Hugging Face's logo

royallab
/
LimaRP-ShareGPT-13b-qloras

PEFT
Not-For-All-Audiences
Model card Files Community
2
kingbri commited on
Commit
b32633c
•
1 Parent(s): c2f1218

Add prompt-a onechar

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-a/onechar/README.md +21 -0
  2. prompt-a/onechar/adapter_config.json +26 -0
  3. prompt-a/onechar/adapter_model.bin +3 -0
  4. prompt-a/onechar/added_tokens.json +3 -0
  5. prompt-a/onechar/checkpoint-100/README.md +21 -0
  6. prompt-a/onechar/checkpoint-100/adapter_config.json +26 -0
  7. prompt-a/onechar/checkpoint-100/adapter_model.bin +3 -0
  8. prompt-a/onechar/checkpoint-100/adapter_model/README.md +21 -0
  9. prompt-a/onechar/checkpoint-100/adapter_model/adapter_config.json +26 -0
  10. prompt-a/onechar/checkpoint-100/adapter_model/adapter_model.bin +3 -0
  11. prompt-a/onechar/checkpoint-100/optimizer.pt +3 -0
  12. prompt-a/onechar/checkpoint-100/rng_state_0.pth +3 -0
  13. prompt-a/onechar/checkpoint-100/rng_state_1.pth +3 -0
  14. prompt-a/onechar/checkpoint-100/rng_state_2.pth +3 -0
  15. prompt-a/onechar/checkpoint-100/rng_state_3.pth +3 -0
  16. prompt-a/onechar/checkpoint-100/rng_state_4.pth +3 -0
  17. prompt-a/onechar/checkpoint-100/rng_state_5.pth +3 -0
  18. prompt-a/onechar/checkpoint-100/rng_state_6.pth +3 -0
  19. prompt-a/onechar/checkpoint-100/rng_state_7.pth +3 -0
  20. prompt-a/onechar/checkpoint-100/scheduler.pt +3 -0
  21. prompt-a/onechar/checkpoint-100/trainer_state.json +659 -0
  22. prompt-a/onechar/checkpoint-100/training_args.bin +3 -0
  23. prompt-a/onechar/checkpoint-150/README.md +21 -0
  24. prompt-a/onechar/checkpoint-150/adapter_config.json +26 -0
  25. prompt-a/onechar/checkpoint-150/adapter_model.bin +3 -0
  26. prompt-a/onechar/checkpoint-150/adapter_model/README.md +21 -0
  27. prompt-a/onechar/checkpoint-150/adapter_model/adapter_config.json +26 -0
  28. prompt-a/onechar/checkpoint-150/adapter_model/adapter_model.bin +3 -0
  29. prompt-a/onechar/checkpoint-150/optimizer.pt +3 -0
  30. prompt-a/onechar/checkpoint-150/rng_state_0.pth +3 -0
  31. prompt-a/onechar/checkpoint-150/rng_state_1.pth +3 -0
  32. prompt-a/onechar/checkpoint-150/rng_state_2.pth +3 -0
  33. prompt-a/onechar/checkpoint-150/rng_state_3.pth +3 -0
  34. prompt-a/onechar/checkpoint-150/rng_state_4.pth +3 -0
  35. prompt-a/onechar/checkpoint-150/rng_state_5.pth +3 -0
  36. prompt-a/onechar/checkpoint-150/rng_state_6.pth +3 -0
  37. prompt-a/onechar/checkpoint-150/rng_state_7.pth +3 -0
  38. prompt-a/onechar/checkpoint-150/scheduler.pt +3 -0
  39. prompt-a/onechar/checkpoint-150/trainer_state.json +975 -0
  40. prompt-a/onechar/checkpoint-150/training_args.bin +3 -0
  41. prompt-a/onechar/checkpoint-200/README.md +21 -0
  42. prompt-a/onechar/checkpoint-200/adapter_config.json +26 -0
  43. prompt-a/onechar/checkpoint-200/adapter_model.bin +3 -0
  44. prompt-a/onechar/checkpoint-200/adapter_model/README.md +21 -0
  45. prompt-a/onechar/checkpoint-200/adapter_model/adapter_config.json +26 -0
  46. prompt-a/onechar/checkpoint-200/adapter_model/adapter_model.bin +3 -0
  47. prompt-a/onechar/checkpoint-200/optimizer.pt +3 -0
  48. prompt-a/onechar/checkpoint-200/rng_state_0.pth +3 -0
  49. prompt-a/onechar/checkpoint-200/rng_state_1.pth +3 -0
  50. prompt-a/onechar/checkpoint-200/rng_state_2.pth +3 -0
prompt-a/onechar/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-a/onechar/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
+ "q_proj",
19
+ "k_proj",
20
+ "down_proj",
21
+ "up_proj",
22
+ "gate_proj",
23
+ "v_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
prompt-a/onechar/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1f5ec19329cacf861536c142e358f467bfaf4ef5352429c4826c33d208347a3d
3
+ size 125374989
prompt-a/onechar/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[PAD]": 32000
3
+ }
prompt-a/onechar/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-a/onechar/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
+ "q_proj",
19
+ "k_proj",
20
+ "down_proj",
21
+ "up_proj",
22
+ "gate_proj",
23
+ "v_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
prompt-a/onechar/checkpoint-100/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ff610c0b53c8af0e9c984b4ed0725380085f987aa69e16c979a257578cfabcb
3
+ size 125374989
prompt-a/onechar/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-a/onechar/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
+ "q_proj",
19
+ "k_proj",
20
+ "down_proj",
21
+ "up_proj",
22
+ "gate_proj",
23
+ "v_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
prompt-a/onechar/checkpoint-100/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ff610c0b53c8af0e9c984b4ed0725380085f987aa69e16c979a257578cfabcb
3
+ size 125374989
prompt-a/onechar/checkpoint-100/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f7f1f403d38dc995a8098ed409cbd760e111bd882f7a68df45055032258b88b6
3
+ size 250681597
prompt-a/onechar/checkpoint-100/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4102fd8971c12e358d8aa24117ab73918cc1aaac07c4678184c87371a7cce1c6
3
+ size 21687
prompt-a/onechar/checkpoint-100/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a35f8c8b1beb2b76ef2ca1aa345895c6db951b1551c9042228eb0ea741ce12c
3
+ size 21687
prompt-a/onechar/checkpoint-100/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa20b68c489c5a2cd2428b1f3e5072cdf8f2417980be271f9db72c4049028bb3
3
+ size 21687
prompt-a/onechar/checkpoint-100/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a828f6b225941a386fb560abfcd7d4dee497153fe1cf40b8a87e5174aa75bed3
3
+ size 21687
