Hugging Face's logo

royallab
/
LimaRP-ShareGPT-13b-qloras

PEFT
Not-For-All-Audiences
Model card Files Community
2
kingbri commited on
Commit
8f823de
·
1 Parent(s): 1b62f17

Add prompt-b 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. prompt-b/twochar/README.md +21 -0
  2. prompt-b/twochar/adapter_config.json +26 -0
  3. prompt-b/twochar/adapter_model.bin +3 -0
  4. prompt-b/twochar/added_tokens.json +3 -0
  5. prompt-b/twochar/checkpoint-100/README.md +21 -0
  6. prompt-b/twochar/checkpoint-100/adapter_config.json +26 -0
  7. prompt-b/twochar/checkpoint-100/adapter_model.bin +3 -0
  8. prompt-b/twochar/checkpoint-100/adapter_model/README.md +21 -0
  9. prompt-b/twochar/checkpoint-100/adapter_model/adapter_config.json +26 -0
  10. prompt-b/twochar/checkpoint-100/adapter_model/adapter_model.bin +3 -0
  11. prompt-b/twochar/checkpoint-100/optimizer.pt +3 -0
  12. prompt-b/twochar/checkpoint-100/rng_state_0.pth +3 -0
  13. prompt-b/twochar/checkpoint-100/rng_state_1.pth +3 -0
  14. prompt-b/twochar/checkpoint-100/rng_state_2.pth +3 -0
  15. prompt-b/twochar/checkpoint-100/rng_state_3.pth +3 -0
  16. prompt-b/twochar/checkpoint-100/rng_state_4.pth +3 -0
  17. prompt-b/twochar/checkpoint-100/rng_state_5.pth +3 -0
  18. prompt-b/twochar/checkpoint-100/rng_state_6.pth +3 -0
  19. prompt-b/twochar/checkpoint-100/rng_state_7.pth +3 -0
  20. prompt-b/twochar/checkpoint-100/scheduler.pt +3 -0
  21. prompt-b/twochar/checkpoint-100/trainer_state.json +659 -0
  22. prompt-b/twochar/checkpoint-100/training_args.bin +3 -0
  23. prompt-b/twochar/checkpoint-150/README.md +21 -0
  24. prompt-b/twochar/checkpoint-150/adapter_config.json +26 -0
  25. prompt-b/twochar/checkpoint-150/adapter_model.bin +3 -0
  26. prompt-b/twochar/checkpoint-150/adapter_model/README.md +21 -0
  27. prompt-b/twochar/checkpoint-150/adapter_model/adapter_config.json +26 -0
  28. prompt-b/twochar/checkpoint-150/adapter_model/adapter_model.bin +3 -0
  29. prompt-b/twochar/checkpoint-150/optimizer.pt +3 -0
  30. prompt-b/twochar/checkpoint-150/rng_state_0.pth +3 -0
  31. prompt-b/twochar/checkpoint-150/rng_state_1.pth +3 -0
  32. prompt-b/twochar/checkpoint-150/rng_state_2.pth +3 -0
  33. prompt-b/twochar/checkpoint-150/rng_state_3.pth +3 -0
  34. prompt-b/twochar/checkpoint-150/rng_state_4.pth +3 -0
  35. prompt-b/twochar/checkpoint-150/rng_state_5.pth +3 -0
  36. prompt-b/twochar/checkpoint-150/rng_state_6.pth +3 -0
  37. prompt-b/twochar/checkpoint-150/rng_state_7.pth +3 -0
  38. prompt-b/twochar/checkpoint-150/scheduler.pt +3 -0
  39. prompt-b/twochar/checkpoint-150/trainer_state.json +975 -0
  40. prompt-b/twochar/checkpoint-150/training_args.bin +3 -0
  41. prompt-b/twochar/checkpoint-200/README.md +21 -0
  42. prompt-b/twochar/checkpoint-200/adapter_config.json +26 -0
  43. prompt-b/twochar/checkpoint-200/adapter_model.bin +3 -0
  44. prompt-b/twochar/checkpoint-200/adapter_model/README.md +21 -0
  45. prompt-b/twochar/checkpoint-200/adapter_model/adapter_config.json +26 -0
  46. prompt-b/twochar/checkpoint-200/adapter_model/adapter_model.bin +3 -0
  47. prompt-b/twochar/checkpoint-200/optimizer.pt +3 -0
  48. prompt-b/twochar/checkpoint-200/rng_state_0.pth +3 -0
  49. prompt-b/twochar/checkpoint-200/rng_state_1.pth +3 -0
  50. prompt-b/twochar/checkpoint-200/rng_state_2.pth +3 -0
prompt-b/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
prompt-b/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
+ "o_proj",
18
+ "v_proj",
19
+ "down_proj",
20
+ "up_proj",
21
+ "q_proj",
22
+ "k_proj",
23
+ "gate_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
prompt-b/twochar/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d91a76c67c2c30d3fe135aecd598332808db8738437b03cde99f20ed724b6a5
3
+ size 125374989
prompt-b/twochar/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[PAD]": 32000
3
+ }
prompt-b/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
prompt-b/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
+ "o_proj",
18
+ "v_proj",
19
+ "down_proj",
20
+ "up_proj",
21
+ "q_proj",
22
+ "k_proj",
23
+ "gate_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
prompt-b/twochar/checkpoint-100/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:580b1096ca85b6bdd0f09aa9a4f6dd811138ea9fcf6b6decdbb1b6a49b4f63a1
3
+ size 125374989
prompt-b/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
prompt-b/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
+ "o_proj",
18
+ "v_proj",
19
+ "down_proj",
20
+ "up_proj",
21
+ "q_proj",
22
+ "k_proj",
23
+ "gate_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
prompt-b/twochar/checkpoint-100/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:580b1096ca85b6bdd0f09aa9a4f6dd811138ea9fcf6b6decdbb1b6a49b4f63a1
3
+ size 125374989
prompt-b/twochar/checkpoint-100/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:32bdb3245ed186baa39f6e598aa0811cfebf0ddfd240d79b447a4815fffcb77a
3
+ size 250681597
prompt-b/twochar/checkpoint-100/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c43bb7d6f4d23f4437efd65b1367f1598d1da341e8e5b04d1630ed20fbe186cf
3
+ size 21687
prompt-b/twochar/checkpoint-100/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b827087b2b8d41b36f24a7f1107e81325efba6a38dfd3fad03688f21ad8226be
3
+ size 21687
prompt-b/twochar/checkpoint-100/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:682ccf2c2c58d50194b5c57bd6fd7f18e57a8f578094e1faffbbade5154b2257
3
+ size 21687
prompt-b/twochar/checkpoint-100/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d462010cd13028cb8baa2e3404d499203043d664dad1231eea108884627422d
