Hugging Face's logo

royallab
/
LimaRP-ShareGPT-13b-qloras

PEFT
Not-For-All-Audiences
Model card Files Community
2
kingbri commited on
Commit
85aac7e
•
1 Parent(s): f413d6b

Add multi-prompt twochar

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. multi-prompt/twochar/README.md +21 -0
  2. multi-prompt/twochar/adapter_config.json +26 -0
  3. multi-prompt/twochar/adapter_model.bin +3 -0
  4. multi-prompt/twochar/added_tokens.json +3 -0
  5. multi-prompt/twochar/checkpoint-100/README.md +21 -0
  6. multi-prompt/twochar/checkpoint-100/adapter_config.json +26 -0
  7. multi-prompt/twochar/checkpoint-100/adapter_model.bin +3 -0
  8. multi-prompt/twochar/checkpoint-100/adapter_model/README.md +21 -0
  9. multi-prompt/twochar/checkpoint-100/adapter_model/adapter_config.json +26 -0
  10. multi-prompt/twochar/checkpoint-100/adapter_model/adapter_model.bin +3 -0
  11. multi-prompt/twochar/checkpoint-100/optimizer.pt +3 -0
  12. multi-prompt/twochar/checkpoint-100/rng_state_0.pth +3 -0
  13. multi-prompt/twochar/checkpoint-100/rng_state_1.pth +3 -0
  14. multi-prompt/twochar/checkpoint-100/rng_state_2.pth +3 -0
  15. multi-prompt/twochar/checkpoint-100/rng_state_3.pth +3 -0
  16. multi-prompt/twochar/checkpoint-100/rng_state_4.pth +3 -0
  17. multi-prompt/twochar/checkpoint-100/rng_state_5.pth +3 -0
  18. multi-prompt/twochar/checkpoint-100/rng_state_6.pth +3 -0
  19. multi-prompt/twochar/checkpoint-100/rng_state_7.pth +3 -0
  20. multi-prompt/twochar/checkpoint-100/scheduler.pt +3 -0
  21. multi-prompt/twochar/checkpoint-100/trainer_state.json +659 -0
  22. multi-prompt/twochar/checkpoint-100/training_args.bin +3 -0
  23. multi-prompt/twochar/checkpoint-150/README.md +21 -0
  24. multi-prompt/twochar/checkpoint-150/adapter_config.json +26 -0
  25. multi-prompt/twochar/checkpoint-150/adapter_model.bin +3 -0
  26. multi-prompt/twochar/checkpoint-150/adapter_model/README.md +21 -0
  27. multi-prompt/twochar/checkpoint-150/adapter_model/adapter_config.json +26 -0
  28. multi-prompt/twochar/checkpoint-150/adapter_model/adapter_model.bin +3 -0
  29. multi-prompt/twochar/checkpoint-150/optimizer.pt +3 -0
  30. multi-prompt/twochar/checkpoint-150/rng_state_0.pth +3 -0
  31. multi-prompt/twochar/checkpoint-150/rng_state_1.pth +3 -0
  32. multi-prompt/twochar/checkpoint-150/rng_state_2.pth +3 -0
  33. multi-prompt/twochar/checkpoint-150/rng_state_3.pth +3 -0
  34. multi-prompt/twochar/checkpoint-150/rng_state_4.pth +3 -0
  35. multi-prompt/twochar/checkpoint-150/rng_state_5.pth +3 -0
  36. multi-prompt/twochar/checkpoint-150/rng_state_6.pth +3 -0
  37. multi-prompt/twochar/checkpoint-150/rng_state_7.pth +3 -0
  38. multi-prompt/twochar/checkpoint-150/scheduler.pt +3 -0
  39. multi-prompt/twochar/checkpoint-150/trainer_state.json +975 -0
  40. multi-prompt/twochar/checkpoint-150/training_args.bin +3 -0
  41. multi-prompt/twochar/checkpoint-200/README.md +21 -0
  42. multi-prompt/twochar/checkpoint-200/adapter_config.json +26 -0
  43. multi-prompt/twochar/checkpoint-200/adapter_model.bin +3 -0
  44. multi-prompt/twochar/checkpoint-200/adapter_model/README.md +21 -0
  45. multi-prompt/twochar/checkpoint-200/adapter_model/adapter_config.json +26 -0
  46. multi-prompt/twochar/checkpoint-200/adapter_model/adapter_model.bin +3 -0
  47. multi-prompt/twochar/checkpoint-200/optimizer.pt +3 -0
  48. multi-prompt/twochar/checkpoint-200/rng_state_0.pth +3 -0
  49. multi-prompt/twochar/checkpoint-200/rng_state_1.pth +3 -0
  50. multi-prompt/twochar/checkpoint-200/rng_state_2.pth +3 -0
multi-prompt/twochar/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 0.6.0.dev0
multi-prompt/twochar/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "meta-llama/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": 0.01,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "v_proj",
18
+ "q_proj",
19
+ "down_proj",
20
+ "k_proj",
21
+ "gate_proj",
22
+ "o_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
multi-prompt/twochar/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:433a53226b2a4581547b8a2e381d865d58aa0575acf593735501071cf46d7fef
3
+ size 125374989
multi-prompt/twochar/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[PAD]": 32000
3
+ }
multi-prompt/twochar/checkpoint-100/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 0.6.0.dev0
multi-prompt/twochar/checkpoint-100/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "meta-llama/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": 0.01,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "v_proj",
18
+ "q_proj",
19
+ "down_proj",
20
+ "k_proj",
21
+ "gate_proj",
22
+ "o_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
multi-prompt/twochar/checkpoint-100/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0708b4cbcf4d2eab6bb13a3b593efb8f9ffd0d3554bd6f0d85be7c2443d26b80
3
+ size 125374989
multi-prompt/twochar/checkpoint-100/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 0.6.0.dev0
multi-prompt/twochar/checkpoint-100/adapter_model/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "meta-llama/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": 0.01,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "v_proj",
18
+ "q_proj",
19
+ "down_proj",
20
+ "k_proj",
21
+ "gate_proj",
22
+ "o_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
multi-prompt/twochar/checkpoint-100/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0708b4cbcf4d2eab6bb13a3b593efb8f9ffd0d3554bd6f0d85be7c2443d26b80
3
+ size 125374989
multi-prompt/twochar/checkpoint-100/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0221233935a3705a400234359b0d51ecb64e96906b36ad8742d87e4e81167bd8
3
+ size 250681597
multi-prompt/twochar/checkpoint-100/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef3b38ac0e685cf7486af95e3e44d65bd9d5f0166e4b7cda4087f3df177fe2ac
