ },
173
+ {
174
+ "epoch": 1.58,
175
+ "learning_rate": 4.963384589619233e-05,
176
+ "loss": 1.4956,
177
+ "step": 28
178
+ },
179
+ {
180
+ "epoch": 1.63,
181
+ "learning_rate": 4.959214604826831e-05,
182
+ "loss": 1.2682,
183
+ "step": 29
184
+ },
185
+ {
186
+ "epoch": 1.69,
187
+ "learning_rate": 4.9548217431567665e-05,
188
+ "loss": 1.391,
189
+ "step": 30
190
+ },
191
+ {
192
+ "epoch": 1.75,
193
+ "learning_rate": 4.9502064027309836e-05,
194
+ "loss": 1.3102,
195
+ "step": 31
196
+ },
197
+ {
198
+ "epoch": 1.8,
199
+ "learning_rate": 4.9453690018345144e-05,
200
+ "loss": 1.3891,
201
+ "step": 32
202
+ },
203
+ {
204
+ "epoch": 1.86,
205
+ "learning_rate": 4.9403099788775754e-05,
206
+ "loss": 1.3538,
207
+ "step": 33
208
+ },
209
+ {
210
+ "epoch": 1.92,
211
+ "learning_rate": 4.935029792355834e-05,
212
+ "loss": 1.3443,
213
+ "step": 34
214
+ },
215
+ {
216
+ "epoch": 1.97,
217
+ "learning_rate": 4.929528920808854e-05,
218
+ "loss": 1.3384,
219
+ "step": 35
220
+ },
221
+ {
222
+ "epoch": 2.03,
223
+ "learning_rate": 4.923807862776728e-05,
224
+ "loss": 1.2968,
225
+ "step": 36
226
+ },
227
+ {
228
+ "epoch": 2.08,
229
+ "learning_rate": 4.917867136754893e-05,
230
+ "loss": 1.2624,
231
+ "step": 37
232
+ },
233
+ {
234
+ "epoch": 2.14,
235
+ "learning_rate": 4.91170728114714e-05,
236
+ "loss": 1.3547,
237
+ "step": 38
238
+ },
239
+ {
240
+ "epoch": 2.2,
241
+ "learning_rate": 4.9053288542168185e-05,
242
+ "loss": 1.3857,
243
+ "step": 39
244
+ },
245
+ {
246
+ "epoch": 2.25,
247
+ "learning_rate": 4.898732434036244e-05,
248
+ "loss": 1.2261,
249
+ "step": 40
250
+ },
251
+ {
252
+ "epoch": 2.31,
253
+ "learning_rate": 4.8919186184343046e-05,
254
+ "loss": 1.4072,
255
+ "step": 41
256
+ },
257
+ {
258
+ "epoch": 2.37,
259
+ "learning_rate": 4.8848880249422815e-05,
260
+ "loss": 1.3283,
261
+ "step": 42
262
+ },
263
+ {
264
+ "epoch": 2.42,
265
+ "learning_rate": 4.877641290737884e-05,
266
+ "loss": 1.2464,
267
+ "step": 43
268
+ },
269
+ {
270
+ "epoch": 2.48,
271
+ "learning_rate": 4.870179072587499e-05,
272
+ "loss": 1.3433,
273
+ "step": 44
274
+ },
275
+ {
276
+ "epoch": 2.54,
277
+ "learning_rate": 4.862502046786671e-05,
278
+ "loss": 1.2861,
279
+ "step": 45
280
+ },
281
+ {
282
+ "epoch": 2.59,
283
+ "learning_rate": 4.854610909098812e-05,
284
+ "loss": 1.1888,
285
+ "step": 46
286
+ },
287
+ {
288
+ "epoch": 2.65,
289
+ "learning_rate": 4.8465063746921395e-05,
290
+ "loss": 1.2509,
291
+ "step": 47
292
+ },
293
+ {
294
+ "epoch": 2.7,
295
+ "learning_rate": 4.838189178074867e-05,
296
+ "loss": 1.4449,
297
+ "step": 48
298
+ },
299
+ {
300
+ "epoch": 2.76,
301
+ "learning_rate": 4.829660073028631e-05,
302
+ "loss": 1.4645,
303
+ "step": 49
304
+ },
305
+ {
306
+ "epoch": 2.82,
307
+ "learning_rate": 4.8209198325401815e-05,
308
+ "loss": 1.3586,
309
+ "step": 50
310
+ },
311
+ {
312
+ "epoch": 2.87,
313
+ "learning_rate": 4.811969248731323e-05,
314
+ "loss": 1.3542,
315
+ "step": 51
316
+ },
317
+ {
318
+ "epoch": 2.93,
319
+ "learning_rate": 4.802809132787125e-05,
320
+ "loss": 1.3145,
321
+ "step": 52
322
+ },
323
+ {
324
+ "epoch": 2.99,
325
+ "learning_rate": 4.793440314882408e-05,
326
+ "loss": 1.2343,
327
+ "step": 53
328
+ },
329
+ {
330
+ "epoch": 3.04,
331
+ "learning_rate": 4.783863644106502e-05,
332
+ "loss": 1.2946,
333
+ "step": 54
334
+ },
335
+ {
336
+ "epoch": 3.1,
337
+ "learning_rate": 4.774079988386296e-05,
338
+ "loss": 1.2621,
339
+ "step": 55
340
+ },
341
+ {
342
+ "epoch": 3.15,
343
+ "learning_rate": 4.764090234407577e-05,
344
+ "loss": 1.3581,
345
+ "step": 56
346
+ },
347
+ {
348
+ "epoch": 3.21,
349
+ "learning_rate": 4.753895287534673e-05,
350
+ "loss": 1.3891,
351
+ "step": 57
352
+ },
353
+ {
354
+ "epoch": 3.27,
355
+ "learning_rate": 4.743496071728396e-05,
356
+ "loss": 1.334,
357
+ "step": 58
358
+ },
359
+ {
360
+ "epoch": 3.32,
361
+ "learning_rate": 4.73289352946231e-05,
362
+ "loss": 1.3105,
363
+ "step": 59
364
+ },
365
+ {
366
+ "epoch": 3.38,
367
+ "learning_rate": 4.722088621637309e-05,
368
+ "loss": 1.2921,
369
+ "step": 60
370
+ },
371
+ {
372
+ "epoch": 3.44,
373
+ "learning_rate": 4.711082327494536e-05,
374
+ "loss": 1.2406,
375
+ "step": 61
376
+ },
377
+ {
378
+ "epoch": 3.49,
379
+ "learning_rate": 4.6998756445266336e-05,
380
+ "loss": 1.4246,
381
+ "step": 62
382
+ },
383
+ {
384
+ "epoch": 3.55,
385
+ "learning_rate": 4.688469588387339e-05,
386
+ "loss": 1.2126,
387
+ "step": 63
388
+ },
389
+ {
390
+ "epoch": 3.61,
391
+ "learning_rate": 4.6768651927994434e-05,
392
+ "loss": 1.3662,
393
+ "step": 64
394
+ },
395
+ {
396
+ "epoch": 3.66,
397
+ "learning_rate": 4.665063509461097e-05,
398
+ "loss": 1.2044,
399
+ "step": 65
400
+ },
401
+ {
402
+ "epoch": 3.72,
403
+ "learning_rate": 4.653065607950502e-05,
404
+ "loss": 1.4344,
405
+ "step": 66
406
+ },
407
+ {
408
+ "epoch": 3.77,
409
+ "learning_rate": 4.640872575628973e-05,
410
+ "loss": 1.3652,
411
+ "step": 67
412
+ },
413
+ {
414
+ "epoch": 3.83,
415
+ "learning_rate": 4.628485517542392e-05,
416
+ "loss": 1.231,
417
+ "step": 68
418
+ },
419
+ {
420
+ "epoch": 3.89,
421
+ "learning_rate": 4.6159055563210604e-05,
422
+ "loss": 1.1948,