3
+ size 21687
prompt-b/twochar/checkpoint-100/rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93718c74afcd0b7d1e9d1f2fcd5511ab87ce04e88a13f23e01f82c8bf7e25ba5
3
+ size 21687
prompt-b/twochar/checkpoint-100/rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99778c05c5758e7b13e7638f494ec33532dca8ab5fb4f762413dda3ea184a60b
3
+ size 21687
prompt-b/twochar/checkpoint-100/rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ecd84cb532c329375227fbc7266b9200e88f39dfa919483634b4cc7b6f681ff3
3
+ size 21687
prompt-b/twochar/checkpoint-100/rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:30f1b98aca8391e5a61b036b986e0813302d0e20952a6b77fe4f7f0fe176a22a
3
+ size 21687
prompt-b/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
prompt-b/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": 2.0614,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.02,
19
+ "learning_rate": 1.3e-05,
20
+ "loss": 2.109,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.03,
25
+ "learning_rate": 1.9499999999999996e-05,
26
+ "loss": 2.1164,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.04,
31
+ "learning_rate": 2.6e-05,
32
+ "loss": 2.1487,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.05,
37
+ "learning_rate": 3.25e-05,
38
+ "loss": 1.9267,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.06,
43
+ "learning_rate": 3.899999999999999e-05,
44
+ "loss": 2.1892,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.07,
49
+ "learning_rate": 4.5499999999999995e-05,
50
+ "loss": 2.0891,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.08,
55
+ "learning_rate": 5.2e-05,
56
+ "loss": 2.1076,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.09,
61
+ "learning_rate": 5.85e-05,
62
+ "loss": 2.1815,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.1,
67
+ "learning_rate": 6.5e-05,
68
+ "loss": 2.1353,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.11,
73
+ "learning_rate": 6.499564948206075e-05,
74
+ "loss": 2.1232,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.12,
79
+ "learning_rate": 6.498259909298188e-05,
80
+ "loss": 2.1035,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.13,
85
+ "learning_rate": 6.49608523266681e-05,
86
+ "loss": 2.1099,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.14,
91
+ "learning_rate": 6.493041500525461e-05,
92
+ "loss": 2.1483,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.15,
97
+ "learning_rate": 6.489129527754836e-05,
98
+ "loss": 2.0477,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.16,
103
+ "learning_rate": 6.484350361684639e-05,
104
+ "loss": 2.0376,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.17,
109
+ "learning_rate": 6.478705281813194e-05,
110
+ "loss": 2.1192,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 0.18,
115
+ "learning_rate": 6.472195799464884e-05,
116
+ "loss": 2.1356,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 0.19,
121
+ "learning_rate": 6.464823657385538e-05,
122
+ "loss": 2.1215,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 0.2,
127
+ "learning_rate": 6.456590829275857e-05,
128
+ "loss": 2.2004,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 0.2,
133
+ "eval_loss": 2.0963854789733887,
134
+ "eval_runtime": 1.0635,
135
+ "eval_samples_per_second": 7.523,
136
+ "eval_steps_per_second": 0.94,
137
+ "step": 20
138
+ },
139
+ {
140
+ "epoch": 0.21,
141
+ "learning_rate": 6.447499519263001e-05,
142
+ "loss": 2.106,
143
+ "step": 21
144
+ },
145
+ {
146
+ "epoch": 0.22,
147
+ "learning_rate": 6.437552161310498e-05,
148
+ "loss": 2.0924,
149
+ "step": 22
150
+ },
151
+ {
152
+ "epoch": 0.23,
153
+ "learning_rate": 6.426751418566609e-05,
154
+ "loss": 2.1906,
155
+ "step": 23
156
+ },
157
+ {
158
+ "epoch": 0.24,
159
+ "learning_rate": 6.415100182651334e-05,
160
+ "loss": 2.1954,
161
+ "step": 24
162
+ },
163
+ {
164
+ "epoch": 0.25,
165
+ "learning_rate": 6.402601572882268e-05,
166
+ "loss": 2.0621,
167
+ "step": 25
168
+ },
169
+ {
170
+ "epoch": 0.26,
171
+ "learning_rate": 6.38925893543947e-05,
172
+ "loss": 2.0017,
173
+ "step": 26
174
+ },
175
+ {
176
+ "epoch": 0.27,
177
+ "learning_rate": 6.375075842469626e-05,
178
+ "loss": 2.1176,
179
+ "step": 27
180
+ },
181
+ {
182
+ "epoch": 0.28,
183
+ "learning_rate": 6.360056091129678e-05,
184
+ "loss": 2.0863,
185
+ "step": 28
186
+ },
187
+ {
188
+ "epoch": 0.29,
189
+ "learning_rate": 6.344203702570254e-05,
190
+ "loss": 2.0782,
191
+ "step": 29
192
+ },
193
+ {
194
+ "epoch": 0.3,
195
+ "learning_rate": 6.327522920859093e-05,
196
+ "loss": 2.1925,
197
+ "step": 30
198
+ },
199
+ {
200
+ "epoch": 0.31,
201
+ "learning_rate": 6.310018211844817e-05,
202
+ "loss": 2.1874,
203
+ "step": 31
204
+ },
205
+ {
206
+ "epoch": 0.32,
207
+ "learning_rate": 6.291694261961308e-05,
208
+ "loss": 2.1522,
209
+ "step": 32
210
+ },
211
+ {
212
+ "epoch": 0.33,
213
+ "learning_rate": 6.272555976973039e-05,
214
+ "loss": 1.8911,
215
+ "step": 33
216
+ },
217
+ {
218
+ "epoch": 0.34,
219
+ "learning_rate": 6.252608480661681e-05,
220
+ "loss": 2.1279,
221
+ "step": 34
222
+ },
223
+ {
224
+ "epoch": 0.35,
225
+ "learning_rate": 6.231857113454346e-05,
226
+ "loss": 1.9683,
227
+ "step": 35
228
+ },
229
+ {
230
+ "epoch": 0.36,
231
+ "learning_rate": 6.21030743099382e-05,
232
+ "loss": 2.0233,
233
+ "step": 36
234
+ },
235
+ {
236
+ "epoch": 0.37,
237
+ "learning_rate": 6.18796520265119e-05,
238
+ "loss": 2.0614,
239
+ "step": 37
240
+ },
241
+ {
242
+ "epoch": 0.38,