3
+ size 21687
multi-prompt/twochar/checkpoint-100/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:96f602cf80bbe8b7b20f61815b81a180e05c4349767b3055d38cee11e56d2f52
3
+ size 21687
multi-prompt/twochar/checkpoint-100/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c4c642ebadc5d5d3863c2af80d71300c9b960385b08a7e97d47221351e923536
3
+ size 21687
multi-prompt/twochar/checkpoint-100/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bdbab4ca5a74f90e77ec13e5c8a300b12d2ec0ffa5f7150ef45db3b39c0ba3fd
3
+ size 21687
multi-prompt/twochar/checkpoint-100/rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8b9e295735ec1efb6e545052a0a5b42c962567ac43c037ef05cd67f65aac86e5
3
+ size 21687
multi-prompt/twochar/checkpoint-100/rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bf0b7826ae8f9d6847c100c005a195903f6702ffefbcf20043e504db1cbb9cf0
3
+ size 21687
multi-prompt/twochar/checkpoint-100/rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3c9b6a36344bf4dd837aca81f6f221eaaf42fdaf4ad38def06c6583bd294478b
3
+ size 21687
multi-prompt/twochar/checkpoint-100/rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fe85df90b682b7f8f20f078de75c81ecf060770f57b3f4465e9d00985eaa8a39
3
+ size 21687
multi-prompt/twochar/checkpoint-100/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b8c8c5e17820ade3bd0f83be62a888f354614fa9e9626e94814a99519af6a262
3
+ size 627
multi-prompt/twochar/checkpoint-100/trainer_state.json ADDED
@@ -0,0 +1,659 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9900990099009901,
5
+ "eval_steps": 20,
6
+ "global_step": 100,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 6.5e-06,
14
+ "loss": 2.0407,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.02,
19
+ "learning_rate": 1.3e-05,
20
+ "loss": 2.3107,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.03,
25
+ "learning_rate": 1.9499999999999996e-05,
26
+ "loss": 2.1613,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.04,
31
+ "learning_rate": 2.6e-05,
32
+ "loss": 2.0444,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.05,
37
+ "learning_rate": 3.25e-05,
38
+ "loss": 2.2512,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.06,
43
+ "learning_rate": 3.899999999999999e-05,
44
+ "loss": 2.0516,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.07,
49
+ "learning_rate": 4.5499999999999995e-05,
50
+ "loss": 2.2006,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.08,
55
+ "learning_rate": 5.2e-05,
56
+ "loss": 2.1229,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.09,
61
+ "learning_rate": 5.85e-05,
62
+ "loss": 2.2498,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.1,
67
+ "learning_rate": 6.5e-05,
68
+ "loss": 2.0293,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.11,
73
+ "learning_rate": 6.499564948206075e-05,
74
+ "loss": 2.1316,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.12,
79
+ "learning_rate": 6.498259909298188e-05,
80
+ "loss": 2.1274,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.13,
85
+ "learning_rate": 6.49608523266681e-05,
86
+ "loss": 2.1129,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.14,
91
+ "learning_rate": 6.493041500525461e-05,
92
+ "loss": 2.1517,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.15,
97
+ "learning_rate": 6.489129527754836e-05,
98
+ "loss": 2.1078,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.16,
103
+ "learning_rate": 6.484350361684639e-05,
104
+ "loss": 2.0362,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.17,
109
+ "learning_rate": 6.478705281813194e-05,
110
+ "loss": 2.0553,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 0.18,
115
+ "learning_rate": 6.472195799464884e-05,
116
+ "loss": 2.1223,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 0.19,
121
+ "learning_rate": 6.464823657385538e-05,
122
+ "loss": 2.0326,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 0.2,
127
+ "learning_rate": 6.456590829275857e-05,
128
+ "loss": 2.1802,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 0.2,
133
+ "eval_loss": 2.012301445007324,
134
+ "eval_runtime": 1.4855,
135
+ "eval_samples_per_second": 5.385,
136
+ "eval_steps_per_second": 0.673,
137
+ "step": 20
138
+ },
139
+ {
140
+ "epoch": 0.21,
141
+ "learning_rate": 6.447499519263001e-05,
142
+ "loss": 2.0665,
143
+ "step": 21
144
+ },
145
+ {
146
+ "epoch": 0.22,
147
+ "learning_rate": 6.437552161310498e-05,
148
+ "loss": 2.0354,
149
+ "step": 22
150
+ },
151
+ {
152
+ "epoch": 0.23,
153
+ "learning_rate": 6.426751418566609e-05,
154
+ "loss": 2.1289,
155
+ "step": 23
156
+ },
157
+ {
158
+ "epoch": 0.24,
159
+ "learning_rate": 6.415100182651334e-05,
160
+ "loss": 2.0742,
161
+ "step": 24
162
+ },
163
+ {
164
+ "epoch": 0.25,
165
+ "learning_rate": 6.402601572882268e-05,
166
+ "loss": 2.0412,
167
+ "step": 25
168
+ },
169
+ {
170
+ "epoch": 0.26,
171
+ "learning_rate": 6.38925893543947e-05,
172
+ "loss": 2.1974,
173
+ "step": 26
174
+ },
175
+ {
176
+ "epoch": 0.27,
177
+ "learning_rate": 6.375075842469626e-05,
178
+ "loss": 2.196,
179
+ "step": 27
180
+ },
181
+ {
182
+ "epoch": 0.28,
183
+ "learning_rate": 6.360056091129678e-05,
184
+ "loss": 2.0775,
185
+ "step": 28
186
+ },
187
+ {
188
+ "epoch": 0.29,
189
+ "learning_rate": 6.344203702570254e-05,
190
+ "loss": 2.0366,
191
+ "step": 29
192
+ },
193
+ {
194
+ "epoch": 0.3,
195
+ "learning_rate": 6.327522920859093e-05,
196
+ "loss": 2.1438,
197
+ "step": 30
198
+ },
199
+ {
200
+ "epoch": 0.31,
201
+ "learning_rate": 6.310018211844817e-05,
202
+ "loss": 2.0639,
203
+ "step": 31
204
+ },
205
+ {
206
+ "epoch": 0.32,
207
+ "learning_rate": 6.291694261961308e-05,
208
+ "loss": 2.1794,
209
+ "step": 32
210
+ },
211
+ {
212
+ "epoch": 0.33,
213
+ "learning_rate": 6.272555976973039e-05,
214
+ "loss": 2.0592,
215
+ "step": 33
216
+ },
217
+ {
218
+ "epoch": 0.34,
219