prompt-a/onechar/checkpoint-100/rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1af5550599dc6617b8105e29364818d1dbea0b5d63b4e86bd3c1697f930ae248
3
+ size 21687
prompt-a/onechar/checkpoint-100/rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b13498715ccaef6df8b69c2dab24d427d1152ac7f5c20f3a7aa0579d6df2f93
3
+ size 21687
prompt-a/onechar/checkpoint-100/rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e0a13f52d9df555a3b9fac609325509aebf8adff3267140830db81111f0c8f4a
3
+ size 21687
prompt-a/onechar/checkpoint-100/rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f7afe0df70969c4c1a997671975d84b8bdde138351b49fb8396fec85cf845301
3
+ size 21687
prompt-a/onechar/checkpoint-100/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:781197e4490645acd969affd7bc482901bb63fc73f416de5de064f3460ca305a
3
+ size 627
prompt-a/onechar/checkpoint-100/trainer_state.json ADDED
@@ -0,0 +1,659 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9523809523809523,
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": 2.1535,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.02,
19
+ "learning_rate": 1.3e-05,
20
+ "loss": 2.1328,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.03,
25
+ "learning_rate": 1.9499999999999996e-05,
26
+ "loss": 2.0918,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.04,
31
+ "learning_rate": 2.6e-05,
32
+ "loss": 2.1194,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.05,
37
+ "learning_rate": 3.25e-05,
38
+ "loss": 2.2342,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.06,
43
+ "learning_rate": 3.899999999999999e-05,
44
+ "loss": 2.1168,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.07,
49
+ "learning_rate": 4.5499999999999995e-05,
50
+ "loss": 2.1761,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.08,
55
+ "learning_rate": 5.2e-05,
56
+ "loss": 2.0764,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.09,
61
+ "learning_rate": 5.85e-05,
62
+ "loss": 2.1711,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.1,
67
+ "learning_rate": 6.5e-05,
68
+ "loss": 2.0885,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.1,
73
+ "learning_rate": 6.499599055565396e-05,
74
+ "loss": 2.0196,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.11,
79
+ "learning_rate": 6.498396321188627e-05,
80
+ "loss": 2.1799,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.12,
85
+ "learning_rate": 6.496392093626402e-05,
86
+ "loss": 2.1322,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.13,
91
+ "learning_rate": 6.493586867391882e-05,
92
+ "loss": 2.1486,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.14,
97
+ "learning_rate": 6.489981334632665e-05,
98
+ "loss": 2.1351,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.15,
103
+ "learning_rate": 6.48557638496001e-05,
104
+ "loss": 2.1366,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.16,
109
+ "learning_rate": 6.480373105229334e-05,
110
+ "loss": 2.0727,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 0.17,
115
+ "learning_rate": 6.474372779272053e-05,
116
+ "loss": 2.1123,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 0.18,
121
+ "learning_rate": 6.467576887578811e-05,
122
+ "loss": 2.0652,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 0.19,
127
+ "learning_rate": 6.459987106934198e-05,
128
+ "loss": 2.0008,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 0.19,
133
+ "eval_loss": 2.0444116592407227,
134
+ "eval_runtime": 2.798,
135
+ "eval_samples_per_second": 3.574,
136
+ "eval_steps_per_second": 0.715,
137
+ "step": 20
138
+ },
139
+ {
140
+ "epoch": 0.2,
141
+ "learning_rate": 6.451605310003014e-05,
142
+ "loss": 2.0769,
143
+ "step": 21
144
+ },
145
+ {
146
+ "epoch": 0.21,
147
+ "learning_rate": 6.442433564868238e-05,
148
+ "loss": 1.9949,
149
+ "step": 22
150
+ },
151
+ {
152
+ "epoch": 0.22,
153
+ "learning_rate": 6.432474134520738e-05,
154
+ "loss": 2.1341,
155
+ "step": 23
156
+ },
157
+ {
158
+ "epoch": 0.23,
159
+ "learning_rate": 6.421729476300929e-05,
160
+ "loss": 2.1433,
161
+ "step": 24
162
+ },
163
+ {
164
+ "epoch": 0.24,
165
+ "learning_rate": 6.410202241292448e-05,
166
+ "loss": 2.0168,
167
+ "step": 25
168
+ },
169
+ {
170
+ "epoch": 0.25,
171
+ "learning_rate": 6.39789527366805e-05,
172
+ "loss": 2.0403,
173
+ "step": 26
174
+ },
175
+ {
176
+ "epoch": 0.26,
177
+ "learning_rate": 6.384811609987843e-05,
178
+ "loss": 2.1983,
179
+ "step": 27
180
+ },
181
+ {
182
+ "epoch": 0.27,
183
+ "learning_rate": 6.370954478450064e-05,
184
+ "loss": 2.0625,
185
+ "step": 28
186
+ },
187
+ {
188
+ "epoch": 0.28,
189
+ "learning_rate": 6.356327298094572e-05,
190
+ "loss": 2.204,
191
+ "step": 29
192
+ },
193
+ {
194
+ "epoch": 0.29,
195
+ "learning_rate": 6.340933677959248e-05,
196
+ "loss": 2.118,
197
+ "step": 30
198
+ },
199
+ {
200
+ "epoch": 0.3,
201
+ "learning_rate": 6.324777416189522e-05,
202
+ "loss": 2.0997,
203
+ "step": 31
204
+ },
205
+ {
206
+ "epoch": 0.3,
207
+ "learning_rate": 6.307862499101232e-05,
208
+ "loss": 2.1154,
209
+ "step": 32
210
+ },
211
+ {
212
+ "epoch": 0.31,
213
+ "learning_rate": 6.290193100197069e-05,
214
+ "loss": 2.0845,
215
+ "step": 33
216
+ },
217
+ {
218
+ "epoch": 0.32,
219
+ "learning_rate": 6.271773579136816e-05,
220
+ "loss": 2.0052,
221
+ "step": 34
222
+ },
223
+ {
224
+ "epoch": 0.33,
225
+ "learning_rate": 6.252608480661681e-05,
226
+ "loss": 1.956,
227
+ "step": 35
228
+ },
229
+ {
230
+ "epoch": 0.34,
231
+ "learning_rate": 6.232702533472938e-05,
232
+ "loss": 2.0958,
233
+ "step": 36
234
+ },
235
+ {
236
+ "epoch": 0.35,
237
+ "learning_rate": 6.212060649065197e-05,
238
+ "loss": 2.0634,