243
+ "learning_rate": 6.164836409981237e-05,
244
+ "loss": 1.9792,
245
+ "step": 38
246
+ },
247
+ {
248
+ "epoch": 0.39,
249
+ "learning_rate": 6.140927245121032e-05,
250
+ "loss": 2.0157,
251
+ "step": 39
252
+ },
253
+ {
254
+ "epoch": 0.4,
255
+ "learning_rate": 6.116244109132153e-05,
256
+ "loss": 2.092,
257
+ "step": 40
258
+ },
259
+ {
260
+ "epoch": 0.4,
261
+ "eval_loss": 2.067598581314087,
262
+ "eval_runtime": 1.0677,
263
+ "eval_samples_per_second": 7.493,
264
+ "eval_steps_per_second": 0.937,
265
+ "step": 40
266
+ },
267
+ {
268
+ "epoch": 0.41,
269
+ "learning_rate": 6.0907936102869656e-05,
270
+ "loss": 1.928,
271
+ "step": 41
272
+ },
273
+ {
274
+ "epoch": 0.42,
275
+ "learning_rate": 6.0645825622994254e-05,
276
+ "loss": 2.1825,
277
+ "step": 42
278
+ },
279
+ {
280
+ "epoch": 0.43,
281
+ "learning_rate": 6.037617982500884e-05,
282
+ "loss": 2.0793,
283
+ "step": 43
284
+ },
285
+ {
286
+ "epoch": 0.44,
287
+ "learning_rate": 6.009907089961381e-05,
288
+ "loss": 1.9642,
289
+ "step": 44
290
+ },
291
+ {
292
+ "epoch": 0.45,
293
+ "learning_rate": 5.981457303556923e-05,
294
+ "loss": 2.0201,
295
+ "step": 45
296
+ },
297
+ {
298
+ "epoch": 0.46,
299
+ "learning_rate": 5.9522762399832716e-05,
300
+ "loss": 2.0672,
301
+ "step": 46
302
+ },
303
+ {
304
+ "epoch": 0.47,
305
+ "learning_rate": 5.922371711716768e-05,
306
+ "loss": 2.122,
307
+ "step": 47
308
+ },
309
+ {
310
+ "epoch": 0.48,
311
+ "learning_rate": 5.891751724922749e-05,
312
+ "loss": 2.035,
313
+ "step": 48
314
+ },
315
+ {
316
+ "epoch": 0.49,
317
+ "learning_rate": 5.860424477312095e-05,
318
+ "loss": 2.1025,
319
+ "step": 49
320
+ },
321
+ {
322
+ "epoch": 0.5,
323
+ "learning_rate": 5.828398355946514e-05,
324
+ "loss": 2.0633,
325
+ "step": 50
326
+ },
327
+ {
328
+ "epoch": 0.5,
329
+ "learning_rate": 5.795681934993113e-05,
330
+ "loss": 2.1863,
331
+ "step": 51
332
+ },
333
+ {
334
+ "epoch": 0.51,
335
+ "learning_rate": 5.7622839734288945e-05,
336
+ "loss": 2.1907,
337
+ "step": 52
338
+ },
339
+ {
340
+ "epoch": 0.52,
341
+ "learning_rate": 5.728213412695761e-05,
342
+ "loss": 2.0571,
343
+ "step": 53
344
+ },
345
+ {
346
+ "epoch": 0.53,
347
+ "learning_rate": 5.693479374306676e-05,
348
+ "loss": 1.992,
349
+ "step": 54
350
+ },
351
+ {
352
+ "epoch": 0.54,
353
+ "learning_rate": 5.658091157403617e-05,
354
+ "loss": 2.1608,
355
+ "step": 55
356
+ },
357
+ {
358
+ "epoch": 0.55,
359
+ "learning_rate": 5.622058236267965e-05,
360
+ "loss": 1.9824,
361
+ "step": 56
362
+ },
363
+ {
364
+ "epoch": 0.56,
365
+ "learning_rate": 5.585390257784018e-05,
366
+ "loss": 2.1031,
367
+ "step": 57
368
+ },
369
+ {
370
+ "epoch": 0.57,
371
+ "learning_rate": 5.548097038856279e-05,
372
+ "loss": 2.0112,
373
+ "step": 58
374
+ },
375
+ {
376
+ "epoch": 0.58,
377
+ "learning_rate": 5.5101885637812374e-05,
378
+ "loss": 2.0949,
379
+ "step": 59
380
+ },
381
+ {
382
+ "epoch": 0.59,
383
+ "learning_rate": 5.4716749815743304e-05,
384
+ "loss": 2.0847,
385
+ "step": 60
386
+ },
387
+ {
388
+ "epoch": 0.59,
389
+ "eval_loss": 2.0599141120910645,
390
+ "eval_runtime": 1.0691,
391
+ "eval_samples_per_second": 7.483,
392
+ "eval_steps_per_second": 0.935,
393
+ "step": 60
394
+ },
395
+ {
396
+ "epoch": 0.6,
397
+ "learning_rate": 5.432566603252809e-05,
398
+ "loss": 1.98,
399
+ "step": 61
400
+ },
401
+ {
402
+ "epoch": 0.61,
403
+ "learning_rate": 5.3928738990752234e-05,
404
+ "loss": 2.0836,
405
+ "step": 62
406
+ },
407
+ {
408
+ "epoch": 0.62,
409
+ "learning_rate": 5.3526074957382866e-05,
410
+ "loss": 2.2208,
411
+ "step": 63
412
+ },
413
+ {
414
+ "epoch": 0.63,
415
+ "learning_rate": 5.311778173531847e-05,
416
+ "loss": 2.0723,
417
+ "step": 64
418
+ },
419
+ {
420
+ "epoch": 0.64,
421
+ "learning_rate": 5.2703968634527514e-05,
422
+ "loss": 2.0783,
423
+ "step": 65
424
+ },
425
+ {
426
+ "epoch": 0.65,
427
+ "learning_rate": 5.2284746442783414e-05,
428
+ "loss": 2.0462,
429
+ "step": 66
430
+ },
431
+ {
432
+ "epoch": 0.66,
433
+ "learning_rate": 5.186022739600408e-05,
434
+ "loss": 2.1665,
435
+ "step": 67
436
+ },
437
+ {
438
+ "epoch": 0.67,
439
+ "learning_rate": 5.143052514820357e-05,
440
+ "loss": 2.0825,
441
+ "step": 68
442
+ },
443
+ {
444
+ "epoch": 0.68,
445
+ "learning_rate": 5.099575474106419e-05,
446
+ "loss": 1.9172,
447
+ "step": 69
448
+ },
449
+ {
450
+ "epoch": 0.69,
451
+ "learning_rate": 5.055603257313707e-05,
452
+ "loss": 2.0618,
453
+ "step": 70
454
+ },
455
+ {
456
+ "epoch": 0.7,
457
+ "learning_rate": 5.011147636867943e-05,
458
+ "loss": 1.9883,
459
+ "step": 71
460
+ },
461
+ {
462
+ "epoch": 0.71,
463
+ "learning_rate": 4.9662205146136955e-05,
464
+ "loss": 2.0498,
465
+ "step": 72
466
+ },
467
+ {
468
+ "epoch": 0.72,
469
+ "learning_rate": 4.92083391862797e-05,
470
+ "loss": 1.8797,
471
+ "step": 73
472
+ },
473
+ {
474
+ "epoch": 0.73,
475
+ "learning_rate": 4.874999999999999e-05,
476
+ "loss": 2.0852,
477
+ "step": 74
478
+ },
479
+ {
480
+ "epoch": 0.74,
481
+ "learning_rate": 4.828731029578105e-05,
482
+ "loss": 2.0565,
483
+ "step": 75
484
+ },
485
+ {
486
+ "epoch": 0.75,
487
+ "learning_rate": 4.7820393946844926e-05,
488
+ "loss": 2.1096,
489
+ "step": 76
490
+ },
491
+ {
492
+ "epoch": 0.76,
493
+ "learning_rate": 4.734937595798867e-05,
494
+ "loss": 2.0701,
495
+ "step": 77
496
+ },
497
+ {
498
+ "epoch": 0.77,
499
+ "learning_rate": 4.687438243211754e-05,
500
+ "loss": 2.1631,
501
+ "step": 78
502
+ },
503
+ {
504
+ "epoch": 0.78,
505
+ "learning_rate": 4.639554053648416e-05,