423
+ "step": 69
424
+ },
425
+ {
426
+ "epoch": 3.94,
427
+ "learning_rate": 4.6031338320779534e-05,
428
+ "loss": 1.2484,
429
+ "step": 70
430
+ },
431
+ {
432
+ "epoch": 4.0,
433
+ "learning_rate": 4.59017150230539e-05,
434
+ "loss": 1.2968,
435
+ "step": 71
436
+ },
437
+ {
438
+ "epoch": 4.06,
439
+ "learning_rate": 4.5770197417701365e-05,
440
+ "loss": 1.1691,
441
+ "step": 72
442
+ },
443
+ {
444
+ "epoch": 4.11,
445
+ "learning_rate": 4.563679742406935e-05,
446
+ "loss": 1.224,
447
+ "step": 73
448
+ },
449
+ {
450
+ "epoch": 4.17,
451
+ "learning_rate": 4.550152713210478e-05,
452
+ "loss": 1.2669,
453
+ "step": 74
454
+ },
455
+ {
456
+ "epoch": 4.23,
457
+ "learning_rate": 4.5364398801258396e-05,
458
+ "loss": 1.3218,
459
+ "step": 75
460
+ },
461
+ {
462
+ "epoch": 4.28,
463
+ "learning_rate": 4.522542485937369e-05,
464
+ "loss": 1.3309,
465
+ "step": 76
466
+ },
467
+ {
468
+ "epoch": 4.34,
469
+ "learning_rate": 4.508461790156056e-05,
470
+ "loss": 1.3356,
471
+ "step": 77
472
+ },
473
+ {
474
+ "epoch": 4.39,
475
+ "learning_rate": 4.4941990689053886e-05,
476
+ "loss": 1.3367,
477
+ "step": 78
478
+ },
479
+ {
480
+ "epoch": 4.45,
481
+ "learning_rate": 4.479755614805688e-05,
482
+ "loss": 1.3568,
483
+ "step": 79
484
+ },
485
+ {
486
+ "epoch": 4.51,
487
+ "learning_rate": 4.465132736856969e-05,
488
+ "loss": 1.29,
489
+ "step": 80
490
+ },
491
+ {
492
+ "epoch": 4.56,
493
+ "learning_rate": 4.450331760320302e-05,
494
+ "loss": 1.3522,
495
+ "step": 81
496
+ },
497
+ {
498
+ "epoch": 4.62,
499
+ "learning_rate": 4.4353540265977064e-05,
500
+ "loss": 1.3021,
501
+ "step": 82
502
+ },
503
+ {
504
+ "epoch": 4.68,
505
+ "learning_rate": 4.4202008931105795e-05,
506
+ "loss": 1.303,
507
+ "step": 83
508
+ },
509
+ {
510
+ "epoch": 4.73,
511
+ "learning_rate": 4.404873733176678e-05,
512
+ "loss": 1.313,
513
+ "step": 84
514
+ },
515
+ {
516
+ "epoch": 4.79,
517
+ "learning_rate": 4.389373935885646e-05,
518
+ "loss": 1.231,
519
+ "step": 85
520
+ },
521
+ {
522
+ "epoch": 4.85,
523
+ "learning_rate": 4.373702905973135e-05,
524
+ "loss": 1.2908,
525
+ "step": 86
526
+ },
527
+ {
528
+ "epoch": 4.9,
529
+ "learning_rate": 4.357862063693486e-05,
530
+ "loss": 1.2575,
531
+ "step": 87
532
+ },
533
+ {
534
+ "epoch": 4.96,
535
+ "learning_rate": 4.341852844691012e-05,
536
+ "loss": 1.1942,
537
+ "step": 88
538
+ },
539
+ {
540
+ "epoch": 5.01,
541
+ "learning_rate": 4.3256766998698936e-05,
542
+ "loss": 1.1607,
543
+ "step": 89
544
+ },
545
+ {
546
+ "epoch": 5.07,
547
+ "learning_rate": 4.309335095262676e-05,
548
+ "loss": 1.3386,
549
+ "step": 90
550
+ },
551
+ {
552
+ "epoch": 5.13,
553
+ "learning_rate": 4.292829511897409e-05,
554
+ "loss": 1.371,
555
+ "step": 91
556
+ },
557
+ {
558
+ "epoch": 5.18,
559
+ "learning_rate": 4.276161445663423e-05,
560
+ "loss": 1.236,
561
+ "step": 92
562
+ },
563
+ {
564
+ "epoch": 5.24,
565
+ "learning_rate": 4.259332407175751e-05,
566
+ "loss": 1.2531,
567
+ "step": 93
568
+ },
569
+ {
570
+ "epoch": 5.3,
571
+ "learning_rate": 4.242343921638234e-05,
572
+ "loss": 1.2593,
573
+ "step": 94
574
+ },
575
+ {
576
+ "epoch": 5.35,
577
+ "learning_rate": 4.2251975287052804e-05,
578
+ "loss": 1.1245,
579
+ "step": 95
580
+ },
581
+ {
582
+ "epoch": 5.41,
583
+ "learning_rate": 4.2078947823423364e-05,
584
+ "loss": 1.2538,
585
+ "step": 96
586
+ },
587
+ {
588
+ "epoch": 5.46,
589
+ "learning_rate": 4.1904372506850484e-05,
590
+ "loss": 1.2847,
591
+ "step": 97
592
+ },
593
+ {
594
+ "epoch": 5.52,
595
+ "learning_rate": 4.172826515897146e-05,
596
+ "loss": 1.2071,
597
+ "step": 98
598
+ },
599
+ {
600
+ "epoch": 5.58,
601
+ "learning_rate": 4.155064174027047e-05,
602
+ "loss": 1.2664,
603
+ "step": 99
604
+ },
605
+ {
606
+ "epoch": 5.63,
607
+ "learning_rate": 4.137151834863213e-05,
608
+ "loss": 1.1919,
609
+ "step": 100
610
+ },
611
+ {
612
+ "epoch": 5.63,
613
+ "eval_loss": 1.2261024713516235,
614
+ "eval_runtime": 1.4353,
615
+ "eval_samples_per_second": 6.967,
616
+ "eval_steps_per_second": 3.484,
617
+ "step": 100
618
+ },
619
+ {
620
+ "epoch": 5.69,
621
+ "learning_rate": 4.119091121788256e-05,
622
+ "loss": 1.1509,
623
+ "step": 101
624
+ },
625
+ {
626
+ "epoch": 5.75,
627
+ "learning_rate": 4.100883671631806e-05,
628
+ "loss": 1.3013,
629
+ "step": 102
630
+ },
631
+ {
632
+ "epoch": 5.8,
633
+ "learning_rate": 4.082531134522176e-05,
634
+ "loss": 1.313,
635
+ "step": 103
636
+ },
637
+ {
638
+ "epoch": 5.86,
639
+ "learning_rate": 4.064035173736804e-05,
640
+ "loss": 1.2176,
641
+ "step": 104
642
+ },
643
+ {
644
+ "epoch": 5.92,
645
+ "learning_rate": 4.045397465551513e-05,
646
+ "loss": 1.1004,
647
+ "step": 105
648
+ },
649
+ {
650
+ "epoch": 5.97,
651
+ "learning_rate": 4.0266196990885955e-05,
652
+ "loss": 1.3575,
653
+ "step": 106
654
+ },
655
+ {
656
+ "epoch": 6.03,
657
+ "learning_rate": 4.007703576163724e-05,
658
+ "loss": 1.2923,
659
+ "step": 107
660
+ },
661
+ {
662
+ "epoch": 6.08,
663
+ "learning_rate": 3.98865081113172e-05,
664
+ "loss": 1.2083,
665
+ "step": 108
666
+ },
667
+ {
668
+ "epoch": 6.14,
669
+ "learning_rate": 3.969463130731183e-05,
670
+ "loss": 1.2141,
671
+ "step": 109