239
+ "step": 37
240
+ },
241
+ {
242
+ "epoch": 0.36,
243
+ "learning_rate": 6.190687920514563e-05,
244
+ "loss": 2.0033,
245
+ "step": 38
246
+ },
247
+ {
248
+ "epoch": 0.37,
249
+ "learning_rate": 6.168589621222e-05,
250
+ "loss": 2.1426,
251
+ "step": 39
252
+ },
253
+ {
254
+ "epoch": 0.38,
255
+ "learning_rate": 6.145771203612195e-05,
256
+ "loss": 1.9823,
257
+ "step": 40
258
+ },
259
+ {
260
+ "epoch": 0.38,
261
+ "eval_loss": 2.0226035118103027,
262
+ "eval_runtime": 2.8049,
263
+ "eval_samples_per_second": 3.565,
264
+ "eval_steps_per_second": 0.713,
265
+ "step": 40
266
+ },
267
+ {
268
+ "epoch": 0.39,
269
+ "learning_rate": 6.122238297788254e-05,
270
+ "loss": 2.2016,
271
+ "step": 41
272
+ },
273
+ {
274
+ "epoch": 0.4,
275
+ "learning_rate": 6.097996710142556e-05,
276
+ "loss": 2.1008,
277
+ "step": 42
278
+ },
279
+ {
280
+ "epoch": 0.41,
281
+ "learning_rate": 6.0730524219241214e-05,
282
+ "loss": 2.0821,
283
+ "step": 43
284
+ },
285
+ {
286
+ "epoch": 0.42,
287
+ "learning_rate": 6.047411587762816e-05,
288
+ "loss": 2.092,
289
+ "step": 44
290
+ },
291
+ {
292
+ "epoch": 0.43,
293
+ "learning_rate": 6.0210805341507994e-05,
294
+ "loss": 1.9617,
295
+ "step": 45
296
+ },
297
+ {
298
+ "epoch": 0.44,
299
+ "learning_rate": 5.994065757881549e-05,
300
+ "loss": 1.9358,
301
+ "step": 46
302
+ },
303
+ {
304
+ "epoch": 0.45,
305
+ "learning_rate": 5.966373924446878e-05,
306
+ "loss": 2.0573,
307
+ "step": 47
308
+ },
309
+ {
310
+ "epoch": 0.46,
311
+ "learning_rate": 5.938011866392325e-05,
312
+ "loss": 2.0731,
313
+ "step": 48
314
+ },
315
+ {
316
+ "epoch": 0.47,
317
+ "learning_rate": 5.9089865816313253e-05,
318
+ "loss": 2.2842,
319
+ "step": 49
320
+ },
321
+ {
322
+ "epoch": 0.48,
323
+ "learning_rate": 5.8793052317185785e-05,
324
+ "loss": 2.2145,
325
+ "step": 50
326
+ },
327
+ {
328
+ "epoch": 0.49,
329
+ "learning_rate": 5.848975140083044e-05,
330
+ "loss": 1.9923,
331
+ "step": 51
332
+ },
333
+ {
334
+ "epoch": 0.5,
335
+ "learning_rate": 5.818003790220993e-05,
336
+ "loss": 2.1065,
337
+ "step": 52
338
+ },
339
+ {
340
+ "epoch": 0.5,
341
+ "learning_rate": 5.786398823849571e-05,
342
+ "loss": 2.0009,
343
+ "step": 53
344
+ },
345
+ {
346
+ "epoch": 0.51,
347
+ "learning_rate": 5.754168039021315e-05,
348
+ "loss": 1.9785,
349
+ "step": 54
350
+ },
351
+ {
352
+ "epoch": 0.52,
353
+ "learning_rate": 5.7213193882001006e-05,
354
+ "loss": 1.9504,
355
+ "step": 55
356
+ },
357
+ {
358
+ "epoch": 0.53,
359
+ "learning_rate": 5.687860976298993e-05,
360
+ "loss": 2.1977,
361
+ "step": 56
362
+ },
363
+ {
364
+ "epoch": 0.54,
365
+ "learning_rate": 5.6538010586804816e-05,
366
+ "loss": 2.0423,
367
+ "step": 57
368
+ },
369
+ {
370
+ "epoch": 0.55,
371
+ "learning_rate": 5.619148039119587e-05,
372
+ "loss": 2.0536,
373
+ "step": 58
374
+ },
375
+ {
376
+ "epoch": 0.56,
377
+ "learning_rate": 5.5839104677303635e-05,
378
+ "loss": 2.0286,
379
+ "step": 59
380
+ },
381
+ {
382
+ "epoch": 0.57,
383
+ "learning_rate": 5.548097038856279e-05,
384
+ "loss": 2.1143,
385
+ "step": 60
386
+ },
387
+ {
388
+ "epoch": 0.57,
389
+ "eval_loss": 2.0128190517425537,
390
+ "eval_runtime": 2.8105,
391
+ "eval_samples_per_second": 3.558,
392
+ "eval_steps_per_second": 0.712,
393
+ "step": 60
394
+ },
395
+ {
396
+ "epoch": 0.58,
397
+ "learning_rate": 5.511716588925021e-05,
398
+ "loss": 2.1397,
399
+ "step": 61
400
+ },
401
+ {
402
+ "epoch": 0.59,
403
+ "learning_rate": 5.474778094268238e-05,
404
+ "loss": 2.0601,
405
+ "step": 62
406
+ },
407
+ {
408
+ "epoch": 0.6,
409
+ "learning_rate": 5.437290668906763e-05,
410
+ "loss": 2.1707,
411
+ "step": 63
412
+ },
413
+ {
414
+ "epoch": 0.61,
415
+ "learning_rate": 5.399263562301868e-05,
416
+ "loss": 2.0649,
417
+ "step": 64
418
+ },
419
+ {
420
+ "epoch": 0.62,
421
+ "learning_rate": 5.360706157073096e-05,
422
+ "loss": 2.0314,
423
+ "step": 65
424
+ },
425
+ {
426
+ "epoch": 0.63,
427
+ "learning_rate": 5.321627966683241e-05,
428
+ "loss": 2.1277,
429
+ "step": 66
430
+ },
431
+ {
432
+ "epoch": 0.64,
433
+ "learning_rate": 5.2820386330910424e-05,
434
+ "loss": 2.0534,
435
+ "step": 67
436
+ },
437
+ {
438
+ "epoch": 0.65,
439
+ "learning_rate": 5.2419479243721734e-05,
440
+ "loss": 1.9976,
441
+ "step": 68
442
+ },
443
+ {
444
+ "epoch": 0.66,
445
+ "learning_rate": 5.2013657323091226e-05,
446
+ "loss": 2.0262,
447
+ "step": 69
448
+ },
449
+ {
450
+ "epoch": 0.67,
451
+ "learning_rate": 5.160302069950537e-05,
452
+ "loss": 2.0093,
453
+ "step": 70
454
+ },
455
+ {
456
+ "epoch": 0.68,
457
+ "learning_rate": 5.118767069140655e-05,
458
+ "loss": 2.1088,
459
+ "step": 71
460
+ },
461
+ {
462
+ "epoch": 0.69,
463
+ "learning_rate": 5.076770978019424e-05,
464
+ "loss": 1.9841,
465
+ "step": 72
466
+ },
467
+ {
468
+ "epoch": 0.7,
469
+ "learning_rate": 5.034324158493928e-05,
470
+ "loss": 2.0081,
471
+ "step": 73
472
+ },
473
+ {
474
+ "epoch": 0.7,
475
+ "learning_rate": 4.991437083681739e-05,
476
+ "loss": 2.0179,
477
+ "step": 74
478
+ },
479
+ {
480
+ "epoch": 0.71,
481
+ "learning_rate": 4.948120335326833e-05,
482
+ "loss": 2.1304,
483
+ "step": 75
484
+ },
485
+ {
486
+ "epoch": 0.72,
487
+ "learning_rate": 4.904384601188706e-05,
488
+ "loss": 1.9213,
489
+ "step": 76
490
+ },
491
+ {
492
+ "epoch": 0.73,
493
+ "learning_rate": 4.860240672405324e-05,
494
+ "loss": 2.2153,
495
+ "step": 77
496
+ },
497
+ {
498
+ "epoch": 0.74,
499
+ "learning_rate": 4.815699440830574e-05,
500
+ "loss": 1.9365,
501
+ "step": 78
502
+ },
503
+ {
504
+ "epoch": 0.75,
505
+ "learning_rate": 4.770771896346863e-05,
506
+ "loss": 2.021,