506
+ "loss": 2.1411,
507
+ "step": 79
508
+ },
509
+ {
510
+ "epoch": 0.79,
511
+ "learning_rate": 4.5912978468642824e-05,
512
+ "loss": 2.165,
513
+ "step": 80
514
+ },
515
+ {
516
+ "epoch": 0.79,
517
+ "eval_loss": 2.0530834197998047,
518
+ "eval_runtime": 1.068,
519
+ "eval_samples_per_second": 7.491,
520
+ "eval_steps_per_second": 0.936,
521
+ "step": 80
522
+ },
523
+ {
524
+ "epoch": 0.8,
525
+ "learning_rate": 4.542682542212785e-05,
526
+ "loss": 2.2604,
527
+ "step": 81
528
+ },
529
+ {
530
+ "epoch": 0.81,
531
+ "learning_rate": 4.4937211551865415e-05,
532
+ "loss": 2.0536,
533
+ "step": 82
534
+ },
535
+ {
536
+ "epoch": 0.82,
537
+ "learning_rate": 4.444426793932787e-05,
538
+ "loss": 2.2147,
539
+ "step": 83
540
+ },
541
+ {
542
+ "epoch": 0.83,
543
+ "learning_rate": 4.3948126557440085e-05,
544
+ "loss": 2.0179,
545
+ "step": 84
546
+ },
547
+ {
548
+ "epoch": 0.84,
549
+ "learning_rate": 4.344892023524714e-05,
550
+ "loss": 2.0337,
551
+ "step": 85
552
+ },
553
+ {
554
+ "epoch": 0.85,
555
+ "learning_rate": 4.2946782622352746e-05,
556
+ "loss": 1.9833,
557
+ "step": 86
558
+ },
559
+ {
560
+ "epoch": 0.86,
561
+ "learning_rate": 4.244184815313799e-05,
562
+ "loss": 1.9425,
563
+ "step": 87
564
+ },
565
+ {
566
+ "epoch": 0.87,
567
+ "learning_rate": 4.193425201077002e-05,
568
+ "loss": 2.075,
569
+ "step": 88
570
+ },
571
+ {
572
+ "epoch": 0.88,
573
+ "learning_rate": 4.14241300910103e-05,
574
+ "loss": 2.0486,
575
+ "step": 89
576
+ },
577
+ {
578
+ "epoch": 0.89,
579
+ "learning_rate": 4.091161896583192e-05,
580
+ "loss": 2.1619,
581
+ "step": 90
582
+ },
583
+ {
584
+ "epoch": 0.9,
585
+ "learning_rate": 4.0396855846856076e-05,
586
+ "loss": 2.0541,
587
+ "step": 91
588
+ },
589
+ {
590
+ "epoch": 0.91,
591
+ "learning_rate": 3.9879978548617125e-05,
592
+ "loss": 2.1311,
593
+ "step": 92
594
+ },
595
+ {
596
+ "epoch": 0.92,
597
+ "learning_rate": 3.936112545166636e-05,
598
+ "loss": 1.8837,
599
+ "step": 93
600
+ },
601
+ {
602
+ "epoch": 0.93,
603
+ "learning_rate": 3.884043546552417e-05,
604
+ "loss": 2.0533,
605
+ "step": 94
606
+ },
607
+ {
608
+ "epoch": 0.94,
609
+ "learning_rate": 3.831804799149057e-05,
610
+ "loss": 2.0174,
611
+ "step": 95
612
+ },
613
+ {
614
+ "epoch": 0.95,
615
+ "learning_rate": 3.779410288532413e-05,
616
+ "loss": 1.9667,
617
+ "step": 96
618
+ },
619
+ {
620
+ "epoch": 0.96,
621
+ "learning_rate": 3.726874041979925e-05,
622
+ "loss": 2.08,
623
+ "step": 97
624
+ },
625
+ {
626
+ "epoch": 0.97,
627
+ "learning_rate": 3.674210124715168e-05,
628
+ "loss": 1.9554,
629
+ "step": 98
630
+ },
631
+ {
632
+ "epoch": 0.98,
633
+ "learning_rate": 3.621432636142251e-05,
634
+ "loss": 1.9813,
635
+ "step": 99
636
+ },
637
+ {
638
+ "epoch": 0.99,
639
+ "learning_rate": 3.5685557060710726e-05,
640
+ "loss": 2.0757,
641
+ "step": 100
642
+ },
643
+ {
644
+ "epoch": 0.99,
645
+ "eval_loss": 2.046072244644165,
646
+ "eval_runtime": 1.0684,
647
+ "eval_samples_per_second": 7.488,
648
+ "eval_steps_per_second": 0.936,
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": 8.380441623003136e+16,
657
+ "trial_name": null,
658
+ "trial_params": null
659
+ }
prompt-b/twochar/checkpoint-100/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2d2f8ff2b343a2aacf7504038a10d85428289fb723bea73597030a3ed0e83d3f
3
+ size 4411
prompt-b/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
prompt-b/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
+ "o_proj",
18
+ "v_proj",
19
+ "down_proj",
20
+ "up_proj",
21
+ "q_proj",
22
+ "k_proj",
23
+ "gate_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
prompt-b/twochar/checkpoint-150/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21cfb4c53d8bbfa9e3d21adc5ddae32d060c1ffc0bb385777e3146e094b589e4
3
+ size 125374989
prompt-b/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
prompt-b/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
+ "o_proj",
18
+ "v_proj",
19
+ "down_proj",
20
+ "up_proj",
21
+ "q_proj",
22
+ "k_proj",
23
+ "gate_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
prompt-b/twochar/checkpoint-150/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21cfb4c53d8bbfa9e3d21adc5ddae32d060c1ffc0bb385777e3146e094b589e4
3
+ size 125374989
prompt-b/twochar/checkpoint-150/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1e95795d0a53bfae1ce84df4f03902b9e43b644616398d19f08b1831168c8f65
3
+ size 250681597
prompt-b/twochar/checkpoint-150/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:07e2bd8fe03ea9be448f1369f4b46b8c236b991a81b5f1d8b7f48d1789502f38
3
+ size 21687
prompt-b/twochar/checkpoint-150/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f2d76b558a68f978caef73ad18126b7eb9d1a6225002f77ca351ed17b3c62f8e
3
+ size 21687
prompt-b/twochar/checkpoint-150/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:454326193d4450f9c7b04cd0441f952bb85684eeed68721f01ab61cb5390c74e
3
+ size 21687
prompt-b/twochar/checkpoint-150/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb63f4a1885b169a112ad1321ff81dd29cc1f219ae423ff1628d51555b46b06c
3
+ size 21687
prompt-b/twochar/checkpoint-150/rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd2b3191b61423781dce3fa1f578683506ea2fb9e4ce4de25890264d178e1373
3
+ size 21687
prompt-b/twochar/checkpoint-150/rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b685e895dd06971990f9400c64124dc422991636dbff94076b9e32a519f5b22d
3
+ size 21687
prompt-b/twochar/checkpoint-150/rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4c7134ba41bb59d1c554c9a7f20a5bd2fac6b50560431b8312e4078d2db7d363