+ "learning_rate": 6.252608480661681e-05,
220
+ "loss": 2.1628,
221
+ "step": 34
222
+ },
223
+ {
224
+ "epoch": 0.35,
225
+ "learning_rate": 6.231857113454346e-05,
226
+ "loss": 2.1851,
227
+ "step": 35
228
+ },
229
+ {
230
+ "epoch": 0.36,
231
+ "learning_rate": 6.21030743099382e-05,
232
+ "loss": 2.0531,
233
+ "step": 36
234
+ },
235
+ {
236
+ "epoch": 0.37,
237
+ "learning_rate": 6.18796520265119e-05,
238
+ "loss": 2.0284,
239
+ "step": 37
240
+ },
241
+ {
242
+ "epoch": 0.38,
243
+ "learning_rate": 6.164836409981237e-05,
244
+ "loss": 2.2359,
245
+ "step": 38
246
+ },
247
+ {
248
+ "epoch": 0.39,
249
+ "learning_rate": 6.140927245121032e-05,
250
+ "loss": 1.8967,
251
+ "step": 39
252
+ },
253
+ {
254
+ "epoch": 0.4,
255
+ "learning_rate": 6.116244109132153e-05,
256
+ "loss": 2.0275,
257
+ "step": 40
258
+ },
259
+ {
260
+ "epoch": 0.4,
261
+ "eval_loss": 1.9861761331558228,
262
+ "eval_runtime": 1.4896,
263
+ "eval_samples_per_second": 5.37,
264
+ "eval_steps_per_second": 0.671,
265
+ "step": 40
266
+ },
267
+ {
268
+ "epoch": 0.41,
269
+ "learning_rate": 6.0907936102869656e-05,
270
+ "loss": 2.0316,
271
+ "step": 41
272
+ },
273
+ {
274
+ "epoch": 0.42,
275
+ "learning_rate": 6.0645825622994254e-05,
276
+ "loss": 2.0768,
277
+ "step": 42
278
+ },
279
+ {
280
+ "epoch": 0.43,
281
+ "learning_rate": 6.037617982500884e-05,
282
+ "loss": 2.0058,
283
+ "step": 43
284
+ },
285
+ {
286
+ "epoch": 0.44,
287
+ "learning_rate": 6.009907089961381e-05,
288
+ "loss": 2.2319,
289
+ "step": 44
290
+ },
291
+ {
292
+ "epoch": 0.45,
293
+ "learning_rate": 5.981457303556923e-05,
294
+ "loss": 2.1693,
295
+ "step": 45
296
+ },
297
+ {
298
+ "epoch": 0.46,
299
+ "learning_rate": 5.9522762399832716e-05,
300
+ "loss": 2.0858,
301
+ "step": 46
302
+ },
303
+ {
304
+ "epoch": 0.47,
305
+ "learning_rate": 5.922371711716768e-05,
306
+ "loss": 2.0576,
307
+ "step": 47
308
+ },
309
+ {
310
+ "epoch": 0.48,
311
+ "learning_rate": 5.891751724922749e-05,
312
+ "loss": 1.9946,
313
+ "step": 48
314
+ },
315
+ {
316
+ "epoch": 0.49,
317
+ "learning_rate": 5.860424477312095e-05,
318
+ "loss": 2.0166,
319
+ "step": 49
320
+ },
321
+ {
322
+ "epoch": 0.5,
323
+ "learning_rate": 5.828398355946514e-05,
324
+ "loss": 2.1482,
325
+ "step": 50
326
+ },
327
+ {
328
+ "epoch": 0.5,
329
+ "learning_rate": 5.795681934993113e-05,
330
+ "loss": 2.0698,
331
+ "step": 51
332
+ },
333
+ {
334
+ "epoch": 0.51,
335
+ "learning_rate": 5.7622839734288945e-05,
336
+ "loss": 1.9317,
337
+ "step": 52
338
+ },
339
+ {
340
+ "epoch": 0.52,
341
+ "learning_rate": 5.728213412695761e-05,
342
+ "loss": 2.1399,
343
+ "step": 53
344
+ },
345
+ {
346
+ "epoch": 0.53,
347
+ "learning_rate": 5.693479374306676e-05,
348
+ "loss": 2.1637,
349
+ "step": 54
350
+ },
351
+ {
352
+ "epoch": 0.54,
353
+ "learning_rate": 5.658091157403617e-05,
354
+ "loss": 2.0557,
355
+ "step": 55
356
+ },
357
+ {
358
+ "epoch": 0.55,
359
+ "learning_rate": 5.622058236267965e-05,
360
+ "loss": 1.926,
361
+ "step": 56
362
+ },
363
+ {
364
+ "epoch": 0.56,
365
+ "learning_rate": 5.585390257784018e-05,
366
+ "loss": 2.0984,
367
+ "step": 57
368
+ },
369
+ {
370
+ "epoch": 0.57,
371
+ "learning_rate": 5.548097038856279e-05,
372
+ "loss": 2.0428,
373
+ "step": 58
374
+ },
375
+ {
376
+ "epoch": 0.58,
377
+ "learning_rate": 5.5101885637812374e-05,
378
+ "loss": 1.9624,
379
+ "step": 59
380
+ },
381
+ {
382
+ "epoch": 0.59,
383
+ "learning_rate": 5.4716749815743304e-05,
384
+ "loss": 1.913,
385
+ "step": 60
386
+ },
387
+ {
388
+ "epoch": 0.59,
389
+ "eval_loss": 1.97765052318573,
390
+ "eval_runtime": 1.4886,
391
+ "eval_samples_per_second": 5.374,
392
+ "eval_steps_per_second": 0.672,
393
+ "step": 60
394
+ },
395
+ {
396
+ "epoch": 0.6,
397
+ "learning_rate": 5.432566603252809e-05,
398
+ "loss": 2.0125,
399
+ "step": 61
400
+ },
401
+ {
402
+ "epoch": 0.61,
403
+ "learning_rate": 5.3928738990752234e-05,
404
+ "loss": 2.0721,
405
+ "step": 62
406
+ },
407
+ {
408
+ "epoch": 0.62,
409
+ "learning_rate": 5.3526074957382866e-05,
410
+ "loss": 2.1046,
411
+ "step": 63
412
+ },
413
+ {
414
+ "epoch": 0.63,
415
+ "learning_rate": 5.311778173531847e-05,
416
+ "loss": 2.0074,
417
+ "step": 64
418
+ },
419
+ {
420
+ "epoch": 0.64,
421
+ "learning_rate": 5.2703968634527514e-05,
422
+ "loss": 2.1667,
423
+ "step": 65
424
+ },
425
+ {
426
+ "epoch": 0.65,
427
+ "learning_rate": 5.2284746442783414e-05,
428
+ "loss": 2.0677,
429
+ "step": 66
430
+ },
431
+ {
432
+ "epoch": 0.66,
433
+ "learning_rate": 5.186022739600408e-05,
434
+ "loss": 2.1798,
435
+ "step": 67
436
+ },
437
+ {
438
+ "epoch": 0.67,
439
+ "learning_rate": 5.143052514820357e-05,
440
+ "loss": 2.105,
441
+ "step": 68
442
+ },
443
+ {
444
+ "epoch": 0.68,
445
+ "learning_rate": 5.099575474106419e-05,
446
+ "loss": 2.118,
447
+ "step": 69
448
+ },
449
+ {
450
+ "epoch": 0.69,
451
+ "learning_rate": 5.055603257313707e-05,
452
+ "loss": 2.0791,
453
+ "step": 70
454
+ },
455
+ {
456
+ "epoch": 0.7,
457
+ "learning_rate": 5.011147636867943e-05,
458
+ "loss": 2.032,
459
+ "step": 71
460
+ },
461
+ {
462
+ "epoch": 0.71,
463
+ "learning_rate": 4.9662205146136955e-05,
464
+ "loss": 1.9331,
465
+ "step": 72
466
+ },
467
+ {
468
+ "epoch": 0.72,
469
+ "learning_rate": 4.92083391862797e-05,
470
+ "loss": 2.0077,
471
+ "step": 73
472
+ },
473
+ {
474
+ "epoch": 0.73,
475
+ "learning_rate": 4.874999999999999e-05,
476
+ "loss": 2.0253,
477
+ "step": 74
478
+ },
479
+ {
480
+ "epoch": 0.74,
481
+ "learning_rate": 4.828731029578105e-05,
482
+ "loss": 1.9922,
483
+ "step": 75
484
+ },
485
+ {
486
+ "epoch": 0.75,
487
+ "learning_rate": 4.7820393946844926e-05,
488
+ "loss": 2.0295,
489
+ "step": 76
490
+ },
491
+ {
492
+ "epoch": 0.76,
493
+ "learning_rate": 4.734937595798867e-05,
494
+ "loss": 1.9514,
495
+ "step": 77
496
+ },
497
+ {