507
+ "step": 79
508
+ },
509
+ {
510
+ "epoch": 0.76,
511
+ "learning_rate": 4.725469124153526e-05,
512
+ "loss": 2.0327,
513
+ "step": 80
514
+ },
515
+ {
516
+ "epoch": 0.76,
517
+ "eval_loss": 2.0059409141540527,
518
+ "eval_runtime": 2.8077,
519
+ "eval_samples_per_second": 3.562,
520
+ "eval_steps_per_second": 0.712,
521
+ "step": 80
522
+ },
523
+ {
524
+ "epoch": 0.77,
525
+ "learning_rate": 4.679802302031724e-05,
526
+ "loss": 2.1576,
527
+ "step": 81
528
+ },
529
+ {
530
+ "epoch": 0.78,
531
+ "learning_rate": 4.6337826975864856e-05,
532
+ "loss": 1.9468,
533
+ "step": 82
534
+ },
535
+ {
536
+ "epoch": 0.79,
537
+ "learning_rate": 4.587421665466603e-05,
538
+ "loss": 2.0795,
539
+ "step": 83
540
+ },
541
+ {
542
+ "epoch": 0.8,
543
+ "learning_rate": 4.540730644563036e-05,
544
+ "loss": 2.0566,
545
+ "step": 84
546
+ },
547
+ {
548
+ "epoch": 0.81,
549
+ "learning_rate": 4.4937211551865415e-05,
550
+ "loss": 2.0819,
551
+ "step": 85
552
+ },
553
+ {
554
+ "epoch": 0.82,
555
+ "learning_rate": 4.4464047962252035e-05,
556
+ "loss": 2.0542,
557
+ "step": 86
558
+ },
559
+ {
560
+ "epoch": 0.83,
561
+ "learning_rate": 4.398793242282586e-05,
562
+ "loss": 2.0006,
563
+ "step": 87
564
+ },
565
+ {
566
+ "epoch": 0.84,
567
+ "learning_rate": 4.350898240797197e-05,
568
+ "loss": 1.9953,
569
+ "step": 88
570
+ },
571
+ {
572
+ "epoch": 0.85,
573
+ "learning_rate": 4.3027316091439854e-05,
574
+ "loss": 2.067,
575
+ "step": 89
576
+ },
577
+ {
578
+ "epoch": 0.86,
579
+ "learning_rate": 4.254305231718579e-05,
580
+ "loss": 1.9909,
581
+ "step": 90
582
+ },
583
+ {
584
+ "epoch": 0.87,
585
+ "learning_rate": 4.205631057004987e-05,
586
+ "loss": 1.9436,
587
+ "step": 91
588
+ },
589
+ {
590
+ "epoch": 0.88,
591
+ "learning_rate": 4.156721094627495e-05,
592
+ "loss": 2.0643,
593
+ "step": 92
594
+ },
595
+ {
596
+ "epoch": 0.89,
597
+ "learning_rate": 4.107587412387462e-05,
598
+ "loss": 2.1244,
599
+ "step": 93
600
+ },
601
+ {
602
+ "epoch": 0.9,
603
+ "learning_rate": 4.058242133285779e-05,
604
+ "loss": 1.9795,
605
+ "step": 94
606
+ },
607
+ {
608
+ "epoch": 0.9,
609
+ "learning_rate": 4.008697432531693e-05,
610
+ "loss": 2.1568,
611
+ "step": 95
612
+ },
613
+ {
614
+ "epoch": 0.91,
615
+ "learning_rate": 3.9589655345387633e-05,
616
+ "loss": 1.9829,
617
+ "step": 96
618
+ },
619
+ {
620
+ "epoch": 0.92,
621
+ "learning_rate": 3.909058709908665e-05,
622
+ "loss": 2.031,
623
+ "step": 97
624
+ },
625
+ {
626
+ "epoch": 0.93,
627
+ "learning_rate": 3.858989272403605e-05,
628
+ "loss": 2.2042,
629
+ "step": 98
630
+ },
631
+ {
632
+ "epoch": 0.94,
633
+ "learning_rate": 3.8087695759080804e-05,
634
+ "loss": 1.9775,
635
+ "step": 99
636
+ },
637
+ {
638
+ "epoch": 0.95,
639
+ "learning_rate": 3.75841201138075e-05,
640
+ "loss": 2.0274,
641
+ "step": 100
642
+ },
643
+ {
644
+ "epoch": 0.95,
645
+ "eval_loss": 2.0003483295440674,
646
+ "eval_runtime": 2.8075,
647
+ "eval_samples_per_second": 3.562,
648
+ "eval_steps_per_second": 0.712,
649
+ "step": 100
650
+ }
651
+ ],
652
+ "logging_steps": 1,
653
+ "max_steps": 210,
654
+ "num_train_epochs": 2,
655
+ "save_steps": 50,
656
+ "total_flos": 8.146151922047386e+16,
657
+ "trial_name": null,
658
+ "trial_params": null
659
+ }
prompt-a/onechar/checkpoint-100/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ff85b6b2e99bcdbf030f4c883beac4553b8cad32f31f2d286263119b723c469
3
+ size 4411
prompt-a/onechar/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-a/onechar/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
+ "q_proj",
19
+ "k_proj",
20
+ "down_proj",
21
+ "up_proj",
22
+ "gate_proj",
23
+ "v_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
prompt-a/onechar/checkpoint-150/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:16b699052f2d128f83c6010ef6f104f26dcbf6cb4325eedcc9a7b58bc08c5c9f
3
+ size 125374989
prompt-a/onechar/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-a/onechar/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
+ "q_proj",
19
+ "k_proj",
20
+ "down_proj",
21
+ "up_proj",
22
+ "gate_proj",
23
+ "v_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
prompt-a/onechar/checkpoint-150/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:16b699052f2d128f83c6010ef6f104f26dcbf6cb4325eedcc9a7b58bc08c5c9f
3
+ size 125374989
prompt-a/onechar/checkpoint-150/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6697d2772f8bdabc0de5308ce64fed55e96f213ec392cc350ff9e35b3419011a
3
+ size 250681597
prompt-a/onechar/checkpoint-150/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:39b0267998bb7a1cfb283b31e1dd7afbe75c84407c53c4f1c298028d9daacc6e
3
+ size 21687
prompt-a/onechar/checkpoint-150/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:50e5dc96be0a674ddf476b916d068682bfc8d9f342c0213d1ba280d171cbc1d0
3
+ size 21687
prompt-a/onechar/checkpoint-150/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5f1ec8428ac308b6c6ffb63f0d2137a6c43b3b0f57031e333c2cde00dfbfedff
3
+ size 21687
prompt-a/onechar/checkpoint-150/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:832989103ea74caa2d18f41014c7ec87b33f665d85bbfabeac5e440a24a4e438
3
+ size 21687
prompt-a/onechar/checkpoint-150/rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c7c038cf0ec9ace1d9317b463a053999ebbd8efbc6568a0c63333109129be6de
3
+ size 21687
prompt-a/onechar/checkpoint-150/rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ea0b0a35fbb56436b89de780a77c2850d7ceb494df91befbfc0be4b4388f8c31
3
+ size 21687
prompt-a/onechar/checkpoint-150/rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:24a1fd82be4a9e4def96881742849b0f5444737a5c55f1de3d18ff9b31a835e5