3
+ size 21687
prompt-b/twochar/checkpoint-150/rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bec58c75ab37f0499c7bd1a7246e09952496e0e302ee11fd74c5a0267b8d0553
3
+ size 21687
prompt-b/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
prompt-b/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": 2.0614,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.02,
19
+ "learning_rate": 1.3e-05,
20
+ "loss": 2.109,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.03,
25
+ "learning_rate": 1.9499999999999996e-05,
26
+ "loss": 2.1164,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.04,
31
+ "learning_rate": 2.6e-05,
32
+ "loss": 2.1487,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.05,
37
+ "learning_rate": 3.25e-05,
38
+ "loss": 1.9267,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.06,
43
+ "learning_rate": 3.899999999999999e-05,
44
+ "loss": 2.1892,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.07,
49
+ "learning_rate": 4.5499999999999995e-05,
50
+ "loss": 2.0891,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.08,
55
+ "learning_rate": 5.2e-05,
56
+ "loss": 2.1076,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.09,
61
+ "learning_rate": 5.85e-05,
62
+ "loss": 2.1815,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.1,
67
+ "learning_rate": 6.5e-05,
68
+ "loss": 2.1353,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.11,
73
+ "learning_rate": 6.499564948206075e-05,
74
+ "loss": 2.1232,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.12,
79
+ "learning_rate": 6.498259909298188e-05,
80
+ "loss": 2.1035,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.13,
85
+ "learning_rate": 6.49608523266681e-05,
86
+ "loss": 2.1099,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.14,
91
+ "learning_rate": 6.493041500525461e-05,
92
+ "loss": 2.1483,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.15,
97
+ "learning_rate": 6.489129527754836e-05,
98
+ "loss": 2.0477,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.16,
103
+ "learning_rate": 6.484350361684639e-05,
104
+ "loss": 2.0376,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.17,
109
+ "learning_rate": 6.478705281813194e-05,
110
+ "loss": 2.1192,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 0.18,
115
+ "learning_rate": 6.472195799464884e-05,
116
+ "loss": 2.1356,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 0.19,
121
+ "learning_rate": 6.464823657385538e-05,
122
+ "loss": 2.1215,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 0.2,
127
+ "learning_rate": 6.456590829275857e-05,
128
+ "loss": 2.2004,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 0.2,
133
+ "eval_loss": 2.0963854789733887,
134
+ "eval_runtime": 1.0635,
135
+ "eval_samples_per_second": 7.523,
136
+ "eval_steps_per_second": 0.94,
137
+ "step": 20
138
+ },
139
+ {
140
+ "epoch": 0.21,
141
+ "learning_rate": 6.447499519263001e-05,
142
+ "loss": 2.106,
143
+ "step": 21
144
+ },
145
+ {
146
+ "epoch": 0.22,
147
+ "learning_rate": 6.437552161310498e-05,
148
+ "loss": 2.0924,
149
+ "step": 22
150
+ },
151
+ {
152
+ "epoch": 0.23,
153
+ "learning_rate": 6.426751418566609e-05,
154
+ "loss": 2.1906,
155
+ "step": 23
156
+ },
157
+ {
158
+ "epoch": 0.24,
159
+ "learning_rate": 6.415100182651334e-05,
160
+ "loss": 2.1954,
161
+ "step": 24
162
+ },
163
+ {
164
+ "epoch": 0.25,
165
+ "learning_rate": 6.402601572882268e-05,
166
+ "loss": 2.0621,
167
+ "step": 25
168
+ },
169
+ {
170
+ "epoch": 0.26,
171
+ "learning_rate": 6.38925893543947e-05,
172
+ "loss": 2.0017,
173
+ "step": 26
174
+ },
175
+ {
176
+ "epoch": 0.27,
177
+ "learning_rate": 6.375075842469626e-05,
178
+ "loss": 2.1176,
179
+ "step": 27
180
+ },
181
+ {
182
+ "epoch": 0.28,
183
+ "learning_rate": 6.360056091129678e-05,
184
+ "loss": 2.0863,
185
+ "step": 28
186
+ },
187
+ {
188
+ "epoch": 0.29,
189
+ "learning_rate": 6.344203702570254e-05,
190
+ "loss": 2.0782,
191
+ "step": 29
192
+ },
193
+ {
194
+ "epoch": 0.3,
195
+ "learning_rate": 6.327522920859093e-05,
196
+ "loss": 2.1925,
197
+ "step": 30
198
+ },
199
+ {
200
+ "epoch": 0.31,
201
+ "learning_rate": 6.310018211844817e-05,
202
+ "loss": 2.1874,
203
+ "step": 31
204
+ },
205
+ {
206
+ "epoch": 0.32,
207
+ "learning_rate": 6.291694261961308e-05,
208
+ "loss": 2.1522,
209
+ "step": 32
210
+ },
211
+ {
212
+ "epoch": 0.33,
213
+ "learning_rate": 6.272555976973039e-05,
214
+ "loss": 1.8911,
215
+ "step": 33
216
+ },
217
+ {
218
+ "epoch": 0.34,
219
+ "learning_rate": 6.252608480661681e-05,
220
+ "loss": 2.1279,
221
+ "step": 34
222
+ },
223
+ {
224
+ "epoch": 0.35,
225
+ "learning_rate": 6.231857113454346e-05,
226
+ "loss": 1.9683,
227
+ "step": 35
228
+ },
229
+ {
230
+ "epoch": 0.36,
231
+ "learning_rate": 6.21030743099382e-05,
232
+ "loss": 2.0233,
233
+ "step": 36
234
+ },
235
+ {
236
+ "epoch": 0.37,
237
+ "learning_rate": 6.18796520265119e-05,
238
+ "loss": 2.0614,
239
+ "step": 37
240
+ },
241
+ {
242
+ "epoch": 0.38,
243
+ "learning_rate": 6.164836409981237e-05,
244
+ "loss": 1.9792,
245
+ "step": 38
246
+ },
247
+ {
248
+ "epoch": 0.39,
249
+ "learning_rate": 6.140927245121032e-05,
250
+ "loss": 2.0157,
251
+ "step": 39
252
+ },
253
+ {
254
+ "epoch": 0.4,
255
+ "learning_rate": 6.116244109132153e-05,
256
+ "loss": 2.092,
257
+ "step": 40
258
+ },
259
+ {
260
+ "epoch": 0.4,
261
+ "eval_loss": 2.067598581314087,
262
+ "eval_runtime": 1.0677,
263
+ "eval_samples_per_second": 7.493,
264
+ "eval_steps_per_second": 0.937,
265