498
+ "epoch": 0.77,
499
+ "learning_rate": 4.687438243211754e-05,
500
+ "loss": 2.1018,
501
+ "step": 78
502
+ },
503
+ {
504
+ "epoch": 0.78,
505
+ "learning_rate": 4.639554053648416e-05,
506
+ "loss": 2.1032,
507
+ "step": 79
508
+ },
509
+ {
510
+ "epoch": 0.79,
511
+ "learning_rate": 4.5912978468642824e-05,
512
+ "loss": 2.1258,
513
+ "step": 80
514
+ },
515
+ {
516
+ "epoch": 0.79,
517
+ "eval_loss": 1.972991704940796,
518
+ "eval_runtime": 1.494,
519
+ "eval_samples_per_second": 5.355,
520
+ "eval_steps_per_second": 0.669,
521
+ "step": 80
522
+ },
523
+ {
524
+ "epoch": 0.8,
525
+ "learning_rate": 4.542682542212785e-05,
526
+ "loss": 2.0353,
527
+ "step": 81
528
+ },
529
+ {
530
+ "epoch": 0.81,
531
+ "learning_rate": 4.4937211551865415e-05,
532
+ "loss": 1.9987,
533
+ "step": 82
534
+ },
535
+ {
536
+ "epoch": 0.82,
537
+ "learning_rate": 4.444426793932787e-05,
538
+ "loss": 1.8856,
539
+ "step": 83
540
+ },
541
+ {
542
+ "epoch": 0.83,
543
+ "learning_rate": 4.3948126557440085e-05,
544
+ "loss": 2.0904,
545
+ "step": 84
546
+ },
547
+ {
548
+ "epoch": 0.84,
549
+ "learning_rate": 4.344892023524714e-05,
550
+ "loss": 2.0685,
551
+ "step": 85
552
+ },
553
+ {
554
+ "epoch": 0.85,
555
+ "learning_rate": 4.2946782622352746e-05,
556
+ "loss": 2.1,
557
+ "step": 86
558
+ },
559
+ {
560
+ "epoch": 0.86,
561
+ "learning_rate": 4.244184815313799e-05,
562
+ "loss": 1.9322,
563
+ "step": 87
564
+ },
565
+ {
566
+ "epoch": 0.87,
567
+ "learning_rate": 4.193425201077002e-05,
568
+ "loss": 2.194,
569
+ "step": 88
570
+ },
571
+ {
572
+ "epoch": 0.88,
573
+ "learning_rate": 4.14241300910103e-05,
574
+ "loss": 2.0455,
575
+ "step": 89
576
+ },
577
+ {
578
+ "epoch": 0.89,
579
+ "learning_rate": 4.091161896583192e-05,
580
+ "loss": 2.1239,
581
+ "step": 90
582
+ },
583
+ {
584
+ "epoch": 0.9,
585
+ "learning_rate": 4.0396855846856076e-05,
586
+ "loss": 2.142,
587
+ "step": 91
588
+ },
589
+ {
590
+ "epoch": 0.91,
591
+ "learning_rate": 3.9879978548617125e-05,
592
+ "loss": 2.2304,
593
+ "step": 92
594
+ },
595
+ {
596
+ "epoch": 0.92,
597
+ "learning_rate": 3.936112545166636e-05,
598
+ "loss": 1.925,
599
+ "step": 93
600
+ },
601
+ {
602
+ "epoch": 0.93,
603
+ "learning_rate": 3.884043546552417e-05,
604
+ "loss": 2.1547,
605
+ "step": 94
606
+ },
607
+ {
608
+ "epoch": 0.94,
609
+ "learning_rate": 3.831804799149057e-05,
610
+ "loss": 2.0439,
611
+ "step": 95
612
+ },
613
+ {
614
+ "epoch": 0.95,
615
+ "learning_rate": 3.779410288532413e-05,
616
+ "loss": 1.9763,
617
+ "step": 96
618
+ },
619
+ {
620
+ "epoch": 0.96,
621
+ "learning_rate": 3.726874041979925e-05,
622
+ "loss": 2.114,
623
+ "step": 97
624
+ },
625
+ {
626
+ "epoch": 0.97,
627
+ "learning_rate": 3.674210124715168e-05,
628
+ "loss": 2.0293,
629
+ "step": 98
630
+ },
631
+ {
632
+ "epoch": 0.98,
633
+ "learning_rate": 3.621432636142251e-05,
634
+ "loss": 2.0833,
635
+ "step": 99
636
+ },
637
+ {
638
+ "epoch": 0.99,
639
+ "learning_rate": 3.5685557060710726e-05,
640
+ "loss": 2.0452,
641
+ "step": 100
642
+ },
643
+ {
644
+ "epoch": 0.99,
645
+ "eval_loss": 1.9682413339614868,
646
+ "eval_runtime": 1.4934,
647
+ "eval_samples_per_second": 5.357,
648
+ "eval_steps_per_second": 0.67,
649
+ "step": 100
650
+ }
651
+ ],
652
+ "logging_steps": 1,
653
+ "max_steps": 202,
654
+ "num_train_epochs": 2,
655
+ "save_steps": 50,
656
+ "total_flos": 8.359599153348608e+16,
657
+ "trial_name": null,
658
+ "trial_params": null
659
+ }
multi-prompt/twochar/checkpoint-100/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac3fb199292750362f1e387a5d8bde7088e81478bbe173194e512403378f1c70
3
+ size 4475
multi-prompt/twochar/checkpoint-150/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 0.6.0.dev0
multi-prompt/twochar/checkpoint-150/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "meta-llama/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": 0.01,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "v_proj",
18
+ "q_proj",
19
+ "down_proj",
20
+ "k_proj",
21
+ "gate_proj",
22
+ "o_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
multi-prompt/twochar/checkpoint-150/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:94befd076d47fa24e3ba916e9a3778dd8ece4972bb7c203083cc5f53c5e46bfa
3
+ size 125374989
multi-prompt/twochar/checkpoint-150/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 0.6.0.dev0
multi-prompt/twochar/checkpoint-150/adapter_model/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "meta-llama/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": 0.01,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "v_proj",
18
+ "q_proj",
19
+ "down_proj",
20
+ "k_proj",
21
+ "gate_proj",
22
+ "o_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
multi-prompt/twochar/checkpoint-150/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:94befd076d47fa24e3ba916e9a3778dd8ece4972bb7c203083cc5f53c5e46bfa
3
+ size 125374989
multi-prompt/twochar/checkpoint-150/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60fd84426c8963b86174be0513aea26512fafa4a79de37a27d66e42699c92c1d
3
+ size 250681597
multi-prompt/twochar/checkpoint-150/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d127d4391ee6771b6d20db8ce3f0dd0584c753057cc9bdfdd0be6f8cef84d955
3
+ size 21687
multi-prompt/twochar/checkpoint-150/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d7ccf1beada17255706986caf9094f244684163a8547f833874af8de8dd4daa0
3
+ size 21687
multi-prompt/twochar/checkpoint-150/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:277501f7b2a65d7470e01413187a94540bcb41c51790d63ac495be26b8fdfb89
3
+ size 21687
multi-prompt/twochar/checkpoint-150/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1d55ef5ebc7e10e4cd63d908d65f05dfdc0de99fd4911bc3aa2e57e01940ef54
3
+ size 21687
