Hugging Face's logo

eswardivi
/
medical_qa_llm

PEFT
Model card Files
xet
Community
1
eswardivi commited on
Commit
09f95cc
·
1 Parent(s): 416e14c

Initial commit

Browse files
Files changed (27) hide show
  1. README.md +20 -0
  2. adapter_config.json +22 -0
  3. adapter_model.bin +3 -0
  4. checkpoint-200/README.md +20 -0
  5. checkpoint-200/adapter_config.json +22 -0
  6. checkpoint-200/adapter_model.bin +3 -0
  7. checkpoint-200/optimizer.pt +3 -0
  8. checkpoint-200/rng_state.pth +3 -0
  9. checkpoint-200/scheduler.pt +3 -0
  10. checkpoint-200/trainer_state.json +144 -0
  11. checkpoint-200/training_args.bin +3 -0
  12. checkpoint-400/README.md +20 -0
  13. checkpoint-400/adapter_config.json +22 -0
  14. checkpoint-400/adapter_model.bin +3 -0
  15. checkpoint-400/optimizer.pt +3 -0
  16. checkpoint-400/rng_state.pth +3 -0
  17. checkpoint-400/scheduler.pt +3 -0
  18. checkpoint-400/trainer_state.json +272 -0
  19. checkpoint-400/training_args.bin +3 -0
  20. checkpoint-600/README.md +20 -0
  21. checkpoint-600/adapter_config.json +22 -0
  22. checkpoint-600/adapter_model.bin +3 -0
  23. checkpoint-600/optimizer.pt +3 -0
  24. checkpoint-600/rng_state.pth +3 -0
  25. checkpoint-600/scheduler.pt +3 -0
  26. checkpoint-600/trainer_state.json +400 -0
  27. checkpoint-600/training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: True
9
+ - load_in_4bit: False
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: fp4
15
+ - bnb_4bit_use_double_quant: False
16
+ - bnb_4bit_compute_dtype: float32
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0.dev0
adapter_config.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model_name_or_path": "yahma/llama-7b-hf",
3
+ "bias": "none",
4
+ "fan_in_fan_out": false,
5
+ "inference_mode": true,
6
+ "init_lora_weights": true,
7
+ "layers_pattern": null,
8
+ "layers_to_transform": null,
9
+ "lora_alpha": 16,
10
+ "lora_dropout": 0.05,
11
+ "modules_to_save": null,
12
+ "peft_type": "LORA",
13
+ "r": 16,
14
+ "revision": null,
15
+ "target_modules": [
16
+ "q_proj",
17
+ "k_proj",
18
+ "v_proj",
19
+ "o_proj"
20
+ ],
21
+ "task_type": "CAUSAL_LM"
22
+ }
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f0a8c6aa44f6ee4f436d5eb16124b7291b40797359cb730d9fb877dc53053b63
3
+ size 67201357
checkpoint-200/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: True
9
+ - load_in_4bit: False
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: fp4
15
+ - bnb_4bit_use_double_quant: False
16
+ - bnb_4bit_compute_dtype: float32
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0.dev0
checkpoint-200/adapter_config.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model_name_or_path": "yahma/llama-7b-hf",
3
+ "bias": "none",
4
+ "fan_in_fan_out": false,
5
+ "inference_mode": true,
6
+ "init_lora_weights": true,
7
+ "layers_pattern": null,
8
+ "layers_to_transform": null,
9
+ "lora_alpha": 16,
10
+ "lora_dropout": 0.05,
11
+ "modules_to_save": null,
12
+ "peft_type": "LORA",
13
+ "r": 16,
14
+ "revision": null,
15
+ "target_modules": [
16
+ "q_proj",
17
+ "k_proj",
18
+ "v_proj",
19
+ "o_proj"
20
+ ],
21
+ "task_type": "CAUSAL_LM"
22
+ }
checkpoint-200/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:65e3cf0b652f6f23047fed81f603a8848b40bd5faf583cc91c10d8e08977660e
3
+ size 67201357
checkpoint-200/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:33b0b632ed11a77cef9edfea20b880a7b4991b9f900fbcd552cbb77fe2dfe992
3
+ size 134433093
checkpoint-200/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5bf69a516de178c197028190cf855cb08be74424f51f596123b5920fcd13caee
3
+ size 14575
checkpoint-200/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:14ff332e73d427c2d06dcc79abfec8c5b75daa4bcfdaa702c1f500671e781e6c
3
+ size 627
checkpoint-200/trainer_state.json ADDED
@@ -0,0 +1,144 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 1.4607141017913818,
3
+ "best_model_checkpoint": "./lora-alpaca-1/checkpoint-200",
4
+ "epoch": 4.2283298097251585,
5
+ "global_step": 200,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.21,
12
+ "learning_rate": 2.9999999999999997e-05,
13
+ "loss": 2.9303,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.42,
18
+ "learning_rate": 5.9999999999999995e-05,
19
+ "loss": 2.8354,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.63,
24
+ "learning_rate": 8.999999999999999e-05,
25
+ "loss": 2.4967,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.85,
30
+ "learning_rate": 0.00011999999999999999,
31
+ "loss": 2.0483,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 1.06,
36
+ "learning_rate": 0.00015,
37
+ "loss": 1.6662,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 1.27,
42
+ "learning_rate": 0.00017999999999999998,
43
+ "loss": 1.4603,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 1.48,
48
+ "learning_rate": 0.00020999999999999998,
49
+ "loss": 1.3619,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 1.69,
54
+ "learning_rate": 0.00023999999999999998,
55
+ "loss": 1.3654,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 1.9,
60
+ "learning_rate": 0.00027,
61
+ "loss": 1.4322,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 2.11,
66