3
+ size 21687
prompt-a/onechar/checkpoint-150/rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b924ccabce3c839d1b33edc854332d317e917f50d4f5defffe6a8f741a9641b0
3
+ size 21687
prompt-a/onechar/checkpoint-150/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d9abe1c80a7b715c51418717c88a5924cc27c160619ed0f1dec7010dc1687509
3
+ size 627
prompt-a/onechar/checkpoint-150/trainer_state.json ADDED
@@ -0,0 +1,975 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.4285714285714286,
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": 2.1535,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.02,
19
+ "learning_rate": 1.3e-05,
20
+ "loss": 2.1328,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.03,
25
+ "learning_rate": 1.9499999999999996e-05,
26
+ "loss": 2.0918,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.04,
31
+ "learning_rate": 2.6e-05,
32
+ "loss": 2.1194,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.05,
37
+ "learning_rate": 3.25e-05,
38
+ "loss": 2.2342,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.06,
43
+ "learning_rate": 3.899999999999999e-05,
44
+ "loss": 2.1168,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.07,
49
+ "learning_rate": 4.5499999999999995e-05,
50
+ "loss": 2.1761,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.08,
55
+ "learning_rate": 5.2e-05,
56
+ "loss": 2.0764,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.09,
61
+ "learning_rate": 5.85e-05,
62
+ "loss": 2.1711,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.1,
67
+ "learning_rate": 6.5e-05,
68
+ "loss": 2.0885,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.1,
73
+ "learning_rate": 6.499599055565396e-05,
74
+ "loss": 2.0196,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.11,
79
+ "learning_rate": 6.498396321188627e-05,
80
+ "loss": 2.1799,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.12,
85
+ "learning_rate": 6.496392093626402e-05,
86
+ "loss": 2.1322,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.13,
91
+ "learning_rate": 6.493586867391882e-05,
92
+ "loss": 2.1486,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.14,
97
+ "learning_rate": 6.489981334632665e-05,
98
+ "loss": 2.1351,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.15,
103
+ "learning_rate": 6.48557638496001e-05,
104
+ "loss": 2.1366,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.16,
109
+ "learning_rate": 6.480373105229334e-05,
110
+ "loss": 2.0727,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 0.17,
115
+ "learning_rate": 6.474372779272053e-05,
116
+ "loss": 2.1123,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 0.18,
121
+ "learning_rate": 6.467576887578811e-05,
122
+ "loss": 2.0652,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 0.19,
127
+ "learning_rate": 6.459987106934198e-05,
128
+ "loss": 2.0008,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 0.19,
133
+ "eval_loss": 2.0444116592407227,
134
+ "eval_runtime": 2.798,
135
+ "eval_samples_per_second": 3.574,
136
+ "eval_steps_per_second": 0.715,
137
+ "step": 20
138
+ },
139
+ {
140
+ "epoch": 0.2,
141
+ "learning_rate": 6.451605310003014e-05,
142
+ "loss": 2.0769,
143
+ "step": 21
144
+ },
145
+ {
146
+ "epoch": 0.21,
147
+ "learning_rate": 6.442433564868238e-05,
148
+ "loss": 1.9949,
149
+ "step": 22
150
+ },
151
+ {
152
+ "epoch": 0.22,
153
+ "learning_rate": 6.432474134520738e-05,
154
+ "loss": 2.1341,
155
+ "step": 23
156
+ },
157
+ {
158
+ "epoch": 0.23,
159
+ "learning_rate": 6.421729476300929e-05,
160
+ "loss": 2.1433,
161
+ "step": 24
162
+ },
163
+ {
164
+ "epoch": 0.24,
165
+ "learning_rate": 6.410202241292448e-05,
166
+ "loss": 2.0168,
167
+ "step": 25
168
+ },
169
+ {
170
+ "epoch": 0.25,
171
+ "learning_rate": 6.39789527366805e-05,
172
+ "loss": 2.0403,
173
+ "step": 26
174
+ },
175
+ {
176
+ "epoch": 0.26,
177
+ "learning_rate": 6.384811609987843e-05,
178
+ "loss": 2.1983,
179
+ "step": 27
180
+ },
181
+ {
182
+ "epoch": 0.27,
183
+ "learning_rate": 6.370954478450064e-05,
184
+ "loss": 2.0625,
185
+ "step": 28
186
+ },
187
+ {
188
+ "epoch": 0.28,
189
+ "learning_rate": 6.356327298094572e-05,
190
+ "loss": 2.204,
191
+ "step": 29
192
+ },
193
+ {
194
+ "epoch": 0.29,
195
+ "learning_rate": 6.340933677959248e-05,
196
+ "loss": 2.118,
197
+ "step": 30
198
+ },
199
+ {
200
+ "epoch": 0.3,
201
+ "learning_rate": 6.324777416189522e-05,
202
+ "loss": 2.0997,
203
+ "step": 31
204
+ },
205
+ {
206
+ "epoch": 0.3,
207
+ "learning_rate": 6.307862499101232e-05,
208
+ "loss": 2.1154,
209
+ "step": 32
210
+ },
211
+ {
212
+ "epoch": 0.31,
213
+ "learning_rate": 6.290193100197069e-05,
214
+ "loss": 2.0845,
215
+ "step": 33
216
+ },
217
+ {
218
+ "epoch": 0.32,
219
+ "learning_rate": 6.271773579136816e-05,
220
+ "loss": 2.0052,
221
+ "step": 34
222
+ },
223
+ {
224
+ "epoch": 0.33,
225
+ "learning_rate": 6.252608480661681e-05,
226
+ "loss": 1.956,
227
+ "step": 35
228
+ },
229
+ {
230
+ "epoch": 0.34,
231
+ "learning_rate": 6.232702533472938e-05,
232
+ "loss": 2.0958,
233
+ "step": 36
234
+ },
235
+ {
236
+ "epoch": 0.35,
237
+ "learning_rate": 6.212060649065197e-05,
238
+ "loss": 2.0634,
239
+ "step": 37
240
+ },
241
+ {
242
+ "epoch": 0.36,
243
+ "learning_rate": 6.190687920514563e-05,
244
+ "loss": 2.0033,
245
+ "step": 38
246
+ },
247
+ {
248
+ "epoch": 0.37,
249
+ "learning_rate": 6.168589621222e-05,
250
+ "loss": 2.1426,
251
+ "step": 39
252
+ },
253
+ {
254
+ "epoch": 0.38,
255
+ "learning_rate": 6.145771203612195e-05,
256
+ "loss": 1.9823,
257
+ "step": 40
258
+ },
259
+ {
260
+ "epoch": 0.38,
261