multi-prompt/twochar/checkpoint-150/rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ddc9393cb55ab7d202c2cc2ae933435652c478277daf312e6370eb3404db7e9f
3
+ size 21687
multi-prompt/twochar/checkpoint-150/rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fcbf8760f23fffddb8debdb97db5712925a0c049476c45b8b57488cb099ff82e
3
+ size 21687
multi-prompt/twochar/checkpoint-150/rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:db101559fa9579ad6b9ca8b4a109575c11fc03a27ef55f3e0bb92c5642fc4001
3
+ size 21687
multi-prompt/twochar/checkpoint-150/rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:644bacb2bf138ffa7d1d20a3358a4fab98d6ab8c6ede06df95e5de696a31d2ce
3
+ size 21687
multi-prompt/twochar/checkpoint-150/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c4e06f3498f4a96af42d371e2251f32c6afc73dc33f11f9b431ef7f76337ee64
3
+ size 627
multi-prompt/twochar/checkpoint-150/trainer_state.json ADDED
@@ -0,0 +1,975 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.4851485148514851,
5
+ "eval_steps": 20,
6
+ "global_step": 150,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 6.5e-06,
14
+ "loss": 2.0407,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.02,
19
+ "learning_rate": 1.3e-05,
20
+ "loss": 2.3107,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.03,
25
+ "learning_rate": 1.9499999999999996e-05,
26
+ "loss": 2.1613,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.04,
31
+ "learning_rate": 2.6e-05,
32
+ "loss": 2.0444,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.05,
37
+ "learning_rate": 3.25e-05,
38
+ "loss": 2.2512,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.06,
43
+ "learning_rate": 3.899999999999999e-05,
44
+ "loss": 2.0516,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.07,
49
+ "learning_rate": 4.5499999999999995e-05,
50
+ "loss": 2.2006,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.08,
55
+ "learning_rate": 5.2e-05,
56
+ "loss": 2.1229,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.09,
61
+ "learning_rate": 5.85e-05,
62
+ "loss": 2.2498,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.1,
67
+ "learning_rate": 6.5e-05,
68
+ "loss": 2.0293,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.11,
73
+ "learning_rate": 6.499564948206075e-05,
74
+ "loss": 2.1316,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.12,
79
+ "learning_rate": 6.498259909298188e-05,
80
+ "loss": 2.1274,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.13,
85
+ "learning_rate": 6.49608523266681e-05,
86
+ "loss": 2.1129,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.14,
91
+ "learning_rate": 6.493041500525461e-05,
92
+ "loss": 2.1517,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.15,
97
+ "learning_rate": 6.489129527754836e-05,
98
+ "loss": 2.1078,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.16,
103
+ "learning_rate": 6.484350361684639e-05,
104
+ "loss": 2.0362,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.17,
109
+ "learning_rate": 6.478705281813194e-05,
110
+ "loss": 2.0553,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 0.18,
115
+ "learning_rate": 6.472195799464884e-05,
116
+ "loss": 2.1223,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 0.19,
121
+ "learning_rate": 6.464823657385538e-05,
122
+ "loss": 2.0326,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 0.2,
127
+ "learning_rate": 6.456590829275857e-05,
128
+ "loss": 2.1802,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 0.2,
133
+ "eval_loss": 2.012301445007324,
134
+ "eval_runtime": 1.4855,
135
+ "eval_samples_per_second": 5.385,
136
+ "eval_steps_per_second": 0.673,
137
+ "step": 20
138
+ },
139
+ {
140
+ "epoch": 0.21,
141
+ "learning_rate": 6.447499519263001e-05,
142
+ "loss": 2.0665,
143
+ "step": 21
144
+ },
145
+ {
146
+ "epoch": 0.22,
147
+ "learning_rate": 6.437552161310498e-05,
148
+ "loss": 2.0354,
149
+ "step": 22
150
+ },
151
+ {
152
+ "epoch": 0.23,
153
+ "learning_rate": 6.426751418566609e-05,
154
+ "loss": 2.1289,
155
+ "step": 23
156
+ },
157
+ {
158
+ "epoch": 0.24,
159
+ "learning_rate": 6.415100182651334e-05,
160
+ "loss": 2.0742,
161
+ "step": 24
162
+ },
163
+ {
164
+ "epoch": 0.25,
165
+ "learning_rate": 6.402601572882268e-05,
166
+ "loss": 2.0412,
167
+ "step": 25
168
+ },
169
+ {
170
+ "epoch": 0.26,
171
+ "learning_rate": 6.38925893543947e-05,
172
+ "loss": 2.1974,
173
+ "step": 26
174
+ },
175
+ {
176
+ "epoch": 0.27,
177
+ "learning_rate": 6.375075842469626e-05,
178
+ "loss": 2.196,
179
+ "step": 27
180
+ },
181
+ {
182
+ "epoch": 0.28,
183
+ "learning_rate": 6.360056091129678e-05,
184
+ "loss": 2.0775,
185
+ "step": 28
186
+ },
187
+ {
188
+ "epoch": 0.29,
189
+ "learning_rate": 6.344203702570254e-05,
190
+ "loss": 2.0366,
191
+ "step": 29
192
+ },
193
+ {
194
+ "epoch": 0.3,
195
+ "learning_rate": 6.327522920859093e-05,
196
+ "loss": 2.1438,
197
+ "step": 30
198
+ },
199
+ {
200
+ "epoch": 0.31,
201
+ "learning_rate": 6.310018211844817e-05,
202
+ "loss": 2.0639,
203
+ "step": 31
204
+ },
205
+ {
206
+ "epoch": 0.32,
207
+ "learning_rate": 6.291694261961308e-05,
208
+ "loss": 2.1794,
209
+ "step": 32
210
+ },
211
+ {
212
+ "epoch": 0.33,
213
+ "learning_rate": 6.272555976973039e-05,
214
+ "loss": 2.0592,
215
+ "step": 33
216
+ },
217
+ {
218
+ "epoch": 0.34,
219
+ "learning_rate": 6.252608480661681e-05,
220
+ "loss": 2.1628,
221
+ "step": 34
222
+ },
223
+ {
224
+ "epoch": 0.35,
225
+ "learning_rate": 6.231857113454346e-05,
226
+ "loss": 2.1851,
227
+ "step": 35
228
+ },
229
+ {
230
+ "epoch": 0.36,
231
+ "learning_rate": 6.21030743099382e-05,
232
+ "loss": 2.0531,
233
+ "step": 36
234
+ },
235
+ {
236
+ "epoch": 0.37,
237
+ "learning_rate": 6.18796520265119e-05,
238
+ "loss": 2.0284,
239
+ "step": 37
240
+ },
241
+ {
242
+ "epoch": 0.38,
243
+ "learning_rate": 6.164836409981237e-05,
244
+ "loss": 2.2359,
245
+ "step": 38
246
+ },
247
+ {
248