+ "step": 40
266
+ },
267
+ {
268
+ "epoch": 0.41,
269
+ "learning_rate": 6.0907936102869656e-05,
270
+ "loss": 1.928,
271
+ "step": 41
272
+ },
273
+ {
274
+ "epoch": 0.42,
275
+ "learning_rate": 6.0645825622994254e-05,
276
+ "loss": 2.1825,
277
+ "step": 42
278
+ },
279
+ {
280
+ "epoch": 0.43,
281
+ "learning_rate": 6.037617982500884e-05,
282
+ "loss": 2.0793,
283
+ "step": 43
284
+ },
285
+ {
286
+ "epoch": 0.44,
287
+ "learning_rate": 6.009907089961381e-05,
288
+ "loss": 1.9642,
289
+ "step": 44
290
+ },
291
+ {
292
+ "epoch": 0.45,
293
+ "learning_rate": 5.981457303556923e-05,
294
+ "loss": 2.0201,
295
+ "step": 45
296
+ },
297
+ {
298
+ "epoch": 0.46,
299
+ "learning_rate": 5.9522762399832716e-05,
300
+ "loss": 2.0672,
301
+ "step": 46
302
+ },
303
+ {
304
+ "epoch": 0.47,
305
+ "learning_rate": 5.922371711716768e-05,
306
+ "loss": 2.122,
307
+ "step": 47
308
+ },
309
+ {
310
+ "epoch": 0.48,
311
+ "learning_rate": 5.891751724922749e-05,
312
+ "loss": 2.035,
313
+ "step": 48
314
+ },
315
+ {
316
+ "epoch": 0.49,
317
+ "learning_rate": 5.860424477312095e-05,
318
+ "loss": 2.1025,
319
+ "step": 49
320
+ },
321
+ {
322
+ "epoch": 0.5,
323
+ "learning_rate": 5.828398355946514e-05,
324
+ "loss": 2.0633,
325
+ "step": 50
326
+ },
327
+ {
328
+ "epoch": 0.5,
329
+ "learning_rate": 5.795681934993113e-05,
330
+ "loss": 2.1863,
331
+ "step": 51
332
+ },
333
+ {
334
+ "epoch": 0.51,
335
+ "learning_rate": 5.7622839734288945e-05,
336
+ "loss": 2.1907,
337
+ "step": 52
338
+ },
339
+ {
340
+ "epoch": 0.52,
341
+ "learning_rate": 5.728213412695761e-05,
342
+ "loss": 2.0571,
343
+ "step": 53
344
+ },
345
+ {
346
+ "epoch": 0.53,
347
+ "learning_rate": 5.693479374306676e-05,
348
+ "loss": 1.992,
349
+ "step": 54
350
+ },
351
+ {
352
+ "epoch": 0.54,
353
+ "learning_rate": 5.658091157403617e-05,
354
+ "loss": 2.1608,
355
+ "step": 55
356
+ },
357
+ {
358
+ "epoch": 0.55,
359
+ "learning_rate": 5.622058236267965e-05,
360
+ "loss": 1.9824,
361
+ "step": 56
362
+ },
363
+ {
364
+ "epoch": 0.56,
365
+ "learning_rate": 5.585390257784018e-05,
366
+ "loss": 2.1031,
367
+ "step": 57
368
+ },
369
+ {
370
+ "epoch": 0.57,
371
+ "learning_rate": 5.548097038856279e-05,
372
+ "loss": 2.0112,
373
+ "step": 58
374
+ },
375
+ {
376
+ "epoch": 0.58,
377
+ "learning_rate": 5.5101885637812374e-05,
378
+ "loss": 2.0949,
379
+ "step": 59
380
+ },
381
+ {
382
+ "epoch": 0.59,
383
+ "learning_rate": 5.4716749815743304e-05,
384
+ "loss": 2.0847,
385
+ "step": 60
386
+ },
387
+ {
388
+ "epoch": 0.59,
389
+ "eval_loss": 2.0599141120910645,
390
+ "eval_runtime": 1.0691,
391
+ "eval_samples_per_second": 7.483,
392
+ "eval_steps_per_second": 0.935,
393
+ "step": 60
394
+ },
395
+ {
396
+ "epoch": 0.6,
397
+ "learning_rate": 5.432566603252809e-05,
398
+ "loss": 1.98,
399
+ "step": 61
400
+ },
401
+ {
402
+ "epoch": 0.61,
403
+ "learning_rate": 5.3928738990752234e-05,
404
+ "loss": 2.0836,
405
+ "step": 62
406
+ },
407
+ {
408
+ "epoch": 0.62,
409
+ "learning_rate": 5.3526074957382866e-05,
410
+ "loss": 2.2208,
411
+ "step": 63
412
+ },
413
+ {
414
+ "epoch": 0.63,
415
+ "learning_rate": 5.311778173531847e-05,
416
+ "loss": 2.0723,
417
+ "step": 64
418
+ },
419
+ {
420
+ "epoch": 0.64,
421
+ "learning_rate": 5.2703968634527514e-05,
422
+ "loss": 2.0783,
423
+ "step": 65
424
+ },
425
+ {
426
+ "epoch": 0.65,
427
+ "learning_rate": 5.2284746442783414e-05,
428
+ "loss": 2.0462,
429
+ "step": 66
430
+ },
431
+ {
432
+ "epoch": 0.66,
433
+ "learning_rate": 5.186022739600408e-05,
434
+ "loss": 2.1665,
435
+ "step": 67
436
+ },
437
+ {
438
+ "epoch": 0.67,
439
+ "learning_rate": 5.143052514820357e-05,
440
+ "loss": 2.0825,
441
+ "step": 68
442
+ },
443
+ {
444
+ "epoch": 0.68,
445
+ "learning_rate": 5.099575474106419e-05,
446
+ "loss": 1.9172,
447
+ "step": 69
448
+ },
449
+ {
450
+ "epoch": 0.69,
451
+ "learning_rate": 5.055603257313707e-05,
452
+ "loss": 2.0618,
453
+ "step": 70
454
+ },
455
+ {
456
+ "epoch": 0.7,
457
+ "learning_rate": 5.011147636867943e-05,
458
+ "loss": 1.9883,
459
+ "step": 71
460
+ },
461
+ {
462
+ "epoch": 0.71,
463
+ "learning_rate": 4.9662205146136955e-05,
464
+ "loss": 2.0498,
465
+ "step": 72
466
+ },
467
+ {
468
+ "epoch": 0.72,
469
+ "learning_rate": 4.92083391862797e-05,
470
+ "loss": 1.8797,
471
+ "step": 73
472
+ },
473
+ {
474
+ "epoch": 0.73,
475
+ "learning_rate": 4.874999999999999e-05,
476
+ "loss": 2.0852,
477
+ "step": 74
478
+ },
479
+ {
480
+ "epoch": 0.74,
481
+ "learning_rate": 4.828731029578105e-05,
482
+ "loss": 2.0565,
483
+ "step": 75
484
+ },
485
+ {
486
+ "epoch": 0.75,
487
+ "learning_rate": 4.7820393946844926e-05,
488
+ "loss": 2.1096,
489
+ "step": 76
490
+ },
491
+ {
492
+ "epoch": 0.76,
493
+ "learning_rate": 4.734937595798867e-05,
494
+ "loss": 2.0701,
495
+ "step": 77
496
+ },
497
+ {
498
+ "epoch": 0.77,
499
+ "learning_rate": 4.687438243211754e-05,
500
+ "loss": 2.1631,
501
+ "step": 78
502
+ },
503
+ {
504
+ "epoch": 0.78,
505
+ "learning_rate": 4.639554053648416e-05,
506
+ "loss": 2.1411,
507
+ "step": 79
508
+ },
509
+ {
510
+ "epoch": 0.79,
511
+ "learning_rate": 4.5912978468642824e-05,
512
+ "loss": 2.165,
513
+ "step": 80
514
+ },
515
+ {
516
+ "epoch": 0.79,
517
+ "eval_loss": 2.0530834197998047,
518
+ "eval_runtime": 1.068,
519
+ "eval_samples_per_second": 7.491,
520
+ "eval_steps_per_second": 0.936,
521
+ "step": 80
522
+ },
523
+ {
524
+ "epoch": 0.8,
525
+ "learning_rate": 4.542682542212785e-05,
526
+ "loss": 2.2604,
527
+ "step": 81
528
+ },
529
+ {
530
+ "epoch": 0.81,