+ "eval_loss": 2.0226035118103027,
262
+ "eval_runtime": 2.8049,
263
+ "eval_samples_per_second": 3.565,
264
+ "eval_steps_per_second": 0.713,
265
+ "step": 40
266
+ },
267
+ {
268
+ "epoch": 0.39,
269
+ "learning_rate": 6.122238297788254e-05,
270
+ "loss": 2.2016,
271
+ "step": 41
272
+ },
273
+ {
274
+ "epoch": 0.4,
275
+ "learning_rate": 6.097996710142556e-05,
276
+ "loss": 2.1008,
277
+ "step": 42
278
+ },
279
+ {
280
+ "epoch": 0.41,
281
+ "learning_rate": 6.0730524219241214e-05,
282
+ "loss": 2.0821,
283
+ "step": 43
284
+ },
285
+ {
286
+ "epoch": 0.42,
287
+ "learning_rate": 6.047411587762816e-05,
288
+ "loss": 2.092,
289
+ "step": 44
290
+ },
291
+ {
292
+ "epoch": 0.43,
293
+ "learning_rate": 6.0210805341507994e-05,
294
+ "loss": 1.9617,
295
+ "step": 45
296
+ },
297
+ {
298
+ "epoch": 0.44,
299
+ "learning_rate": 5.994065757881549e-05,
300
+ "loss": 1.9358,
301
+ "step": 46
302
+ },
303
+ {
304
+ "epoch": 0.45,
305
+ "learning_rate": 5.966373924446878e-05,
306
+ "loss": 2.0573,
307
+ "step": 47
308
+ },
309
+ {
310
+ "epoch": 0.46,
311
+ "learning_rate": 5.938011866392325e-05,
312
+ "loss": 2.0731,
313
+ "step": 48
314
+ },
315
+ {
316
+ "epoch": 0.47,
317
+ "learning_rate": 5.9089865816313253e-05,
318
+ "loss": 2.2842,
319
+ "step": 49
320
+ },
321
+ {
322
+ "epoch": 0.48,
323
+ "learning_rate": 5.8793052317185785e-05,
324
+ "loss": 2.2145,
325
+ "step": 50
326
+ },
327
+ {
328
+ "epoch": 0.49,
329
+ "learning_rate": 5.848975140083044e-05,
330
+ "loss": 1.9923,
331
+ "step": 51
332
+ },
333
+ {
334
+ "epoch": 0.5,
335
+ "learning_rate": 5.818003790220993e-05,
336
+ "loss": 2.1065,
337
+ "step": 52
338
+ },
339
+ {
340
+ "epoch": 0.5,
341
+ "learning_rate": 5.786398823849571e-05,
342
+ "loss": 2.0009,
343
+ "step": 53
344
+ },
345
+ {
346
+ "epoch": 0.51,
347
+ "learning_rate": 5.754168039021315e-05,
348
+ "loss": 1.9785,
349
+ "step": 54
350
+ },
351
+ {
352
+ "epoch": 0.52,
353
+ "learning_rate": 5.7213193882001006e-05,
354
+ "loss": 1.9504,
355
+ "step": 55
356
+ },
357
+ {
358
+ "epoch": 0.53,
359
+ "learning_rate": 5.687860976298993e-05,
360
+ "loss": 2.1977,
361
+ "step": 56
362
+ },
363
+ {
364
+ "epoch": 0.54,
365
+ "learning_rate": 5.6538010586804816e-05,
366
+ "loss": 2.0423,
367
+ "step": 57
368
+ },
369
+ {
370
+ "epoch": 0.55,
371
+ "learning_rate": 5.619148039119587e-05,
372
+ "loss": 2.0536,
373
+ "step": 58
374
+ },
375
+ {
376
+ "epoch": 0.56,
377
+ "learning_rate": 5.5839104677303635e-05,
378
+ "loss": 2.0286,
379
+ "step": 59
380
+ },
381
+ {
382
+ "epoch": 0.57,
383
+ "learning_rate": 5.548097038856279e-05,
384
+ "loss": 2.1143,
385
+ "step": 60
386
+ },
387
+ {
388
+ "epoch": 0.57,
389
+ "eval_loss": 2.0128190517425537,
390
+ "eval_runtime": 2.8105,
391
+ "eval_samples_per_second": 3.558,
392
+ "eval_steps_per_second": 0.712,
393
+ "step": 60
394
+ },
395
+ {
396
+ "epoch": 0.58,
397
+ "learning_rate": 5.511716588925021e-05,
398
+ "loss": 2.1397,
399
+ "step": 61
400
+ },
401
+ {
402
+ "epoch": 0.59,
403
+ "learning_rate": 5.474778094268238e-05,
404
+ "loss": 2.0601,
405
+ "step": 62
406
+ },
407
+ {
408
+ "epoch": 0.6,
409
+ "learning_rate": 5.437290668906763e-05,
410
+ "loss": 2.1707,
411
+ "step": 63
412
+ },
413
+ {
414
+ "epoch": 0.61,
415
+ "learning_rate": 5.399263562301868e-05,
416
+ "loss": 2.0649,
417
+ "step": 64
418
+ },
419
+ {
420
+ "epoch": 0.62,
421
+ "learning_rate": 5.360706157073096e-05,
422
+ "loss": 2.0314,
423
+ "step": 65
424
+ },
425
+ {
426
+ "epoch": 0.63,
427
+ "learning_rate": 5.321627966683241e-05,
428
+ "loss": 2.1277,
429
+ "step": 66
430
+ },
431
+ {
432
+ "epoch": 0.64,
433
+ "learning_rate": 5.2820386330910424e-05,
434
+ "loss": 2.0534,
435
+ "step": 67
436
+ },
437
+ {
438
+ "epoch": 0.65,
439
+ "learning_rate": 5.2419479243721734e-05,
440
+ "loss": 1.9976,
441
+ "step": 68
442
+ },
443
+ {
444
+ "epoch": 0.66,
445
+ "learning_rate": 5.2013657323091226e-05,
446
+ "loss": 2.0262,
447
+ "step": 69
448
+ },
449
+ {
450
+ "epoch": 0.67,
451
+ "learning_rate": 5.160302069950537e-05,
452
+ "loss": 2.0093,
453
+ "step": 70
454
+ },
455
+ {
456
+ "epoch": 0.68,
457
+ "learning_rate": 5.118767069140655e-05,
458
+ "loss": 2.1088,
459
+ "step": 71
460
+ },
461
+ {
462
+ "epoch": 0.69,
463
+ "learning_rate": 5.076770978019424e-05,
464
+ "loss": 1.9841,
465
+ "step": 72
466
+ },
467
+ {
468
+ "epoch": 0.7,
469
+ "learning_rate": 5.034324158493928e-05,
470
+ "loss": 2.0081,
471
+ "step": 73
472
+ },
473
+ {
474
+ "epoch": 0.7,
475
+ "learning_rate": 4.991437083681739e-05,
476
+ "loss": 2.0179,
477
+ "step": 74
478
+ },
479
+ {
480
+ "epoch": 0.71,
481
+ "learning_rate": 4.948120335326833e-05,
482
+ "loss": 2.1304,
483
+ "step": 75
484
+ },
485
+ {
486
+ "epoch": 0.72,
487
+ "learning_rate": 4.904384601188706e-05,
488
+ "loss": 1.9213,
489
+ "step": 76
490
+ },
491
+ {
492
+ "epoch": 0.73,
493
+ "learning_rate": 4.860240672405324e-05,
494
+ "loss": 2.2153,
495
+ "step": 77
496
+ },
497
+ {
498
+ "epoch": 0.74,
499
+ "learning_rate": 4.815699440830574e-05,
500
+ "loss": 1.9365,
501
+ "step": 78
502
+ },
503
+ {
504
+ "epoch": 0.75,
505
+ "learning_rate": 4.770771896346863e-05,
506
+ "loss": 2.021,
507
+ "step": 79
508
+ },
509
+ {
510
+ "epoch": 0.76,
511
+ "learning_rate": 4.725469124153526e-05,
512
+ "loss": 2.0327,
513
+ "step": 80
514
+ },
515
+ {
516
+ "epoch": 0.76,
517
+ "eval_loss": 2.0059409141540527,
518
+ "eval_runtime": 2.8077,
519
+ "eval_samples_per_second": 3.562,
520
+ "eval_steps_per_second": 0.712,
521
+ "step": 80
522
+ },
523
+ {
524
+ "epoch": 0.77,
525
+ "learning_rate": 4.679802302031724e-05,