672
+ },
673
+ {
674
+ "epoch": 6.2,
675
+ "learning_rate": 3.9501422739279956e-05,
676
+ "loss": 1.3297,
677
+ "step": 110
678
+ },
679
+ {
680
+ "epoch": 6.25,
681
+ "learning_rate": 3.9306899917577245e-05,
682
+ "loss": 1.0901,
683
+ "step": 111
684
+ },
685
+ {
686
+ "epoch": 6.31,
687
+ "learning_rate": 3.911108047166924e-05,
688
+ "loss": 1.2235,
689
+ "step": 112
690
+ },
691
+ {
692
+ "epoch": 6.37,
693
+ "learning_rate": 3.89139821485336e-05,
694
+ "loss": 1.341,
695
+ "step": 113
696
+ },
697
+ {
698
+ "epoch": 6.42,
699
+ "learning_rate": 3.871562281105175e-05,
700
+ "loss": 1.362,
701
+ "step": 114
702
+ },
703
+ {
704
+ "epoch": 6.48,
705
+ "learning_rate": 3.851602043638994e-05,
706
+ "loss": 1.2738,
707
+ "step": 115
708
+ },
709
+ {
710
+ "epoch": 6.54,
711
+ "learning_rate": 3.8315193114369996e-05,
712
+ "loss": 1.2264,
713
+ "step": 116
714
+ },
715
+ {
716
+ "epoch": 6.59,
717
+ "learning_rate": 3.8113159045829864e-05,
718
+ "loss": 1.171,
719
+ "step": 117
720
+ },
721
+ {
722
+ "epoch": 6.65,
723
+ "learning_rate": 3.790993654097405e-05,
724
+ "loss": 1.1848,
725
+ "step": 118
726
+ },
727
+ {
728
+ "epoch": 6.7,
729
+ "learning_rate": 3.770554401771423e-05,
730
+ "loss": 1.1545,
731
+ "step": 119
732
+ },
733
+ {
734
+ "epoch": 6.76,
735
+ "learning_rate": 3.7500000000000003e-05,
736
+ "loss": 1.2799,
737
+ "step": 120
738
+ },
739
+ {
740
+ "epoch": 6.82,
741
+ "learning_rate": 3.72933231161401e-05,
742
+ "loss": 1.2278,
743
+ "step": 121
744
+ },
745
+ {
746
+ "epoch": 6.87,
747
+ "learning_rate": 3.708553209711409e-05,
748
+ "loss": 1.1959,
749
+ "step": 122
750
+ },
751
+ {
752
+ "epoch": 6.93,
753
+ "learning_rate": 3.6876645774874876e-05,
754
+ "loss": 1.1151,
755
+ "step": 123
756
+ },
757
+ {
758
+ "epoch": 6.99,
759
+ "learning_rate": 3.6666683080641846e-05,
760
+ "loss": 1.2178,
761
+ "step": 124
762
+ },
763
+ {
764
+ "epoch": 7.04,
765
+ "learning_rate": 3.645566304318526e-05,
766
+ "loss": 1.1584,
767
+ "step": 125
768
+ },
769
+ {
770
+ "epoch": 7.1,
771
+ "learning_rate": 3.624360478710165e-05,
772
+ "loss": 1.0959,
773
+ "step": 126
774
+ },
775
+ {
776
+ "epoch": 7.15,
777
+ "learning_rate": 3.603052753108053e-05,
778
+ "loss": 1.3026,
779
+ "step": 127
780
+ },
781
+ {
782
+ "epoch": 7.21,
783
+ "learning_rate": 3.581645058616271e-05,
784
+ "loss": 1.2007,
785
+ "step": 128
786
+ },
787
+ {
788
+ "epoch": 7.27,
789
+ "learning_rate": 3.5601393353990046e-05,
790
+ "loss": 1.2242,
791
+ "step": 129
792
+ },
793
+ {
794
+ "epoch": 7.32,
795
+ "learning_rate": 3.5385375325047166e-05,
796
+ "loss": 1.2737,
797
+ "step": 130
798
+ },
799
+ {
800
+ "epoch": 7.38,
801
+ "learning_rate": 3.516841607689501e-05,
802
+ "loss": 1.2724,
803
+ "step": 131
804
+ },
805
+ {
806
+ "epoch": 7.44,
807
+ "learning_rate": 3.495053527239656e-05,
808
+ "loss": 1.1916,
809
+ "step": 132
810
+ },
811
+ {
812
+ "epoch": 7.49,
813
+ "learning_rate": 3.4731752657934794e-05,
814
+ "loss": 1.2107,
815
+ "step": 133
816
+ },
817
+ {
818
+ "epoch": 7.55,
819
+ "learning_rate": 3.4512088061623075e-05,
820
+ "loss": 1.246,
821
+ "step": 134
822
+ },
823
+ {
824
+ "epoch": 7.61,
825
+ "learning_rate": 3.4291561391508185e-05,
826
+ "loss": 1.1841,
827
+ "step": 135
828
+ },
829
+ {
830
+ "epoch": 7.66,
831
+ "learning_rate": 3.4070192633766025e-05,
832
+ "loss": 1.1898,
833
+ "step": 136
834
+ },
835
+ {
836
+ "epoch": 7.72,
837
+ "learning_rate": 3.3848001850890296e-05,
838
+ "loss": 1.1195,
839
+ "step": 137
840
+ },
841
+ {
842
+ "epoch": 7.77,
843
+ "learning_rate": 3.3625009179874265e-05,
844
+ "loss": 1.1685,
845
+ "step": 138
846
+ },
847
+ {
848
+ "epoch": 7.83,
849
+ "learning_rate": 3.3401234830385756e-05,
850
+ "loss": 1.2748,
851
+ "step": 139
852
+ },
853
+ {
854
+ "epoch": 7.89,
855
+ "learning_rate": 3.3176699082935545e-05,
856
+ "loss": 1.1697,
857
+ "step": 140
858
+ },
859
+ {
860
+ "epoch": 7.94,
861
+ "learning_rate": 3.295142228703938e-05,
862
+ "loss": 1.1624,
863
+ "step": 141
864
+ },
865
+ {
866
+ "epoch": 8.0,
867
+ "learning_rate": 3.272542485937369e-05,
868
+ "loss": 1.1679,
869
+ "step": 142
870
+ },
871
+ {
872
+ "epoch": 8.06,
873
+ "learning_rate": 3.249872728192527e-05,
874
+ "loss": 1.1163,
875
+ "step": 143
876
+ },
877
+ {
878
+ "epoch": 8.11,
879
+ "learning_rate": 3.2271350100134975e-05,
880
+ "loss": 1.1319,
881
+ "step": 144
882
+ },
883
+ {
884
+ "epoch": 8.17,
885
+ "learning_rate": 3.2043313921035743e-05,
886
+ "loss": 1.2151,
887
+ "step": 145
888
+ },
889
+ {
890
+ "epoch": 8.23,
891
+ "learning_rate": 3.181463941138495e-05,
892
+ "loss": 1.1339,
893
+ "step": 146
894
+ },
895
+ {
896
+ "epoch": 8.28,
897
+ "learning_rate": 3.158534729579142e-05,
898
+ "loss": 1.1503,
899
+ "step": 147
900
+ },
901
+ {
902
+ "epoch": 8.34,
903
+ "learning_rate": 3.135545835483718e-05,
904
+ "loss": 1.2326,
905
+ "step": 148
906
+ },
907
+ {
908
+ "epoch": 8.39,
909
+ "learning_rate": 3.11249934231941e-05,
910
+ "loss": 1.2938,
911
+ "step": 149
912
+ },
913
+ {
914
+ "epoch": 8.45,
915
+ "learning_rate": 3.0893973387735687e-05,
916
+ "loss": 1.1454,
917
+ "step": 150
918
+ },
919
+ {
920
+ "epoch": 8.51,
921