+ "epoch": 0.39,
249
+ "learning_rate": 6.140927245121032e-05,
250
+ "loss": 1.8967,
251
+ "step": 39
252
+ },
253
+ {
254
+ "epoch": 0.4,
255
+ "learning_rate": 6.116244109132153e-05,
256
+ "loss": 2.0275,
257
+ "step": 40
258
+ },
259
+ {
260
+ "epoch": 0.4,
261
+ "eval_loss": 1.9861761331558228,
262
+ "eval_runtime": 1.4896,
263
+ "eval_samples_per_second": 5.37,
264
+ "eval_steps_per_second": 0.671,
265
+ "step": 40
266
+ },
267
+ {
268
+ "epoch": 0.41,
269
+ "learning_rate": 6.0907936102869656e-05,
270
+ "loss": 2.0316,
271
+ "step": 41
272
+ },
273
+ {
274
+ "epoch": 0.42,
275
+ "learning_rate": 6.0645825622994254e-05,
276
+ "loss": 2.0768,
277
+ "step": 42
278
+ },
279
+ {
280
+ "epoch": 0.43,
281
+ "learning_rate": 6.037617982500884e-05,
282
+ "loss": 2.0058,
283
+ "step": 43
284
+ },
285
+ {
286
+ "epoch": 0.44,
287
+ "learning_rate": 6.009907089961381e-05,
288
+ "loss": 2.2319,
289
+ "step": 44
290
+ },
291
+ {
292
+ "epoch": 0.45,
293
+ "learning_rate": 5.981457303556923e-05,
294
+ "loss": 2.1693,
295
+ "step": 45
296
+ },
297
+ {
298
+ "epoch": 0.46,
299
+ "learning_rate": 5.9522762399832716e-05,
300
+ "loss": 2.0858,
301
+ "step": 46
302
+ },
303
+ {
304
+ "epoch": 0.47,
305
+ "learning_rate": 5.922371711716768e-05,
306
+ "loss": 2.0576,
307
+ "step": 47
308
+ },
309
+ {
310
+ "epoch": 0.48,
311
+ "learning_rate": 5.891751724922749e-05,
312
+ "loss": 1.9946,
313
+ "step": 48
314
+ },
315
+ {
316
+ "epoch": 0.49,
317
+ "learning_rate": 5.860424477312095e-05,
318
+ "loss": 2.0166,
319
+ "step": 49
320
+ },
321
+ {
322
+ "epoch": 0.5,
323
+ "learning_rate": 5.828398355946514e-05,
324
+ "loss": 2.1482,
325
+ "step": 50
326
+ },
327
+ {
328
+ "epoch": 0.5,
329
+ "learning_rate": 5.795681934993113e-05,
330
+ "loss": 2.0698,
331
+ "step": 51
332
+ },
333
+ {
334
+ "epoch": 0.51,
335
+ "learning_rate": 5.7622839734288945e-05,
336
+ "loss": 1.9317,
337
+ "step": 52
338
+ },
339
+ {
340
+ "epoch": 0.52,
341
+ "learning_rate": 5.728213412695761e-05,
342
+ "loss": 2.1399,
343
+ "step": 53
344
+ },
345
+ {
346
+ "epoch": 0.53,
347
+ "learning_rate": 5.693479374306676e-05,
348
+ "loss": 2.1637,
349
+ "step": 54
350
+ },
351
+ {
352
+ "epoch": 0.54,
353
+ "learning_rate": 5.658091157403617e-05,
354
+ "loss": 2.0557,
355
+ "step": 55
356
+ },
357
+ {
358
+ "epoch": 0.55,
359
+ "learning_rate": 5.622058236267965e-05,
360
+ "loss": 1.926,
361
+ "step": 56
362
+ },
363
+ {
364
+ "epoch": 0.56,
365
+ "learning_rate": 5.585390257784018e-05,
366
+ "loss": 2.0984,
367
+ "step": 57
368
+ },
369
+ {
370
+ "epoch": 0.57,
371
+ "learning_rate": 5.548097038856279e-05,
372
+ "loss": 2.0428,
373
+ "step": 58
374
+ },
375
+ {
376
+ "epoch": 0.58,
377
+ "learning_rate": 5.5101885637812374e-05,
378
+ "loss": 1.9624,
379
+ "step": 59
380
+ },
381
+ {
382
+ "epoch": 0.59,
383
+ "learning_rate": 5.4716749815743304e-05,
384
+ "loss": 1.913,
385
+ "step": 60
386
+ },
387
+ {
388
+ "epoch": 0.59,
389
+ "eval_loss": 1.97765052318573,
390
+ "eval_runtime": 1.4886,
391
+ "eval_samples_per_second": 5.374,
392
+ "eval_steps_per_second": 0.672,
393
+ "step": 60
394
+ },
395
+ {
396
+ "epoch": 0.6,
397
+ "learning_rate": 5.432566603252809e-05,
398
+ "loss": 2.0125,
399
+ "step": 61
400
+ },
401
+ {
402
+ "epoch": 0.61,
403
+ "learning_rate": 5.3928738990752234e-05,
404
+ "loss": 2.0721,
405
+ "step": 62
406
+ },
407
+ {
408
+ "epoch": 0.62,
409
+ "learning_rate": 5.3526074957382866e-05,
410
+ "loss": 2.1046,
411
+ "step": 63
412
+ },
413
+ {
414
+ "epoch": 0.63,
415
+ "learning_rate": 5.311778173531847e-05,
416
+ "loss": 2.0074,
417
+ "step": 64
418
+ },
419
+ {
420
+ "epoch": 0.64,
421
+ "learning_rate": 5.2703968634527514e-05,
422
+ "loss": 2.1667,
423
+ "step": 65
424
+ },
425
+ {
426
+ "epoch": 0.65,
427
+ "learning_rate": 5.2284746442783414e-05,
428
+ "loss": 2.0677,
429
+ "step": 66
430
+ },
431
+ {
432
+ "epoch": 0.66,
433
+ "learning_rate": 5.186022739600408e-05,
434
+ "loss": 2.1798,
435
+ "step": 67
436
+ },
437
+ {
438
+ "epoch": 0.67,
439
+ "learning_rate": 5.143052514820357e-05,
440
+ "loss": 2.105,
441
+ "step": 68
442
+ },
443
+ {
444
+ "epoch": 0.68,
445
+ "learning_rate": 5.099575474106419e-05,
446
+ "loss": 2.118,
447
+ "step": 69
448
+ },
449
+ {
450
+ "epoch": 0.69,
451
+ "learning_rate": 5.055603257313707e-05,
452
+ "loss": 2.0791,
453
+ "step": 70
454
+ },
455
+ {
456
+ "epoch": 0.7,
457
+ "learning_rate": 5.011147636867943e-05,
458
+ "loss": 2.032,
459
+ "step": 71
460
+ },
461
+ {
462
+ "epoch": 0.71,
463
+ "learning_rate": 4.9662205146136955e-05,
464
+ "loss": 1.9331,
465
+ "step": 72
466
+ },
467
+ {
468
+ "epoch": 0.72,
469
+ "learning_rate": 4.92083391862797e-05,
470
+ "loss": 2.0077,
471
+ "step": 73
472
+ },
473
+ {
474
+ "epoch": 0.73,
475
+ "learning_rate": 4.874999999999999e-05,
476
+ "loss": 2.0253,
477
+ "step": 74
478
+ },
479
+ {
480
+ "epoch": 0.74,
481
+ "learning_rate": 4.828731029578105e-05,
482
+ "loss": 1.9922,
483
+ "step": 75
484
+ },
485
+ {
486
+ "epoch": 0.75,
487
+ "learning_rate": 4.7820393946844926e-05,
488
+ "loss": 2.0295,
489
+ "step": 76
490
+ },
491
+ {
492
+ "epoch": 0.76,
493
+ "learning_rate": 4.734937595798867e-05,
494
+ "loss": 1.9514,
495
+ "step": 77
496
+ },
497
+ {
498
+ "epoch": 0.77,
499
+ "learning_rate": 4.687438243211754e-05,
500
+ "loss": 2.1018,
501
+ "step": 78
502
+ },
503
+ {
504
+ "epoch": 0.78,
505
+ "learning_rate": 4.639554053648416e-05,
506
+ "loss": 2.1032,
507
+ "step": 79
508
+ },
509
+ {
510
+ "epoch": 0.79,
511
+ "learning_rate": 4.5912978468642824e-05,
512
+ "loss": 2.1258,
513
+ "step": 80
514
+ },
515
+ {
516
+ "epoch": 0.79,
517
+ "eval_loss": 1.972991704940796,
518
+ "eval_runtime": 1.494,
519
+ "eval_samples_per_second": 5.355,
520
+ "eval_steps_per_second": 0.669,
521
+ "step": 80
522
+ },
523
+ {
524
+ "epoch": 0.8,
525