526
+ "loss": 2.1576,
527
+ "step": 81
528
+ },
529
+ {
530
+ "epoch": 0.78,
531
+ "learning_rate": 4.6337826975864856e-05,
532
+ "loss": 1.9468,
533
+ "step": 82
534
+ },
535
+ {
536
+ "epoch": 0.79,
537
+ "learning_rate": 4.587421665466603e-05,
538
+ "loss": 2.0795,
539
+ "step": 83
540
+ },
541
+ {
542
+ "epoch": 0.8,
543
+ "learning_rate": 4.540730644563036e-05,
544
+ "loss": 2.0566,
545
+ "step": 84
546
+ },
547
+ {
548
+ "epoch": 0.81,
549
+ "learning_rate": 4.4937211551865415e-05,
550
+ "loss": 2.0819,
551
+ "step": 85
552
+ },
553
+ {
554
+ "epoch": 0.82,
555
+ "learning_rate": 4.4464047962252035e-05,
556
+ "loss": 2.0542,
557
+ "step": 86
558
+ },
559
+ {
560
+ "epoch": 0.83,
561
+ "learning_rate": 4.398793242282586e-05,
562
+ "loss": 2.0006,
563
+ "step": 87
564
+ },
565
+ {
566
+ "epoch": 0.84,
567
+ "learning_rate": 4.350898240797197e-05,
568
+ "loss": 1.9953,
569
+ "step": 88
570
+ },
571
+ {
572
+ "epoch": 0.85,
573
+ "learning_rate": 4.3027316091439854e-05,
574
+ "loss": 2.067,
575
+ "step": 89
576
+ },
577
+ {
578
+ "epoch": 0.86,
579
+ "learning_rate": 4.254305231718579e-05,
580
+ "loss": 1.9909,
581
+ "step": 90
582
+ },
583
+ {
584
+ "epoch": 0.87,
585
+ "learning_rate": 4.205631057004987e-05,
586
+ "loss": 1.9436,
587
+ "step": 91
588
+ },
589
+ {
590
+ "epoch": 0.88,
591
+ "learning_rate": 4.156721094627495e-05,
592
+ "loss": 2.0643,
593
+ "step": 92
594
+ },
595
+ {
596
+ "epoch": 0.89,
597
+ "learning_rate": 4.107587412387462e-05,
598
+ "loss": 2.1244,
599
+ "step": 93
600
+ },
601
+ {
602
+ "epoch": 0.9,
603
+ "learning_rate": 4.058242133285779e-05,
604
+ "loss": 1.9795,
605
+ "step": 94
606
+ },
607
+ {
608
+ "epoch": 0.9,
609
+ "learning_rate": 4.008697432531693e-05,
610
+ "loss": 2.1568,
611
+ "step": 95
612
+ },
613
+ {
614
+ "epoch": 0.91,
615
+ "learning_rate": 3.9589655345387633e-05,
616
+ "loss": 1.9829,
617
+ "step": 96
618
+ },
619
+ {
620
+ "epoch": 0.92,
621
+ "learning_rate": 3.909058709908665e-05,
622
+ "loss": 2.031,
623
+ "step": 97
624
+ },
625
+ {
626
+ "epoch": 0.93,
627
+ "learning_rate": 3.858989272403605e-05,
628
+ "loss": 2.2042,
629
+ "step": 98
630
+ },
631
+ {
632
+ "epoch": 0.94,
633
+ "learning_rate": 3.8087695759080804e-05,
634
+ "loss": 1.9775,
635
+ "step": 99
636
+ },
637
+ {
638
+ "epoch": 0.95,
639
+ "learning_rate": 3.75841201138075e-05,
640
+ "loss": 2.0274,
641
+ "step": 100
642
+ },
643
+ {
644
+ "epoch": 0.95,
645
+ "eval_loss": 2.0003483295440674,
646
+ "eval_runtime": 2.8075,
647
+ "eval_samples_per_second": 3.562,
648
+ "eval_steps_per_second": 0.712,
649
+ "step": 100
650
+ },
651
+ {
652
+ "epoch": 0.96,
653
+ "learning_rate": 3.707929003797143e-05,
654
+ "loss": 2.0745,
655
+ "step": 101
656
+ },
657
+ {
658
+ "epoch": 0.97,
659
+ "learning_rate": 3.6573330090839886e-05,
660
+ "loss": 1.9933,
661
+ "step": 102
662
+ },
663
+ {
664
+ "epoch": 0.98,
665
+ "learning_rate": 3.6066365110458966e-05,
666
+ "loss": 2.127,
667
+ "step": 103
668
+ },
669
+ {
670
+ "epoch": 0.99,
671
+ "learning_rate": 3.555852018285172e-05,
672
+ "loss": 2.0499,
673
+ "step": 104
674
+ },
675
+ {
676
+ "epoch": 1.0,
677
+ "learning_rate": 3.504992061115496e-05,
678
+ "loss": 1.9391,
679
+ "step": 105
680
+ },
681
+ {
682
+ "epoch": 1.01,
683
+ "learning_rate": 3.4540691884702686e-05,
684
+ "loss": 2.0216,
685
+ "step": 106
686
+ },
687
+ {
688
+ "epoch": 1.02,
689
+ "learning_rate": 3.4030959648063384e-05,
690
+ "loss": 2.0055,
691
+ "step": 107
692
+ },
693
+ {
694
+ "epoch": 1.03,
695
+ "learning_rate": 3.3520849670039165e-05,
696
+ "loss": 2.0721,
697
+ "step": 108
698
+ },
699
+ {
700
+ "epoch": 1.04,
701
+ "learning_rate": 3.301048781263417e-05,
702
+ "loss": 2.0274,
703
+ "step": 109
704
+ },
705
+ {
706
+ "epoch": 1.05,
707
+ "learning_rate": 3.25e-05,
708
+ "loss": 1.9373,
709
+ "step": 110
710
+ },
711
+ {
712
+ "epoch": 1.06,
713
+ "learning_rate": 3.198951218736583e-05,
714
+ "loss": 1.9034,
715
+ "step": 111
716
+ },
717
+ {
718
+ "epoch": 1.07,
719
+ "learning_rate": 3.147915032996083e-05,
720
+ "loss": 1.8759,
721
+ "step": 112
722
+ },
723
+ {
724
+ "epoch": 1.08,
725
+ "learning_rate": 3.096904035193661e-05,
726
+ "loss": 2.0572,
727
+ "step": 113
728
+ },
729
+ {
730
+ "epoch": 1.09,
731
+ "learning_rate": 3.0459308115297312e-05,
732
+ "loss": 2.0072,
733
+ "step": 114
734
+ },
735
+ {
736
+ "epoch": 1.1,
737
+ "learning_rate": 2.9950079388845042e-05,
738
+ "loss": 1.9129,
739
+ "step": 115
740
+ },
741
+ {
742
+ "epoch": 1.1,
743
+ "learning_rate": 2.944147981714828e-05,
744
+ "loss": 2.0966,
745
+ "step": 116
746
+ },
747
+ {
748
+ "epoch": 1.11,
749
+ "learning_rate": 2.8933634889541028e-05,
750
+ "loss": 2.0903,
751
+ "step": 117
752
+ },
753
+ {
754
+ "epoch": 1.12,
755
+ "learning_rate": 2.8426669909160105e-05,
756
+ "loss": 2.0917,
757
+ "step": 118
758
+ },
759
+ {
760
+ "epoch": 1.13,
761
+ "learning_rate": 2.7920709962028562e-05,
762
+ "loss": 2.089,
763
+ "step": 119
764
+ },
765
+ {
766
+ "epoch": 1.14,
767
+ "learning_rate": 2.741587988619249e-05,
768
+ "loss": 1.9964,
769
+ "step": 120
770
+ },
771
+ {
772
+ "epoch": 1.14,
773
+ "eval_loss": 1.9988892078399658,
774
+ "eval_runtime": 2.8188,
775
+ "eval_samples_per_second": 3.548,
776
+ "eval_steps_per_second": 0.71,
777
+ "step": 120
778
+ },
779
+ {
780
+ "epoch": 1.15,
781
+ "learning_rate": 2.6912304240919183e-05,
782
+ "loss": 1.9754,
783
+ "step": 121
784
+ },
785
+ {
786
+ "epoch": 1.16,
787
+ "learning_rate": 2.641010727596394e-05,
788
+ "loss": 2.1508,
789
+ "step": 122
790
+ },
791
+ {
792
+ "epoch": 1.17,
793