+ "learning_rate": 3.0662419185644115e-05,
922
+ "loss": 1.2504,
923
+ "step": 151
924
+ },
925
+ {
926
+ "epoch": 8.56,
927
+ "learning_rate": 3.0430351802512698e-05,
928
+ "loss": 0.987,
929
+ "step": 152
930
+ },
931
+ {
932
+ "epoch": 8.62,
933
+ "learning_rate": 3.0197792270443982e-05,
934
+ "loss": 1.1387,
935
+ "step": 153
936
+ },
937
+ {
938
+ "epoch": 8.68,
939
+ "learning_rate": 2.996476166614364e-05,
940
+ "loss": 1.1142,
941
+ "step": 154
942
+ },
943
+ {
944
+ "epoch": 8.73,
945
+ "learning_rate": 2.9731281109010256e-05,
946
+ "loss": 1.1714,
947
+ "step": 155
948
+ },
949
+ {
950
+ "epoch": 8.79,
951
+ "learning_rate": 2.9497371759221347e-05,
952
+ "loss": 1.1858,
953
+ "step": 156
954
+ },
955
+ {
956
+ "epoch": 8.85,
957
+ "learning_rate": 2.92630548158156e-05,
958
+ "loss": 1.2169,
959
+ "step": 157
960
+ },
961
+ {
962
+ "epoch": 8.9,
963
+ "learning_rate": 2.9028351514771606e-05,
964
+ "loss": 1.2204,
965
+ "step": 158
966
+ },
967
+ {
968
+ "epoch": 8.96,
969
+ "learning_rate": 2.8793283127083292e-05,
970
+ "loss": 1.0313,
971
+ "step": 159
972
+ },
973
+ {
974
+ "epoch": 9.01,
975
+ "learning_rate": 2.8557870956832132e-05,
976
+ "loss": 1.2432,
977
+ "step": 160
978
+ },
979
+ {
980
+ "epoch": 9.07,
981
+ "learning_rate": 2.8322136339256356e-05,
982
+ "loss": 1.0614,
983
+ "step": 161
984
+ },
985
+ {
986
+ "epoch": 9.13,
987
+ "learning_rate": 2.808610063881737e-05,
988
+ "loss": 1.1682,
989
+ "step": 162
990
+ },
991
+ {
992
+ "epoch": 9.18,
993
+ "learning_rate": 2.7849785247263515e-05,
994
+ "loss": 1.0238,
995
+ "step": 163
996
+ },
997
+ {
998
+ "epoch": 9.24,
999
+ "learning_rate": 2.761321158169134e-05,
1000
+ "loss": 1.166,
1001
+ "step": 164
1002
+ },
1003
+ {
1004
+ "epoch": 9.3,
1005
+ "learning_rate": 2.7376401082604564e-05,
1006
+ "loss": 1.0965,
1007
+ "step": 165
1008
+ },
1009
+ {
1010
+ "epoch": 9.35,
1011
+ "learning_rate": 2.7139375211970996e-05,
1012
+ "loss": 1.096,
1013
+ "step": 166
1014
+ },
1015
+ {
1016
+ "epoch": 9.41,
1017
+ "learning_rate": 2.6902155451277377e-05,
1018
+ "loss": 1.1161,
1019
+ "step": 167
1020
+ },
1021
+ {
1022
+ "epoch": 9.46,
1023
+ "learning_rate": 2.6664763299582602e-05,
1024
+ "loss": 1.1328,
1025
+ "step": 168
1026
+ },
1027
+ {
1028
+ "epoch": 9.52,
1029
+ "learning_rate": 2.6427220271569203e-05,
1030
+ "loss": 1.0747,
1031
+ "step": 169
1032
+ },
1033
+ {
1034
+ "epoch": 9.58,
1035
+ "learning_rate": 2.6189547895593562e-05,
1036
+ "loss": 1.1092,
1037
+ "step": 170
1038
+ },
1039
+ {
1040
+ "epoch": 9.63,
1041
+ "learning_rate": 2.5951767711734753e-05,
1042
+ "loss": 1.1624,
1043
+ "step": 171
1044
+ },
1045
+ {
1046
+ "epoch": 9.69,
1047
+ "learning_rate": 2.5713901269842404e-05,
1048
+ "loss": 1.1434,
1049
+ "step": 172
1050
+ },
1051
+ {
1052
+ "epoch": 9.75,
1053
+ "learning_rate": 2.5475970127583666e-05,
1054
+ "loss": 1.1167,
1055
+ "step": 173
1056
+ },
1057
+ {
1058
+ "epoch": 9.8,
1059
+ "learning_rate": 2.523799584848942e-05,
1060
+ "loss": 1.0509,
1061
+ "step": 174
1062
+ },
1063
+ {
1064
+ "epoch": 9.86,
1065
+ "learning_rate": 2.5e-05,
1066
+ "loss": 1.1449,
1067
+ "step": 175
1068
+ },
1069
+ {
1070
+ "epoch": 9.92,
1071
+ "learning_rate": 2.4762004151510584e-05,
1072
+ "loss": 1.109,
1073
+ "step": 176
1074
+ },
1075
+ {
1076
+ "epoch": 9.97,
1077
+ "learning_rate": 2.4524029872416333e-05,
1078
+ "loss": 1.0272,
1079
+ "step": 177
1080
+ },
1081
+ {
1082
+ "epoch": 10.03,
1083
+ "learning_rate": 2.42860987301576e-05,
1084
+ "loss": 1.1489,
1085
+ "step": 178
1086
+ },
1087
+ {
1088
+ "epoch": 10.08,
1089
+ "learning_rate": 2.4048232288265253e-05,
1090
+ "loss": 1.0263,
1091
+ "step": 179
1092
+ },
1093
+ {
1094
+ "epoch": 10.14,
1095
+ "learning_rate": 2.3810452104406444e-05,
1096
+ "loss": 1.0305,
1097
+ "step": 180
1098
+ },
1099
+ {
1100
+ "epoch": 10.2,
1101
+ "learning_rate": 2.35727797284308e-05,
1102
+ "loss": 0.9612,
1103
+ "step": 181
1104
+ },
1105
+ {
1106
+ "epoch": 10.25,
1107
+ "learning_rate": 2.3335236700417404e-05,
1108
+ "loss": 1.1883,
1109
+ "step": 182
1110
+ },
1111
+ {
1112
+ "epoch": 10.31,
1113
+ "learning_rate": 2.309784454872262e-05,
1114
+ "loss": 1.192,
1115
+ "step": 183
1116
+ },
1117
+ {
1118
+ "epoch": 10.37,
1119
+ "learning_rate": 2.2860624788029013e-05,
1120
+ "loss": 1.0919,
1121
+ "step": 184
1122
+ },
1123
+ {
1124
+ "epoch": 10.42,
1125
+ "learning_rate": 2.2623598917395438e-05,
1126
+ "loss": 1.0974,
1127
+ "step": 185
1128
+ },
1129
+ {
1130
+ "epoch": 10.48,
1131
+ "learning_rate": 2.238678841830867e-05,
1132
+ "loss": 1.0938,
1133
+ "step": 186
1134
+ },
1135
+ {
1136
+ "epoch": 10.54,
1137
+ "learning_rate": 2.2150214752736488e-05,
1138
+ "loss": 0.9879,
1139
+ "step": 187
1140
+ },
1141
+ {
1142
+ "epoch": 10.59,
1143
+ "learning_rate": 2.1913899361182632e-05,
1144
+ "loss": 1.0284,
1145
+ "step": 188
1146
+ },
1147
+ {
1148
+ "epoch": 10.65,
1149
+ "learning_rate": 2.167786366074365e-05,
1150
+ "loss": 1.0184,
1151
+ "step": 189
1152
+ },
1153
+ {
1154
+ "epoch": 10.7,
1155
+ "learning_rate": 2.1442129043167874e-05,
1156
+ "loss": 1.0419,
1157
+ "step": 190
1158
+ },
1159
+ {
1160
+ "epoch": 10.76,
1161