+ "learning_rate": 4.542682542212785e-05,
526
+ "loss": 2.0353,
527
+ "step": 81
528
+ },
529
+ {
530
+ "epoch": 0.81,
531
+ "learning_rate": 4.4937211551865415e-05,
532
+ "loss": 1.9987,
533
+ "step": 82
534
+ },
535
+ {
536
+ "epoch": 0.82,
537
+ "learning_rate": 4.444426793932787e-05,
538
+ "loss": 1.8856,
539
+ "step": 83
540
+ },
541
+ {
542
+ "epoch": 0.83,
543
+ "learning_rate": 4.3948126557440085e-05,
544
+ "loss": 2.0904,
545
+ "step": 84
546
+ },
547
+ {
548
+ "epoch": 0.84,
549
+ "learning_rate": 4.344892023524714e-05,
550
+ "loss": 2.0685,
551
+ "step": 85
552
+ },
553
+ {
554
+ "epoch": 0.85,
555
+ "learning_rate": 4.2946782622352746e-05,
556
+ "loss": 2.1,
557
+ "step": 86
558
+ },
559
+ {
560
+ "epoch": 0.86,
561
+ "learning_rate": 4.244184815313799e-05,
562
+ "loss": 1.9322,
563
+ "step": 87
564
+ },
565
+ {
566
+ "epoch": 0.87,
567
+ "learning_rate": 4.193425201077002e-05,
568
+ "loss": 2.194,
569
+ "step": 88
570
+ },
571
+ {
572
+ "epoch": 0.88,
573
+ "learning_rate": 4.14241300910103e-05,
574
+ "loss": 2.0455,
575
+ "step": 89
576
+ },
577
+ {
578
+ "epoch": 0.89,
579
+ "learning_rate": 4.091161896583192e-05,
580
+ "loss": 2.1239,
581
+ "step": 90
582
+ },
583
+ {
584
+ "epoch": 0.9,
585
+ "learning_rate": 4.0396855846856076e-05,
586
+ "loss": 2.142,
587
+ "step": 91
588
+ },
589
+ {
590
+ "epoch": 0.91,
591
+ "learning_rate": 3.9879978548617125e-05,
592
+ "loss": 2.2304,
593
+ "step": 92
594
+ },
595
+ {
596
+ "epoch": 0.92,
597
+ "learning_rate": 3.936112545166636e-05,
598
+ "loss": 1.925,
599
+ "step": 93
600
+ },
601
+ {
602
+ "epoch": 0.93,
603
+ "learning_rate": 3.884043546552417e-05,
604
+ "loss": 2.1547,
605
+ "step": 94
606
+ },
607
+ {
608
+ "epoch": 0.94,
609
+ "learning_rate": 3.831804799149057e-05,
610
+ "loss": 2.0439,
611
+ "step": 95
612
+ },
613
+ {
614
+ "epoch": 0.95,
615
+ "learning_rate": 3.779410288532413e-05,
616
+ "loss": 1.9763,
617
+ "step": 96
618
+ },
619
+ {
620
+ "epoch": 0.96,
621
+ "learning_rate": 3.726874041979925e-05,
622
+ "loss": 2.114,
623
+ "step": 97
624
+ },
625
+ {
626
+ "epoch": 0.97,
627
+ "learning_rate": 3.674210124715168e-05,
628
+ "loss": 2.0293,
629
+ "step": 98
630
+ },
631
+ {
632
+ "epoch": 0.98,
633
+ "learning_rate": 3.621432636142251e-05,
634
+ "loss": 2.0833,
635
+ "step": 99
636
+ },
637
+ {
638
+ "epoch": 0.99,
639
+ "learning_rate": 3.5685557060710726e-05,
640
+ "loss": 2.0452,
641
+ "step": 100
642
+ },
643
+ {
644
+ "epoch": 0.99,
645
+ "eval_loss": 1.9682413339614868,
646
+ "eval_runtime": 1.4934,
647
+ "eval_samples_per_second": 5.357,
648
+ "eval_steps_per_second": 0.67,
649
+ "step": 100
650
+ },
651
+ {
652
+ "epoch": 1.0,
653
+ "learning_rate": 3.5155934909344214e-05,
654
+ "loss": 1.9682,
655
+ "step": 101
656
+ },
657
+ {
658
+ "epoch": 1.01,
659
+ "learning_rate": 3.462560169997965e-05,
660
+ "loss": 2.0723,
661
+ "step": 102
662
+ },
663
+ {
664
+ "epoch": 1.02,
665
+ "learning_rate": 3.409469941564109e-05,
666
+ "loss": 2.0713,
667
+ "step": 103
668
+ },
669
+ {
670
+ "epoch": 1.03,
671
+ "learning_rate": 3.356337019170772e-05,
672
+ "loss": 2.067,
673
+ "step": 104
674
+ },
675
+ {
676
+ "epoch": 1.04,
677
+ "learning_rate": 3.303175627786082e-05,
678
+ "loss": 1.934,
679
+ "step": 105
680
+ },
681
+ {
682
+ "epoch": 1.05,
683
+ "learning_rate": 3.25e-05,
684
+ "loss": 2.2083,
685
+ "step": 106
686
+ },
687
+ {
688
+ "epoch": 1.06,
689
+ "learning_rate": 3.1968243722139173e-05,
690
+ "loss": 2.045,
691
+ "step": 107
692
+ },
693
+ {
694
+ "epoch": 1.07,
695
+ "learning_rate": 3.143662980829228e-05,
696
+ "loss": 1.9531,
697
+ "step": 108
698
+ },
699
+ {
700
+ "epoch": 1.08,
701
+ "learning_rate": 3.090530058435891e-05,
702
+ "loss": 2.0913,
703
+ "step": 109
704
+ },
705
+ {
706
+ "epoch": 1.09,
707
+ "learning_rate": 3.0374398300020347e-05,
708
+ "loss": 1.9211,
709
+ "step": 110
710
+ },
711
+ {
712
+ "epoch": 1.1,
713
+ "learning_rate": 2.984406509065579e-05,
714
+ "loss": 2.0983,
715
+ "step": 111
716
+ },
717
+ {
718
+ "epoch": 1.11,
719
+ "learning_rate": 2.9314442939289275e-05,
720
+ "loss": 2.1621,
721
+ "step": 112
722
+ },
723
+ {
724
+ "epoch": 1.12,
725
+ "learning_rate": 2.8785673638577486e-05,
726
+ "loss": 2.1728,
727
+ "step": 113
728
+ },
729
+ {
730
+ "epoch": 1.13,
731
+ "learning_rate": 2.825789875284833e-05,
732
+ "loss": 2.0512,
733
+ "step": 114
734
+ },
735
+ {
736
+ "epoch": 1.14,
737
+ "learning_rate": 2.773125958020074e-05,
738
+ "loss": 2.0462,
739
+ "step": 115
740
+ },
741
+ {
742
+ "epoch": 1.15,
743
+ "learning_rate": 2.7205897114675866e-05,
744
+ "loss": 1.9357,
745
+ "step": 116
746
+ },
747
+ {
748
+ "epoch": 1.16,
749
+ "learning_rate": 2.668195200850944e-05,
750
+ "loss": 1.8384,
751
+ "step": 117
752
+ },
753
+ {
754
+ "epoch": 1.17,
755
+ "learning_rate": 2.6159564534475832e-05,
756
+ "loss": 1.9649,
757
+ "step": 118
758
+ },
759
+ {
760
+ "epoch": 1.18,
761
+ "learning_rate": 2.563887454833363e-05,
762
+ "loss": 2.0262,
763
+ "step": 119
764
+ },
765
+ {
766
+ "epoch": 1.19,
767
+ "learning_rate": 2.5120021451382872e-05,
768
+ "loss": 2.1315,
769
+ "step": 120
770
+ },
771
+ {
772
+ "epoch": 1.19,
773
+ "eval_loss": 1.9639294147491455,
774
+ "eval_runtime": 1.4959,
775
+ "eval_samples_per_second": 5.348,
776
+ "eval_steps_per_second": 0.668,
777
+ "step": 120
778
+ },
779
+ {
780
+ "epoch": 1.2,
781
+ "learning_rate": 2.4603144153143925e-05,
782
+ "loss": 1.9284,
783
+ "step": 121
784
+ },
785
+ {
786
+ "epoch": 1.21,
787
+ "learning_rate": 2.4088381034168072e-05,
788
+ "loss": 2.0671,
789
+ "step": 122
790
+ },
791
+ {
792
+ "epoch": 1.22,
793
+ "learning_rate": 2.3575869908989705e-05,
794
+ "loss": 2.0938,
795
+ "step": 123
796
+ },
797
+ {
798
+ "epoch": 1.23,
799
+ "learning_rate": 2.306574798922998e-05,
800
+ "loss": 1.9189,
801
+ "step": 124