531
+ "learning_rate": 4.4937211551865415e-05,
532
+ "loss": 2.0536,
533
+ "step": 82
534
+ },
535
+ {
536
+ "epoch": 0.82,
537
+ "learning_rate": 4.444426793932787e-05,
538
+ "loss": 2.2147,
539
+ "step": 83
540
+ },
541
+ {
542
+ "epoch": 0.83,
543
+ "learning_rate": 4.3948126557440085e-05,
544
+ "loss": 2.0179,
545
+ "step": 84
546
+ },
547
+ {
548
+ "epoch": 0.84,
549
+ "learning_rate": 4.344892023524714e-05,
550
+ "loss": 2.0337,
551
+ "step": 85
552
+ },
553
+ {
554
+ "epoch": 0.85,
555
+ "learning_rate": 4.2946782622352746e-05,
556
+ "loss": 1.9833,
557
+ "step": 86
558
+ },
559
+ {
560
+ "epoch": 0.86,
561
+ "learning_rate": 4.244184815313799e-05,
562
+ "loss": 1.9425,
563
+ "step": 87
564
+ },
565
+ {
566
+ "epoch": 0.87,
567
+ "learning_rate": 4.193425201077002e-05,
568
+ "loss": 2.075,
569
+ "step": 88
570
+ },
571
+ {
572
+ "epoch": 0.88,
573
+ "learning_rate": 4.14241300910103e-05,
574
+ "loss": 2.0486,
575
+ "step": 89
576
+ },
577
+ {
578
+ "epoch": 0.89,
579
+ "learning_rate": 4.091161896583192e-05,
580
+ "loss": 2.1619,
581
+ "step": 90
582
+ },
583
+ {
584
+ "epoch": 0.9,
585
+ "learning_rate": 4.0396855846856076e-05,
586
+ "loss": 2.0541,
587
+ "step": 91
588
+ },
589
+ {
590
+ "epoch": 0.91,
591
+ "learning_rate": 3.9879978548617125e-05,
592
+ "loss": 2.1311,
593
+ "step": 92
594
+ },
595
+ {
596
+ "epoch": 0.92,
597
+ "learning_rate": 3.936112545166636e-05,
598
+ "loss": 1.8837,
599
+ "step": 93
600
+ },
601
+ {
602
+ "epoch": 0.93,
603
+ "learning_rate": 3.884043546552417e-05,
604
+ "loss": 2.0533,
605
+ "step": 94
606
+ },
607
+ {
608
+ "epoch": 0.94,
609
+ "learning_rate": 3.831804799149057e-05,
610
+ "loss": 2.0174,
611
+ "step": 95
612
+ },
613
+ {
614
+ "epoch": 0.95,
615
+ "learning_rate": 3.779410288532413e-05,
616
+ "loss": 1.9667,
617
+ "step": 96
618
+ },
619
+ {
620
+ "epoch": 0.96,
621
+ "learning_rate": 3.726874041979925e-05,
622
+ "loss": 2.08,
623
+ "step": 97
624
+ },
625
+ {
626
+ "epoch": 0.97,
627
+ "learning_rate": 3.674210124715168e-05,
628
+ "loss": 1.9554,
629
+ "step": 98
630
+ },
631
+ {
632
+ "epoch": 0.98,
633
+ "learning_rate": 3.621432636142251e-05,
634
+ "loss": 1.9813,
635
+ "step": 99
636
+ },
637
+ {
638
+ "epoch": 0.99,
639
+ "learning_rate": 3.5685557060710726e-05,
640
+ "loss": 2.0757,
641
+ "step": 100
642
+ },
643
+ {
644
+ "epoch": 0.99,
645
+ "eval_loss": 2.046072244644165,
646
+ "eval_runtime": 1.0684,
647
+ "eval_samples_per_second": 7.488,
648
+ "eval_steps_per_second": 0.936,
649
+ "step": 100
650
+ },
651
+ {
652
+ "epoch": 1.0,
653
+ "learning_rate": 3.5155934909344214e-05,
654
+ "loss": 1.9579,
655
+ "step": 101
656
+ },
657
+ {
658
+ "epoch": 1.01,
659
+ "learning_rate": 3.462560169997965e-05,
660
+ "loss": 1.9566,
661
+ "step": 102
662
+ },
663
+ {
664
+ "epoch": 1.02,
665
+ "learning_rate": 3.409469941564109e-05,
666
+ "loss": 1.8615,
667
+ "step": 103
668
+ },
669
+ {
670
+ "epoch": 1.03,
671
+ "learning_rate": 3.356337019170772e-05,
672
+ "loss": 2.0485,
673
+ "step": 104
674
+ },
675
+ {
676
+ "epoch": 1.04,
677
+ "learning_rate": 3.303175627786082e-05,
678
+ "loss": 2.0282,
679
+ "step": 105
680
+ },
681
+ {
682
+ "epoch": 1.05,
683
+ "learning_rate": 3.25e-05,
684
+ "loss": 1.9859,
685
+ "step": 106
686
+ },
687
+ {
688
+ "epoch": 1.06,
689
+ "learning_rate": 3.1968243722139173e-05,
690
+ "loss": 1.8691,
691
+ "step": 107
692
+ },
693
+ {
694
+ "epoch": 1.07,
695
+ "learning_rate": 3.143662980829228e-05,
696
+ "loss": 2.0285,
697
+ "step": 108
698
+ },
699
+ {
700
+ "epoch": 1.08,
701
+ "learning_rate": 3.090530058435891e-05,
702
+ "loss": 1.9122,
703
+ "step": 109
704
+ },
705
+ {
706
+ "epoch": 1.09,
707
+ "learning_rate": 3.0374398300020347e-05,
708
+ "loss": 2.2074,
709
+ "step": 110
710
+ },
711
+ {
712
+ "epoch": 1.1,
713
+ "learning_rate": 2.984406509065579e-05,
714
+ "loss": 2.0043,
715
+ "step": 111
716
+ },
717
+ {
718
+ "epoch": 1.11,
719
+ "learning_rate": 2.9314442939289275e-05,
720
+ "loss": 1.9841,
721
+ "step": 112
722
+ },
723
+ {
724
+ "epoch": 1.12,
725
+ "learning_rate": 2.8785673638577486e-05,
726
+ "loss": 2.0132,
727
+ "step": 113
728
+ },
729
+ {
730
+ "epoch": 1.13,
731
+ "learning_rate": 2.825789875284833e-05,
732
+ "loss": 1.9829,
733
+ "step": 114
734
+ },
735
+ {
736
+ "epoch": 1.14,
737
+ "learning_rate": 2.773125958020074e-05,
738
+ "loss": 1.966,
739
+ "step": 115
740
+ },
741
+ {
742
+ "epoch": 1.15,
743
+ "learning_rate": 2.7205897114675866e-05,
744
+ "loss": 1.9977,
745
+ "step": 116
746
+ },
747
+ {
748
+ "epoch": 1.16,
749
+ "learning_rate": 2.668195200850944e-05,
750
+ "loss": 2.0548,
751
+ "step": 117
752
+ },
753
+ {
754
+ "epoch": 1.17,
755
+ "learning_rate": 2.6159564534475832e-05,
756
+ "loss": 2.0984,
757
+ "step": 118
758
+ },
759
+ {
760
+ "epoch": 1.18,
761
+ "learning_rate": 2.563887454833363e-05,
762
+ "loss": 2.0417,
763
+ "step": 119
764
+ },
765
+ {
766
+ "epoch": 1.19,
767
+ "learning_rate": 2.5120021451382872e-05,
768
+ "loss": 2.1137,
769
+ "step": 120
770
+ },
771
+ {
772
+ "epoch": 1.19,
773
+ "eval_loss": 2.0429368019104004,
774
+ "eval_runtime": 1.0681,
775
+ "eval_samples_per_second": 7.49,
776
+ "eval_steps_per_second": 0.936,
777
+ "step": 120
778
+ },
779
+ {
780
+ "epoch": 1.2,
781
+ "learning_rate": 2.4603144153143925e-05,
782
+ "loss": 1.9983,
783
+ "step": 121
784
+ },
785
+ {
786
+ "epoch": 1.21,
787
+ "learning_rate": 2.4088381034168072e-05,
788
+ "loss": 2.1263,
789
+ "step": 122
790
+ },
791
+ {
792
+ "epoch": 1.22,
793