+ "learning_rate": 2.1206716872916714e-05,
1162
+ "loss": 1.0631,
1163
+ "step": 191
1164
+ },
1165
+ {
1166
+ "epoch": 10.82,
1167
+ "learning_rate": 2.09716484852284e-05,
1168
+ "loss": 1.0492,
1169
+ "step": 192
1170
+ },
1171
+ {
1172
+ "epoch": 10.87,
1173
+ "learning_rate": 2.0736945184184405e-05,
1174
+ "loss": 1.0924,
1175
+ "step": 193
1176
+ },
1177
+ {
1178
+ "epoch": 10.93,
1179
+ "learning_rate": 2.0502628240778655e-05,
1180
+ "loss": 0.9832,
1181
+ "step": 194
1182
+ },
1183
+ {
1184
+ "epoch": 10.99,
1185
+ "learning_rate": 2.0268718890989753e-05,
1186
+ "loss": 1.1266,
1187
+ "step": 195
1188
+ },
1189
+ {
1190
+ "epoch": 11.04,
1191
+ "learning_rate": 2.003523833385637e-05,
1192
+ "loss": 1.0762,
1193
+ "step": 196
1194
+ },
1195
+ {
1196
+ "epoch": 11.1,
1197
+ "learning_rate": 1.980220772955602e-05,
1198
+ "loss": 1.0011,
1199
+ "step": 197
1200
+ },
1201
+ {
1202
+ "epoch": 11.15,
1203
+ "learning_rate": 1.956964819748731e-05,
1204
+ "loss": 1.0195,
1205
+ "step": 198
1206
+ },
1207
+ {
1208
+ "epoch": 11.21,
1209
+ "learning_rate": 1.9337580814355888e-05,
1210
+ "loss": 1.0442,
1211
+ "step": 199
1212
+ },
1213
+ {
1214
+ "epoch": 11.27,
1215
+ "learning_rate": 1.9106026612264316e-05,
1216
+ "loss": 1.0285,
1217
+ "step": 200
1218
+ },
1219
+ {
1220
+ "epoch": 11.27,
1221
+ "eval_loss": 1.3292417526245117,
1222
+ "eval_runtime": 1.4356,
1223
+ "eval_samples_per_second": 6.966,
1224
+ "eval_steps_per_second": 3.483,
1225
+ "step": 200
1226
+ },
1227
+ {
1228
+ "epoch": 11.32,
1229
+ "learning_rate": 1.8875006576805914e-05,
1230
+ "loss": 0.9805,
1231
+ "step": 201
1232
+ },
1233
+ {
1234
+ "epoch": 11.38,
1235
+ "learning_rate": 1.8644541645162834e-05,
1236
+ "loss": 0.9562,
1237
+ "step": 202
1238
+ },
1239
+ {
1240
+ "epoch": 11.44,
1241
+ "learning_rate": 1.8414652704208583e-05,
1242
+ "loss": 1.0457,
1243
+ "step": 203
1244
+ },
1245
+ {
1246
+ "epoch": 11.49,
1247
+ "learning_rate": 1.8185360588615058e-05,
1248
+ "loss": 1.0217,
1249
+ "step": 204
1250
+ },
1251
+ {
1252
+ "epoch": 11.55,
1253
+ "learning_rate": 1.795668607896426e-05,
1254
+ "loss": 1.1281,
1255
+ "step": 205
1256
+ },
1257
+ {
1258
+ "epoch": 11.61,
1259
+ "learning_rate": 1.7728649899865024e-05,
1260
+ "loss": 0.9654,
1261
+ "step": 206
1262
+ },
1263
+ {
1264
+ "epoch": 11.66,
1265
+ "learning_rate": 1.7501272718074737e-05,
1266
+ "loss": 0.9011,
1267
+ "step": 207
1268
+ },
1269
+ {
1270
+ "epoch": 11.72,
1271
+ "learning_rate": 1.7274575140626318e-05,
1272
+ "loss": 1.097,
1273
+ "step": 208
1274
+ },
1275
+ {
1276
+ "epoch": 11.77,
1277
+ "learning_rate": 1.7048577712960627e-05,
1278
+ "loss": 1.0652,
1279
+ "step": 209
1280
+ },
1281
+ {
1282
+ "epoch": 11.83,
1283
+ "learning_rate": 1.682330091706446e-05,
1284
+ "loss": 1.011,
1285
+ "step": 210
1286
+ },
1287
+ {
1288
+ "epoch": 11.89,
1289
+ "learning_rate": 1.6598765169614243e-05,
1290
+ "loss": 1.1001,
1291
+ "step": 211
1292
+ },
1293
+ {
1294
+ "epoch": 11.94,
1295
+ "learning_rate": 1.637499082012574e-05,
1296
+ "loss": 1.1224,
1297
+ "step": 212
1298
+ },
1299
+ {
1300
+ "epoch": 12.0,
1301
+ "learning_rate": 1.615199814910971e-05,
1302
+ "loss": 0.967,
1303
+ "step": 213
1304
+ },
1305
+ {
1306
+ "epoch": 12.06,
1307
+ "learning_rate": 1.5929807366233977e-05,
1308
+ "loss": 0.9923,
1309
+ "step": 214
1310
+ },
1311
+ {
1312
+ "epoch": 12.11,
1313
+ "learning_rate": 1.5708438608491814e-05,
1314
+ "loss": 1.0722,
1315
+ "step": 215
1316
+ },
1317
+ {
1318
+ "epoch": 12.17,
1319
+ "learning_rate": 1.5487911938376924e-05,
1320
+ "loss": 0.9666,
1321
+ "step": 216
1322
+ },
1323
+ {
1324
+ "epoch": 12.23,
1325
+ "learning_rate": 1.5268247342065215e-05,
1326
+ "loss": 0.9136,
1327
+ "step": 217
1328
+ },
1329
+ {
1330
+ "epoch": 12.28,
1331
+ "learning_rate": 1.5049464727603452e-05,
1332
+ "loss": 0.9274,
1333
+ "step": 218
1334
+ },
1335
+ {
1336
+ "epoch": 12.34,
1337
+ "learning_rate": 1.4831583923104999e-05,
1338
+ "loss": 1.0955,
1339
+ "step": 219
1340
+ },
1341
+ {
1342
+ "epoch": 12.39,
1343
+ "learning_rate": 1.4614624674952842e-05,
1344
+ "loss": 0.9058,
1345
+ "step": 220
1346
+ },
1347
+ {
1348
+ "epoch": 12.45,
1349
+ "learning_rate": 1.4398606646009963e-05,
1350
+ "loss": 0.8351,
1351
+ "step": 221
1352
+ },
1353
+ {
1354
+ "epoch": 12.51,
1355
+ "learning_rate": 1.4183549413837289e-05,
1356
+ "loss": 1.119,
1357
+ "step": 222
1358
+ },
1359
+ {
1360
+ "epoch": 12.56,
1361
+ "learning_rate": 1.3969472468919461e-05,
1362
+ "loss": 1.0418,
1363
+ "step": 223
1364
+ },
1365
+ {
1366
+ "epoch": 12.62,
1367
+ "learning_rate": 1.3756395212898359e-05,
1368
+ "loss": 0.9102,
1369
+ "step": 224
1370
+ },
1371
+ {
1372
+ "epoch": 12.68,
1373
+ "learning_rate": 1.354433695681474e-05,
1374
+ "loss": 0.997,
1375
+ "step": 225
1376
+ },
1377
+ {
1378
+ "epoch": 12.73,
1379
+ "learning_rate": 1.3333316919358157e-05,
1380
+ "loss": 1.1737,
1381
+ "step": 226
1382
+ },
1383
+ {
1384
+ "epoch": 12.79,
1385
+ "learning_rate": 1.3123354225125128e-05,
1386
+ "loss": 0.9887,
1387
+ "step": 227
1388
+ },
1389
+ {
1390
+ "epoch": 12.85,
1391
+ "learning_rate": 1.2914467902885901e-05,
1392
+ "loss": 0.9984,
1393
+ "step": 228