802
+ },
803
+ {
804
+ "epoch": 1.24,
805
+ "learning_rate": 2.2558151846862005e-05,
806
+ "loss": 2.0635,
807
+ "step": 125
808
+ },
809
+ {
810
+ "epoch": 1.25,
811
+ "learning_rate": 2.2053217377647255e-05,
812
+ "loss": 2.1367,
813
+ "step": 126
814
+ },
815
+ {
816
+ "epoch": 1.26,
817
+ "learning_rate": 2.1551079764752848e-05,
818
+ "loss": 1.9967,
819
+ "step": 127
820
+ },
821
+ {
822
+ "epoch": 1.27,
823
+ "learning_rate": 2.105187344255991e-05,
824
+ "loss": 2.1994,
825
+ "step": 128
826
+ },
827
+ {
828
+ "epoch": 1.28,
829
+ "learning_rate": 2.0555732060672138e-05,
830
+ "loss": 2.151,
831
+ "step": 129
832
+ },
833
+ {
834
+ "epoch": 1.29,
835
+ "learning_rate": 2.0062788448134583e-05,
836
+ "loss": 2.0517,
837
+ "step": 130
838
+ },
839
+ {
840
+ "epoch": 1.3,
841
+ "learning_rate": 1.957317457787214e-05,
842
+ "loss": 2.0584,
843
+ "step": 131
844
+ },
845
+ {
846
+ "epoch": 1.31,
847
+ "learning_rate": 1.908702153135717e-05,
848
+ "loss": 2.0868,
849
+ "step": 132
850
+ },
851
+ {
852
+ "epoch": 1.32,
853
+ "learning_rate": 1.860445946351584e-05,
854
+ "loss": 1.9646,
855
+ "step": 133
856
+ },
857
+ {
858
+ "epoch": 1.33,
859
+ "learning_rate": 1.8125617567882463e-05,
860
+ "loss": 1.986,
861
+ "step": 134
862
+ },
863
+ {
864
+ "epoch": 1.34,
865
+ "learning_rate": 1.7650624042011325e-05,
866
+ "loss": 1.9015,
867
+ "step": 135
868
+ },
869
+ {
870
+ "epoch": 1.35,
871
+ "learning_rate": 1.7179606053155072e-05,
872
+ "loss": 2.1968,
873
+ "step": 136
874
+ },
875
+ {
876
+ "epoch": 1.36,
877
+ "learning_rate": 1.6712689704218944e-05,
878
+ "loss": 1.9268,
879
+ "step": 137
880
+ },
881
+ {
882
+ "epoch": 1.37,
883
+ "learning_rate": 1.6250000000000005e-05,
884
+ "loss": 2.0205,
885
+ "step": 138
886
+ },
887
+ {
888
+ "epoch": 1.38,
889
+ "learning_rate": 1.5791660813720294e-05,
890
+ "loss": 2.1183,
891
+ "step": 139
892
+ },
893
+ {
894
+ "epoch": 1.39,
895
+ "learning_rate": 1.533779485386304e-05,
896
+ "loss": 2.08,
897
+ "step": 140
898
+ },
899
+ {
900
+ "epoch": 1.39,
901
+ "eval_loss": 1.9603089094161987,
902
+ "eval_runtime": 1.495,
903
+ "eval_samples_per_second": 5.351,
904
+ "eval_steps_per_second": 0.669,
905
+ "step": 140
906
+ },
907
+ {
908
+ "epoch": 1.4,
909
+ "learning_rate": 1.4888523631320579e-05,
910
+ "loss": 1.9309,
911
+ "step": 141
912
+ },
913
+ {
914
+ "epoch": 1.41,
915
+ "learning_rate": 1.4443967426862935e-05,
916
+ "loss": 2.0579,
917
+ "step": 142
918
+ },
919
+ {
920
+ "epoch": 1.42,
921
+ "learning_rate": 1.4004245258935799e-05,
922
+ "loss": 1.8377,
923
+ "step": 143
924
+ },
925
+ {
926
+ "epoch": 1.43,
927
+ "learning_rate": 1.3569474851796432e-05,
928
+ "loss": 2.0764,
929
+ "step": 144
930
+ },
931
+ {
932
+ "epoch": 1.44,
933
+ "learning_rate": 1.3139772603995914e-05,
934
+ "loss": 2.1289,
935
+ "step": 145
936
+ },
937
+ {
938
+ "epoch": 1.45,
939
+ "learning_rate": 1.2715253557216577e-05,
940
+ "loss": 2.041,
941
+ "step": 146
942
+ },
943
+ {
944
+ "epoch": 1.46,
945
+ "learning_rate": 1.2296031365472491e-05,
946
+ "loss": 1.9343,
947
+ "step": 147
948
+ },
949
+ {
950
+ "epoch": 1.47,
951
+ "learning_rate": 1.1882218264681525e-05,
952
+ "loss": 1.9665,
953
+ "step": 148
954
+ },
955
+ {
956
+ "epoch": 1.48,
957
+ "learning_rate": 1.1473925042617137e-05,
958
+ "loss": 2.1843,
959
+ "step": 149
960
+ },
961
+ {
962
+ "epoch": 1.49,
963
+ "learning_rate": 1.1071261009247762e-05,
964
+ "loss": 2.0125,
965
+ "step": 150
966
+ }
967
+ ],
968
+ "logging_steps": 1,
969
+ "max_steps": 202,
970
+ "num_train_epochs": 2,
971
+ "save_steps": 50,
972
+ "total_flos": 1.2497211511065805e+17,
973
+ "trial_name": null,
974
+ "trial_params": null
975
+ }
multi-prompt/twochar/checkpoint-150/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac3fb199292750362f1e387a5d8bde7088e81478bbe173194e512403378f1c70
3
+ size 4475
multi-prompt/twochar/checkpoint-200/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 0.6.0.dev0
multi-prompt/twochar/checkpoint-200/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "meta-llama/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": 0.01,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "v_proj",
18
+ "q_proj",
19
+ "down_proj",
20
+ "k_proj",
21
+ "gate_proj",
22
+ "o_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
multi-prompt/twochar/checkpoint-200/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c882f2a426060ad44b5cf280eb92814d684c2673107ca7f19c9e34b4c3fe2980
3
+ size 125374989
multi-prompt/twochar/checkpoint-200/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 0.6.0.dev0
multi-prompt/twochar/checkpoint-200/adapter_model/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "meta-llama/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": 0.01,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "v_proj",
18
+ "q_proj",
19
+ "down_proj",
20
+ "k_proj",
21
+ "gate_proj",
22
+ "o_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
multi-prompt/twochar/checkpoint-200/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c882f2a426060ad44b5cf280eb92814d684c2673107ca7f19c9e34b4c3fe2980
3
+ size 125374989
multi-prompt/twochar/checkpoint-200/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:16b546d91cb67c74c723898ad8350b04c2ee27be0612fff40324b5db78dd87a0
3
+ size 250681597
multi-prompt/twochar/checkpoint-200/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:19c4ecd007574b31ef585631572d90a849b7178b232262c2c41125e361bc3517
3
+ size 21687
multi-prompt/twochar/checkpoint-200/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3cc05c6dead1d35e2b0a4b8c8ca26e5b7aeb7547f432cc1c89591900a46a09b1
3
+ size 21687
multi-prompt/twochar/checkpoint-200/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:10c2d8765af7ee78c7ddacf8d9c6ad56309da3711e52aa502ca9e611e8b23df8
3
+ size 21687