+ "learning_rate": 2.3575869908989705e-05,
794
+ "loss": 2.0453,
795
+ "step": 123
796
+ },
797
+ {
798
+ "epoch": 1.23,
799
+ "learning_rate": 2.306574798922998e-05,
800
+ "loss": 1.9909,
801
+ "step": 124
802
+ },
803
+ {
804
+ "epoch": 1.24,
805
+ "learning_rate": 2.2558151846862005e-05,
806
+ "loss": 1.9634,
807
+ "step": 125
808
+ },
809
+ {
810
+ "epoch": 1.25,
811
+ "learning_rate": 2.2053217377647255e-05,
812
+ "loss": 1.9686,
813
+ "step": 126
814
+ },
815
+ {
816
+ "epoch": 1.26,
817
+ "learning_rate": 2.1551079764752848e-05,
818
+ "loss": 2.0998,
819
+ "step": 127
820
+ },
821
+ {
822
+ "epoch": 1.27,
823
+ "learning_rate": 2.105187344255991e-05,
824
+ "loss": 2.0794,
825
+ "step": 128
826
+ },
827
+ {
828
+ "epoch": 1.28,
829
+ "learning_rate": 2.0555732060672138e-05,
830
+ "loss": 1.8749,
831
+ "step": 129
832
+ },
833
+ {
834
+ "epoch": 1.29,
835
+ "learning_rate": 2.0062788448134583e-05,
836
+ "loss": 2.0591,
837
+ "step": 130
838
+ },
839
+ {
840
+ "epoch": 1.3,
841
+ "learning_rate": 1.957317457787214e-05,
842
+ "loss": 2.0994,
843
+ "step": 131
844
+ },
845
+ {
846
+ "epoch": 1.31,
847
+ "learning_rate": 1.908702153135717e-05,
848
+ "loss": 1.8715,
849
+ "step": 132
850
+ },
851
+ {
852
+ "epoch": 1.32,
853
+ "learning_rate": 1.860445946351584e-05,
854
+ "loss": 1.983,
855
+ "step": 133
856
+ },
857
+ {
858
+ "epoch": 1.33,
859
+ "learning_rate": 1.8125617567882463e-05,
860
+ "loss": 2.1738,
861
+ "step": 134
862
+ },
863
+ {
864
+ "epoch": 1.34,
865
+ "learning_rate": 1.7650624042011325e-05,
866
+ "loss": 1.9573,
867
+ "step": 135
868
+ },
869
+ {
870
+ "epoch": 1.35,
871
+ "learning_rate": 1.7179606053155072e-05,
872
+ "loss": 2.0063,
873
+ "step": 136
874
+ },
875
+ {
876
+ "epoch": 1.36,
877
+ "learning_rate": 1.6712689704218944e-05,
878
+ "loss": 2.0043,
879
+ "step": 137
880
+ },
881
+ {
882
+ "epoch": 1.37,
883
+ "learning_rate": 1.6250000000000005e-05,
884
+ "loss": 1.9306,
885
+ "step": 138
886
+ },
887
+ {
888
+ "epoch": 1.38,
889
+ "learning_rate": 1.5791660813720294e-05,
890
+ "loss": 2.0638,
891
+ "step": 139
892
+ },
893
+ {
894
+ "epoch": 1.39,
895
+ "learning_rate": 1.533779485386304e-05,
896
+ "loss": 1.9447,
897
+ "step": 140
898
+ },
899
+ {
900
+ "epoch": 1.39,
901
+ "eval_loss": 2.0422682762145996,
902
+ "eval_runtime": 1.0691,
903
+ "eval_samples_per_second": 7.483,
904
+ "eval_steps_per_second": 0.935,
905
+ "step": 140
906
+ },
907
+ {
908
+ "epoch": 1.4,
909
+ "learning_rate": 1.4888523631320579e-05,
910
+ "loss": 2.0975,
911
+ "step": 141
912
+ },
913
+ {
914
+ "epoch": 1.41,
915
+ "learning_rate": 1.4443967426862935e-05,
916
+ "loss": 1.9837,
917
+ "step": 142
918
+ },
919
+ {
920
+ "epoch": 1.42,
921
+ "learning_rate": 1.4004245258935799e-05,
922
+ "loss": 2.1573,
923
+ "step": 143
924
+ },
925
+ {
926
+ "epoch": 1.43,
927
+ "learning_rate": 1.3569474851796432e-05,
928
+ "loss": 2.133,
929
+ "step": 144
930
+ },
931
+ {
932
+ "epoch": 1.44,
933
+ "learning_rate": 1.3139772603995914e-05,
934
+ "loss": 2.0504,
935
+ "step": 145
936
+ },
937
+ {
938
+ "epoch": 1.45,
939
+ "learning_rate": 1.2715253557216577e-05,
940
+ "loss": 1.9792,
941
+ "step": 146
942
+ },
943
+ {
944
+ "epoch": 1.46,
945
+ "learning_rate": 1.2296031365472491e-05,
946
+ "loss": 1.9987,
947
+ "step": 147
948
+ },
949
+ {
950
+ "epoch": 1.47,
951
+ "learning_rate": 1.1882218264681525e-05,
952
+ "loss": 2.0199,
953
+ "step": 148
954
+ },
955
+ {
956
+ "epoch": 1.48,
957
+ "learning_rate": 1.1473925042617137e-05,
958
+ "loss": 2.0314,
959
+ "step": 149
960
+ },
961
+ {
962
+ "epoch": 1.49,
963
+ "learning_rate": 1.1071261009247762e-05,
964
+ "loss": 2.1465,
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": 1.258108369282007e+17,
973
+ "trial_name": null,
974
+ "trial_params": null
975
+ }
prompt-b/twochar/checkpoint-150/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2d2f8ff2b343a2aacf7504038a10d85428289fb723bea73597030a3ed0e83d3f
3
+ size 4411
prompt-b/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
prompt-b/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
+ "o_proj",
18
+ "v_proj",
19
+ "down_proj",
20
+ "up_proj",
21
+ "q_proj",
22
+ "k_proj",
23
+ "gate_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
prompt-b/twochar/checkpoint-200/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:310ac823e3d830335a1292e6f605c3049e2af8672d18606601f93a64dfd411df
3
+ size 125374989
prompt-b/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
prompt-b/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
+ "o_proj",
18
+ "v_proj",
19
+ "down_proj",
20
+ "up_proj",
21
+ "q_proj",
22
+ "k_proj",
23
+ "gate_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
prompt-b/twochar/checkpoint-200/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:310ac823e3d830335a1292e6f605c3049e2af8672d18606601f93a64dfd411df
3
+ size 125374989
prompt-b/twochar/checkpoint-200/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a1917cbd7a79eec1284b0a0813fb39615252555b61b3288d56789b8bf4b2b76
3
+ size 250681597
prompt-b/twochar/checkpoint-200/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:12a9d90998cbf0d0c19a24d1951a6eb3668083a32192d7b6dd49bb0ddb49b819
3
+ size 21687
prompt-b/twochar/checkpoint-200/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c54741901ef9bd54a0e567804d6b54cc118bbee9eec08756a401e08b6748b19
3
+ size 21687
prompt-b/twochar/checkpoint-200/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:33dc3ea26f3ef71f649b5c031626fcfe8246eddfbda905b6de7fe669731ee707
3
+ size 21687