1394
+ },
1395
+ {
1396
+ "epoch": 12.9,
1397
+ "learning_rate": 1.2706676883859903e-05,
1398
+ "loss": 0.9507,
1399
+ "step": 229
1400
+ },
1401
+ {
1402
+ "epoch": 12.96,
1403
+ "learning_rate": 1.2500000000000006e-05,
1404
+ "loss": 0.9395,
1405
+ "step": 230
1406
+ },
1407
+ {
1408
+ "epoch": 13.01,
1409
+ "learning_rate": 1.2294455982285777e-05,
1410
+ "loss": 1.1445,
1411
+ "step": 231
1412
+ },
1413
+ {
1414
+ "epoch": 13.07,
1415
+ "learning_rate": 1.2090063459025955e-05,
1416
+ "loss": 0.8469,
1417
+ "step": 232
1418
+ },
1419
+ {
1420
+ "epoch": 13.13,
1421
+ "learning_rate": 1.1886840954170142e-05,
1422
+ "loss": 0.9936,
1423
+ "step": 233
1424
+ },
1425
+ {
1426
+ "epoch": 13.18,
1427
+ "learning_rate": 1.1684806885630004e-05,
1428
+ "loss": 0.8797,
1429
+ "step": 234
1430
+ },
1431
+ {
1432
+ "epoch": 13.24,
1433
+ "learning_rate": 1.148397956361007e-05,
1434
+ "loss": 1.1331,
1435
+ "step": 235
1436
+ },
1437
+ {
1438
+ "epoch": 13.3,
1439
+ "learning_rate": 1.1284377188948258e-05,
1440
+ "loss": 0.9613,
1441
+ "step": 236
1442
+ },
1443
+ {
1444
+ "epoch": 13.35,
1445
+ "learning_rate": 1.1086017851466413e-05,
1446
+ "loss": 0.9343,
1447
+ "step": 237
1448
+ },
1449
+ {
1450
+ "epoch": 13.41,
1451
+ "learning_rate": 1.0888919528330777e-05,
1452
+ "loss": 1.0098,
1453
+ "step": 238
1454
+ },
1455
+ {
1456
+ "epoch": 13.46,
1457
+ "learning_rate": 1.0693100082422763e-05,
1458
+ "loss": 0.9669,
1459
+ "step": 239
1460
+ },
1461
+ {
1462
+ "epoch": 13.52,
1463
+ "learning_rate": 1.049857726072005e-05,
1464
+ "loss": 0.9686,
1465
+ "step": 240
1466
+ },
1467
+ {
1468
+ "epoch": 13.58,
1469
+ "learning_rate": 1.0305368692688174e-05,
1470
+ "loss": 0.8916,
1471
+ "step": 241
1472
+ },
1473
+ {
1474
+ "epoch": 13.63,
1475
+ "learning_rate": 1.0113491888682802e-05,
1476
+ "loss": 0.9274,
1477
+ "step": 242
1478
+ },
1479
+ {
1480
+ "epoch": 13.69,
1481
+ "learning_rate": 9.922964238362762e-06,
1482
+ "loss": 0.9307,
1483
+ "step": 243
1484
+ },
1485
+ {
1486
+ "epoch": 13.75,
1487
+ "learning_rate": 9.733803009114045e-06,
1488
+ "loss": 1.0112,
1489
+ "step": 244
1490
+ },
1491
+ {
1492
+ "epoch": 13.8,
1493
+ "learning_rate": 9.546025344484869e-06,
1494
+ "loss": 1.0465,
1495
+ "step": 245
1496
+ },
1497
+ {
1498
+ "epoch": 13.86,
1499
+ "learning_rate": 9.359648262631961e-06,
1500
+ "loss": 1.0822,
1501
+ "step": 246
1502
+ },
1503
+ {
1504
+ "epoch": 13.92,
1505
+ "learning_rate": 9.174688654778243e-06,
1506
+ "loss": 0.8992,
1507
+ "step": 247
1508
+ },
1509
+ {
1510
+ "epoch": 13.97,
1511
+ "learning_rate": 8.991163283681944e-06,
1512
+ "loss": 0.8778,
1513
+ "step": 248
1514
+ },
1515
+ {
1516
+ "epoch": 14.03,
1517
+ "learning_rate": 8.809088782117452e-06,
1518
+ "loss": 1.0079,
1519
+ "step": 249
1520
+ },
1521
+ {
1522
+ "epoch": 14.08,
1523
+ "learning_rate": 8.628481651367876e-06,
1524
+ "loss": 0.9508,
1525
+ "step": 250
1526
+ },
1527
+ {
1528
+ "epoch": 14.14,
1529
+ "learning_rate": 8.449358259729537e-06,
1530
+ "loss": 0.997,
1531
+ "step": 251
1532
+ },
1533
+ {
1534
+ "epoch": 14.2,
1535
+ "learning_rate": 8.271734841028553e-06,
1536
+ "loss": 0.8669,
1537
+ "step": 252
1538
+ },
1539
+ {
1540
+ "epoch": 14.25,
1541
+ "learning_rate": 8.09562749314952e-06,
1542
+ "loss": 0.9721,
1543
+ "step": 253
1544
+ },
1545
+ {
1546
+ "epoch": 14.31,
1547
+ "learning_rate": 7.921052176576644e-06,
1548
+ "loss": 1.0777,
1549
+ "step": 254
1550
+ },
1551
+ {
1552
+ "epoch": 14.37,
1553
+ "learning_rate": 7.748024712947205e-06,
1554
+ "loss": 0.8644,
1555
+ "step": 255
1556
+ },
1557
+ {
1558
+ "epoch": 14.42,
1559
+ "learning_rate": 7.576560783617668e-06,
1560
+ "loss": 0.9119,
1561
+ "step": 256
1562
+ },
1563
+ {
1564
+ "epoch": 14.48,
1565
+ "learning_rate": 7.40667592824249e-06,
1566
+ "loss": 0.8645,
1567
+ "step": 257
1568
+ },
1569
+ {
1570
+ "epoch": 14.54,
1571
+ "learning_rate": 7.238385543365783e-06,
1572
+ "loss": 0.8786,
1573
+ "step": 258
1574
+ },
1575
+ {
1576
+ "epoch": 14.59,
1577
+ "learning_rate": 7.071704881025915e-06,
1578
+ "loss": 0.9539,
1579
+ "step": 259
1580
+ },
1581
+ {
1582
+ "epoch": 14.65,
1583
+ "learning_rate": 6.906649047373246e-06,
1584
+ "loss": 0.9521,
1585
+ "step": 260
1586
+ },
1587
+ {
1588
+ "epoch": 14.7,
1589
+ "learning_rate": 6.743233001301067e-06,
1590
+ "loss": 1.0708,
1591
+ "step": 261
1592
+ },
1593
+ {
1594
+ "epoch": 14.76,
1595
+ "learning_rate": 6.5814715530898745e-06,
1596
+ "loss": 1.033,
1597
+ "step": 262
1598
+ },
1599
+ {
1600
+ "epoch": 14.82,
1601
+ "learning_rate": 6.421379363065142e-06,
1602
+ "loss": 0.9344,
1603
+ "step": 263
1604
+ },
1605
+ {
1606
+ "epoch": 14.87,
1607
+ "learning_rate": 6.2629709402686535e-06,
1608
+ "loss": 0.9491,
1609
+ "step": 264
1610
+ },
1611
+ {
1612
+ "epoch": 14.93,
1613
+ "learning_rate": 6.106260641143546e-06,
1614
+ "loss": 0.8894,
1615
+ "step": 265
1616
+ },
1617
+ {
1618
+ "epoch": 14.99,
1619
+ "learning_rate": 5.951262668233232e-06,
1620
+ "loss": 0.8825,
1621
+ "step": 266
1622
+ },
1623
+ {
1624
+ "epoch": 15.04,
1625
+ "learning_rate": 5.797991068894201e-06,
1626
+ "loss": 0.925,
1627
+ "step": 267
1628
+ },
1629
+ {
1630