+ "learning_rate": 2.5909412900913353e-05,
794
+ "loss": 2.0331,
795
+ "step": 123
796
+ },
797
+ {
798
+ "epoch": 1.18,
799
+ "learning_rate": 2.5410344654612374e-05,
800
+ "loss": 1.9607,
801
+ "step": 124
802
+ },
803
+ {
804
+ "epoch": 1.19,
805
+ "learning_rate": 2.4913025674683083e-05,
806
+ "loss": 2.0356,
807
+ "step": 125
808
+ },
809
+ {
810
+ "epoch": 1.2,
811
+ "learning_rate": 2.4417578667142222e-05,
812
+ "loss": 2.0417,
813
+ "step": 126
814
+ },
815
+ {
816
+ "epoch": 1.21,
817
+ "learning_rate": 2.3924125876125387e-05,
818
+ "loss": 1.9246,
819
+ "step": 127
820
+ },
821
+ {
822
+ "epoch": 1.22,
823
+ "learning_rate": 2.343278905372505e-05,
824
+ "loss": 2.0911,
825
+ "step": 128
826
+ },
827
+ {
828
+ "epoch": 1.23,
829
+ "learning_rate": 2.2943689429950124e-05,
830
+ "loss": 1.9324,
831
+ "step": 129
832
+ },
833
+ {
834
+ "epoch": 1.24,
835
+ "learning_rate": 2.2456947682814208e-05,
836
+ "loss": 1.9987,
837
+ "step": 130
838
+ },
839
+ {
840
+ "epoch": 1.25,
841
+ "learning_rate": 2.1972683908560147e-05,
842
+ "loss": 2.0121,
843
+ "step": 131
844
+ },
845
+ {
846
+ "epoch": 1.26,
847
+ "learning_rate": 2.149101759202803e-05,
848
+ "loss": 2.0002,
849
+ "step": 132
850
+ },
851
+ {
852
+ "epoch": 1.27,
853
+ "learning_rate": 2.1012067577174147e-05,
854
+ "loss": 2.0464,
855
+ "step": 133
856
+ },
857
+ {
858
+ "epoch": 1.28,
859
+ "learning_rate": 2.0535952037747966e-05,
860
+ "loss": 2.1422,
861
+ "step": 134
862
+ },
863
+ {
864
+ "epoch": 1.29,
865
+ "learning_rate": 2.0062788448134583e-05,
866
+ "loss": 1.9796,
867
+ "step": 135
868
+ },
869
+ {
870
+ "epoch": 1.3,
871
+ "learning_rate": 1.959269355436963e-05,
872
+ "loss": 1.9683,
873
+ "step": 136
874
+ },
875
+ {
876
+ "epoch": 1.3,
877
+ "learning_rate": 1.9125783345333963e-05,
878
+ "loss": 2.1488,
879
+ "step": 137
880
+ },
881
+ {
882
+ "epoch": 1.31,
883
+ "learning_rate": 1.866217302413514e-05,
884
+ "loss": 1.9852,
885
+ "step": 138
886
+ },
887
+ {
888
+ "epoch": 1.32,
889
+ "learning_rate": 1.8201976979682756e-05,
890
+ "loss": 2.0087,
891
+ "step": 139
892
+ },
893
+ {
894
+ "epoch": 1.33,
895
+ "learning_rate": 1.774530875846473e-05,
896
+ "loss": 1.9724,
897
+ "step": 140
898
+ },
899
+ {
900
+ "epoch": 1.33,
901
+ "eval_loss": 1.9960596561431885,
902
+ "eval_runtime": 2.8086,
903
+ "eval_samples_per_second": 3.56,
904
+ "eval_steps_per_second": 0.712,
905
+ "step": 140
906
+ },
907
+ {
908
+ "epoch": 1.34,
909
+ "learning_rate": 1.7292281036531367e-05,
910
+ "loss": 2.0022,
911
+ "step": 141
912
+ },
913
+ {
914
+ "epoch": 1.35,
915
+ "learning_rate": 1.6843005591694247e-05,
916
+ "loss": 2.0827,
917
+ "step": 142
918
+ },
919
+ {
920
+ "epoch": 1.36,
921
+ "learning_rate": 1.639759327594675e-05,
922
+ "loss": 2.0052,
923
+ "step": 143
924
+ },
925
+ {
926
+ "epoch": 1.37,
927
+ "learning_rate": 1.595615398811293e-05,
928
+ "loss": 1.9938,
929
+ "step": 144
930
+ },
931
+ {
932
+ "epoch": 1.38,
933
+ "learning_rate": 1.551879664673166e-05,
934
+ "loss": 1.996,
935
+ "step": 145
936
+ },
937
+ {
938
+ "epoch": 1.39,
939
+ "learning_rate": 1.5085629163182601e-05,
940
+ "loss": 2.0065,
941
+ "step": 146
942
+ },
943
+ {
944
+ "epoch": 1.4,
945
+ "learning_rate": 1.4656758415060712e-05,
946
+ "loss": 2.0048,
947
+ "step": 147
948
+ },
949
+ {
950
+ "epoch": 1.41,
951
+ "learning_rate": 1.4232290219805762e-05,
952
+ "loss": 1.8846,
953
+ "step": 148
954
+ },
955
+ {
956
+ "epoch": 1.42,
957
+ "learning_rate": 1.3812329308593457e-05,
958
+ "loss": 1.9582,
959
+ "step": 149
960
+ },
961
+ {
962
+ "epoch": 1.43,
963
+ "learning_rate": 1.3396979300494626e-05,
964
+ "loss": 2.0509,
965
+ "step": 150
966
+ }
967
+ ],
968
+ "logging_steps": 1,
969
+ "max_steps": 210,
970
+ "num_train_epochs": 2,
971
+ "save_steps": 50,
972
+ "total_flos": 1.2206169951043584e+17,
973
+ "trial_name": null,
974
+ "trial_params": null
975
+ }
prompt-a/onechar/checkpoint-150/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ff85b6b2e99bcdbf030f4c883beac4553b8cad32f31f2d286263119b723c469
3
+ size 4411
prompt-a/onechar/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-a/onechar/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
+ "q_proj",
19
+ "k_proj",
20
+ "down_proj",
21
+ "up_proj",
22
+ "gate_proj",
23
+ "v_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
prompt-a/onechar/checkpoint-200/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0aeeda740ab6bb8ee51e6b7ecd7a2848397fd35d1b4db657df6d8fa4dac2f048
3
+ size 125374989
prompt-a/onechar/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-a/onechar/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
+ "q_proj",
19
+ "k_proj",
20
+ "down_proj",
21
+ "up_proj",
22
+ "gate_proj",
23
+ "v_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
prompt-a/onechar/checkpoint-200/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0aeeda740ab6bb8ee51e6b7ecd7a2848397fd35d1b4db657df6d8fa4dac2f048
3
+ size 125374989
prompt-a/onechar/checkpoint-200/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:696142d361db2f7734fc31c617f47091a5690b7733ed8c4a21b5b59a2ee213a8
3
+ size 250681597
prompt-a/onechar/checkpoint-200/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02519202233b2f973159654f70ae8d8d74816c44a0c980e2ec2bc67267e89056
3
+ size 21687
prompt-a/onechar/checkpoint-200/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a7022483dc2b7110805d232d89f98606595b39d55b0309c029f39838461bfa05
3
+ size 21687
prompt-a/onechar/checkpoint-200/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fbee7570d88f852abb0c601b61e086fab06332833e7e65aa5e6cac17720c5a40
3
+ size 21687