+ "epoch": 15.1,
1631
+ "learning_rate": 5.646459734022938e-06,
1632
+ "loss": 0.8431,
1633
+ "step": 268
1634
+ },
1635
+ {
1636
+ "epoch": 15.15,
1637
+ "learning_rate": 5.49668239679699e-06,
1638
+ "loss": 1.036,
1639
+ "step": 269
1640
+ },
1641
+ {
1642
+ "epoch": 15.21,
1643
+ "learning_rate": 5.348672631430318e-06,
1644
+ "loss": 0.9463,
1645
+ "step": 270
1646
+ },
1647
+ {
1648
+ "epoch": 15.27,
1649
+ "learning_rate": 5.202443851943126e-06,
1650
+ "loss": 0.9942,
1651
+ "step": 271
1652
+ },
1653
+ {
1654
+ "epoch": 15.32,
1655
+ "learning_rate": 5.058009310946119e-06,
1656
+ "loss": 0.9657,
1657
+ "step": 272
1658
+ },
1659
+ {
1660
+ "epoch": 15.38,
1661
+ "learning_rate": 4.915382098439436e-06,
1662
+ "loss": 0.9375,
1663
+ "step": 273
1664
+ },
1665
+ {
1666
+ "epoch": 15.44,
1667
+ "learning_rate": 4.7745751406263165e-06,
1668
+ "loss": 0.9347,
1669
+ "step": 274
1670
+ },
1671
+ {
1672
+ "epoch": 15.49,
1673
+ "learning_rate": 4.6356011987416075e-06,
1674
+ "loss": 0.9796,
1675
+ "step": 275
1676
+ },
1677
+ {
1678
+ "epoch": 15.55,
1679
+ "learning_rate": 4.498472867895223e-06,
1680
+ "loss": 0.987,
1681
+ "step": 276
1682
+ },
1683
+ {
1684
+ "epoch": 15.61,
1685
+ "learning_rate": 4.36320257593065e-06,
1686
+ "loss": 0.8946,
1687
+ "step": 277
1688
+ },
1689
+ {
1690
+ "epoch": 15.66,
1691
+ "learning_rate": 4.229802582298633e-06,
1692
+ "loss": 0.9034,
1693
+ "step": 278
1694
+ },
1695
+ {
1696
+ "epoch": 15.72,
1697
+ "learning_rate": 4.098284976946101e-06,
1698
+ "loss": 1.0294,
1699
+ "step": 279
1700
+ },
1701
+ {
1702
+ "epoch": 15.77,
1703
+ "learning_rate": 3.968661679220468e-06,
1704
+ "loss": 0.9631,
1705
+ "step": 280
1706
+ },
1707
+ {
1708
+ "epoch": 15.83,
1709
+ "learning_rate": 3.840944436789393e-06,
1710
+ "loss": 0.8831,
1711
+ "step": 281
1712
+ },
1713
+ {
1714
+ "epoch": 15.89,
1715
+ "learning_rate": 3.715144824576078e-06,
1716
+ "loss": 0.9274,
1717
+ "step": 282
1718
+ },
1719
+ {
1720
+ "epoch": 15.94,
1721
+ "learning_rate": 3.591274243710277e-06,
1722
+ "loss": 0.8896,
1723
+ "step": 283
1724
+ },
1725
+ {
1726
+ "epoch": 16.0,
1727
+ "learning_rate": 3.4693439204949858e-06,
1728
+ "loss": 0.8672,
1729
+ "step": 284
1730
+ },
1731
+ {
1732
+ "epoch": 16.06,
1733
+ "learning_rate": 3.3493649053890326e-06,
1734
+ "loss": 0.9043,
1735
+ "step": 285
1736
+ },
1737
+ {
1738
+ "epoch": 16.11,
1739
+ "learning_rate": 3.2313480720055745e-06,
1740
+ "loss": 1.1059,
1741
+ "step": 286
1742
+ },
1743
+ {
1744
+ "epoch": 16.17,
1745
+ "learning_rate": 3.1153041161266134e-06,
1746
+ "loss": 0.9241,
1747
+ "step": 287
1748
+ },
1749
+ {
1750
+ "epoch": 16.23,
1751
+ "learning_rate": 3.0012435547336737e-06,
1752
+ "loss": 0.7761,
1753
+ "step": 288
1754
+ },
1755
+ {
1756
+ "epoch": 16.28,
1757
+ "learning_rate": 2.889176725054643e-06,
1758
+ "loss": 1.0323,
1759
+ "step": 289
1760
+ },
1761
+ {
1762
+ "epoch": 16.34,
1763
+ "learning_rate": 2.779113783626916e-06,
1764
+ "loss": 0.8534,
1765
+ "step": 290
1766
+ },
1767
+ {
1768
+ "epoch": 16.39,
1769
+ "learning_rate": 2.671064705376905e-06,
1770
+ "loss": 0.7669,
1771
+ "step": 291
1772
+ },
1773
+ {
1774
+ "epoch": 16.45,
1775
+ "learning_rate": 2.565039282716045e-06,
1776
+ "loss": 0.984,
1777
+ "step": 292
1778
+ },
1779
+ {
1780
+ "epoch": 16.51,
1781
+ "learning_rate": 2.461047124653279e-06,
1782
+ "loss": 0.9542,
1783
+ "step": 293
1784
+ },
1785
+ {
1786
+ "epoch": 16.56,
1787
+ "learning_rate": 2.3590976559242278e-06,
1788
+ "loss": 0.9439,
1789
+ "step": 294
1790
+ },
1791
+ {
1792
+ "epoch": 16.62,
1793
+ "learning_rate": 2.2592001161370392e-06,
1794
+ "loss": 0.8776,
1795
+ "step": 295
1796
+ },
1797
+ {
1798
+ "epoch": 16.68,
1799
+ "learning_rate": 2.1613635589349756e-06,
1800
+ "loss": 0.9211,
1801
+ "step": 296
1802
+ },
1803
+ {
1804
+ "epoch": 16.73,
1805
+ "learning_rate": 2.0655968511759134e-06,
1806
+ "loss": 0.9145,
1807
+ "step": 297
1808
+ },
1809
+ {
1810
+ "epoch": 16.79,
1811
+ "learning_rate": 1.97190867212875e-06,
1812
+ "loss": 0.9481,
1813
+ "step": 298
1814
+ },
1815
+ {
1816
+ "epoch": 16.85,
1817
+ "learning_rate": 1.8803075126867715e-06,
1818
+ "loss": 0.9805,
1819
+ "step": 299
1820
+ },
1821
+ {
1822
+ "epoch": 16.9,
1823
+ "learning_rate": 1.790801674598186e-06,
1824
+ "loss": 0.8387,
1825
+ "step": 300
1826
+ },
1827
+ {
1828
+ "epoch": 16.9,
1829
+ "eval_loss": 1.3825798034667969,
1830
+ "eval_runtime": 1.4357,
1831
+ "eval_samples_per_second": 6.965,
1832
+ "eval_steps_per_second": 3.483,
1833
+ "step": 300
1834
+ },
1835
+ {
1836
+ "epoch": 16.96,
1837
+ "learning_rate": 1.703399269713693e-06,
1838
+ "loss": 1.005,
1839
+ "step": 301
1840
+ },
1841
+ {
1842
+ "epoch": 17.01,
1843
+ "learning_rate": 1.6181082192513353e-06,
1844
+ "loss": 0.9032,
1845
+ "step": 302
1846
+ },
1847
+ {
1848
+ "epoch": 17.07,
1849
+ "learning_rate": 1.534936253078606e-06,
1850
+ "loss": 1.0064,
1851
+ "step": 303
1852
+ },
1853
+ {
1854
+ "epoch": 17.13,
1855
+ "learning_rate": 1.4538909090118846e-06,
1856
+ "loss": 0.8834,
1857
+ "step": 304
1858
+ },
1859
+ {
1860
+ "epoch": 17.18,
1861
+ "learning_rate": 1.3749795321332887e-06,
1862
+ "loss": 0.93,
1863
+ "step": 305
1864
+ },
1865
+ {
1866
+ "epoch": 17.24,
1867
+ "learning_rate": 1.2982092741250145e-06,
1868
+ "loss": 0.9181,
1869
+ "step": 306
1870
+ },
1871
+ {
1872
+ "epoch": 17.3,
1873
+ "learning_rate": 1.2235870926211619e-06,
1874
+ "loss": 0.8043,
1875
+ "step": 307
1876
+ },
1877
+ {
1878
+ "epoch": 17.35,
1879
+ "learning_rate": 1.1511197505771843e-06,
1880
+ "loss": 0.8552,
1881
+ "step": 308
1882
+ },
1883
+ {
1884
+ "epoch": 17.41,
1885
+ "learning_rate": 1.0808138156569614e-06,
1886
+ "loss": 0.8864,
1887
+ "step": 309
1888
+ },
1889
+ {
1890
+ "epoch": 17.46,
1891
+ "learning_rate": 1.0126756596375686e-06,
1892
+ "loss": 0.8856,
1893
+ "step": 310
1894
+ },
1895
+ {
1896
+ "epoch": 17.52,
1897
+ "learning_rate": 9.467114578318226e-07,
1898
+ "loss": 0.9447,
1899
+ "step": 311
1900
+ },
1901
+ {
1902
+ "epoch": 17.58,
1903
+ "learning_rate": 8.829271885286094e-07,
1904
+ "loss": 0.8091,
1905
+ "step": 312
1906
+ },
1907
+ {
1908
+ "epoch": 17.63,
1909
+ "learning_rate": 8.213286324510738e-07,
1910
+ "loss": 0.9956,
1911
+ "step": 313
1912
+ },
1913
+ {
1914
+ "epoch": 17.69,
1915
+ "learning_rate": 7.619213722327185e-07,
1916
+ "loss": 0.88,
1917
+ "step": 314
1918
+ },
1919
+ {
1920
+ "epoch": 17.75,
1921
+ "learning_rate": 7.047107919114588e-07,
1922
+ "loss": 0.9142,
1923
+ "step": 315
1924
+ },
1925
+ {
1926
+ "epoch": 17.8,
1927
+ "learning_rate": 6.497020764416633e-07,
1928
+ "loss": 0.9984,
1929
+ "step": 316
1930
+ },
1931
+ {
1932
+ "epoch": 17.86,
1933
+ "learning_rate": 5.969002112242494e-07,
1934
+ "loss": 0.8869,
1935
+ "step": 317
1936
+ },
1937
+ {
1938
+ "epoch": 17.92,
1939
+ "learning_rate": 5.463099816548579e-07,
1940
+ "loss": 0.9627,
1941
+ "step": 318
1942
+ },
1943
+ {
1944
+ "epoch": 17.97,
1945
+ "learning_rate": 4.979359726901639e-07,
1946
+ "loss": 0.9783,
1947
+ "step": 319
1948
+ },
1949
+ {
1950
+ "epoch": 18.03,
1951
+ "learning_rate": 4.517825684323324e-07,
1952
+ "loss": 1.0067,
1953
+ "step": 320
1954
+ },
1955
+ {
1956
+ "epoch": 18.08,
1957
+ "learning_rate": 4.078539517316987e-07,
1958
+ "loss": 0.9261,
1959
+ "step": 321
1960
+ },
1961
+ {
1962
+ "epoch": 18.14,
1963
+ "learning_rate": 3.6615410380767544e-07,
1964
+ "loss": 1.0504,
1965
+ "step": 322
1966
+ },
1967
+ {
1968
+ "epoch": 18.2,
1969
+ "learning_rate": 3.266868038879434e-07,
1970
+ "loss": 0.8995,
1971
+ "step": 323
1972
+ },
1973
+ {
1974
+ "epoch": 18.25,
1975
+ "learning_rate": 2.894556288659395e-07,
1976
+ "loss": 0.915,
1977
+ "step": 324
1978
+ },
1979
+ {
1980
+ "epoch": 18.31,
1981
+ "learning_rate": 2.544639529766829e-07,
1982
+ "loss": 0.8657,
1983
+ "step": 325
1984
+ },
1985
+ {
1986
+ "epoch": 18.37,
1987
+ "learning_rate": 2.2171494749097245e-07,
1988
+ "loss": 0.9171,
1989
+ "step": 326
1990
+ },
1991
+ {
1992
+ "epoch": 18.42,
1993
+ "learning_rate": 1.912115804279746e-07,
1994
+ "loss": 0.7903,
1995
+ "step": 327
1996
+ },
1997
+ {
1998
+ "epoch": 18.48,
1999
+ "learning_rate": 1.6295661628624447e-07,
2000
+ "loss": 0.9827,
2001
+ "step": 328
2002
+ },
2003
+ {
2004
+ "epoch": 18.54,
2005
+ "learning_rate": 1.3695261579316777e-07,
2006
+ "loss": 0.9373,
2007
+ "step": 329
2008
+ },
2009
+ {
2010
+ "epoch": 18.59,
2011
+ "learning_rate": 1.1320193567288529e-07,
2012
+ "loss": 0.915,
2013
+ "step": 330
2014
+ },
2015
+ {
2016
+ "epoch": 18.65,
2017
+ "learning_rate": 9.170672843271666e-08,
2018
+ "loss": 1.0085,
2019
+ "step": 331
2020
+ },
2021
+ {
2022
+ "epoch": 18.7,
2023
+ "learning_rate": 7.246894216806355e-08,
2024
+ "loss": 0.9681,
2025
+ "step": 332
2026
+ },
2027
+ {
2028
+ "epoch": 18.76,
2029
+ "learning_rate": 5.54903203858731e-08,
2030
+ "loss": 0.8392,
2031
+ "step": 333
2032
+ },
2033
+ {
2034
+ "epoch": 18.82,
2035
+ "learning_rate": 4.07724018466088e-08,
2036
+ "loss": 0.9473,
2037
+ "step": 334
2038
+ },
2039
+ {
2040
+ "epoch": 18.87,
2041
+ "learning_rate": 2.831652042480093e-08,
2042
+ "loss": 0.9534,
2043
+ "step": 335
2044
+ },
2045
+ {
2046
+ "epoch": 18.93,
2047
+ "learning_rate": 1.812380498815991e-08,
2048
+ "loss": 0.9152,
2049
+ "step": 336
2050
+ },
2051
+ {
2052
+ "epoch": 18.99,
2053
+ "learning_rate": 1.0195179295269252e-08,
2054
+ "loss": 0.8761,
2055
+ "step": 337
2056
+ },
2057
+ {
2058
+ "epoch": 19.04,
2059
+ "learning_rate": 4.531361911855325e-09,
2060
+ "loss": 0.9116,
2061
+ "step": 338
2062
+ },
2063
+ {
2064
+ "epoch": 19.1,
2065
+ "learning_rate": 1.132866145678313e-09,
2066
+ "loss": 0.9118,
2067
+ "step": 339
2068
+ },
2069
+ {
2070
+ "epoch": 19.15,
2071
+ "learning_rate": 0.0,
2072
+ "loss": 0.8806,
2073
+ "step": 340
2074
+ }
2075
+ ],
2076
+ "logging_steps": 1,
2077
+ "max_steps": 340,
2078
+ "num_train_epochs": 20,
2079
+ "save_steps": 500,
2080
+ "total_flos": 4.253890839379968e+17,
2081
+ "trial_name": null,
2082
+ "trial_params": null
2083
+ }
checkpoint-340/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af7bb882cac5b4d33f676aa38a935fcc9173c63dc3da717a69a225490ea5e0f7
3
+ size 4155
special_tokens_map.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "pad_token": "[PAD]",
5
+ "unk_token": "<unk>"
6
+ }
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723