+ "learning_rate": 0.0003,
67
+ "loss": 1.4571,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 2.33,
72
+ "learning_rate": 0.0002950413223140496,
73
+ "loss": 1.3471,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 2.54,
78
+ "learning_rate": 0.00029008264462809916,
79
+ "loss": 1.3187,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 2.75,
84
+ "learning_rate": 0.0002851239669421488,
85
+ "loss": 1.3126,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 2.96,
90
+ "learning_rate": 0.00028016528925619834,
91
+ "loss": 1.2193,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 3.17,
96
+ "learning_rate": 0.0002752066115702479,
97
+ "loss": 1.3303,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 3.38,
102
+ "learning_rate": 0.0002702479338842975,
103
+ "loss": 1.3181,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 3.59,
108
+ "learning_rate": 0.0002652892561983471,
109
+ "loss": 1.3102,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 3.81,
114
+ "learning_rate": 0.00026033057851239665,
115
+ "loss": 1.2773,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 4.02,
120
+ "learning_rate": 0.00025537190082644627,
121
+ "loss": 1.1511,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 4.23,
126
+ "learning_rate": 0.00025041322314049583,
127
+ "loss": 1.1912,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 4.23,
132
+ "eval_loss": 1.4607141017913818,
133
+ "eval_runtime": 81.1682,
134
+ "eval_samples_per_second": 7.774,
135
+ "eval_steps_per_second": 0.973,
136
+ "step": 200
137
+ }
138
+ ],
139
+ "max_steps": 705,
140
+ "num_train_epochs": 15,
141
+ "total_flos": 1.4226067884736512e+17,
142
+ "trial_name": null,
143
+ "trial_params": null
144
+ }
checkpoint-200/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a46f4da27d88d972b2396250eeef005295a27631879666feecc45de066a6d6c0
3
+ size 3963
checkpoint-400/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: True
9
+ - load_in_4bit: False
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: fp4
15
+ - bnb_4bit_use_double_quant: False
16
+ - bnb_4bit_compute_dtype: float32
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0.dev0
checkpoint-400/adapter_config.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model_name_or_path": "yahma/llama-7b-hf",
3
+ "bias": "none",
4
+ "fan_in_fan_out": false,
5
+ "inference_mode": true,
6
+ "init_lora_weights": true,
7
+ "layers_pattern": null,
8
+ "layers_to_transform": null,
9
+ "lora_alpha": 16,
10
+ "lora_dropout": 0.05,
11
+ "modules_to_save": null,
12
+ "peft_type": "LORA",
13
+ "r": 16,
14
+ "revision": null,
15
+ "target_modules": [
16
+ "q_proj",
17
+ "k_proj",
18
+ "v_proj",
19
+ "o_proj"
20
+ ],
21
+ "task_type": "CAUSAL_LM"
22
+ }
checkpoint-400/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:71391cd42351bcc3a83955a29f0f7ea21443f3a35cc8a53f4f4dc6192b383f5f
3
+ size 67201357
checkpoint-400/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a8b076fadee8020413be33c546197fbe9915a6819c84b2c82d179af9e1464bc0
3
+ size 134433093
checkpoint-400/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:48c31f4c339523e764559633a7a4b525006068964af33e089b9b43ec3c33a8f8
3
+ size 14575
checkpoint-400/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ef6d9110e8161287655fc5b751c12cc12ea3a4485e2e210e659c356df526e4e
3
+ size 627
checkpoint-400/trainer_state.json ADDED
@@ -0,0 +1,272 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 1.4225733280181885,
3
+ "best_model_checkpoint": "./lora-alpaca-1/checkpoint-400",
4
+ "epoch": 8.456659619450317,
5
+ "global_step": 400,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.21,
12
+ "learning_rate": 2.9999999999999997e-05,
13
+ "loss": 2.9303,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.42,
18
+ "learning_rate": 5.9999999999999995e-05,
19
+ "loss": 2.8354,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.63,
24
+ "learning_rate": 8.999999999999999e-05,
25
+ "loss": 2.4967,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.85,
30
+ "learning_rate": 0.00011999999999999999,
31
+ "loss": 2.0483,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 1.06,
36
+ "learning_rate": 0.00015,
37
+ "loss": 1.6662,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 1.27,
42
+ "learning_rate": 0.00017999999999999998,
43
+ "loss": 1.4603,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 1.48,
48
+ "learning_rate": 0.00020999999999999998,
49
+ "loss": 1.3619,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 1.69,
54
+ "learning_rate": 0.00023999999999999998,
55
+ "loss": 1.3654,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 1.9,
60
+ "learning_rate": 0.00027,
61
+ "loss": 1.4322,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 2.11,
66
+ "learning_rate": 0.0003,
67
+ "loss": 1.4571,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 2.33,
72
+ "learning_rate": 0.0002950413223140496,
73
+ "loss": 1.3471,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 2.54,
78
+ "learning_rate": 0.00029008264462809916,
79
+ "loss": 1.3187,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 2.75,
84
+ "learning_rate": 0.0002851239669421488,
85
+ "loss": 1.3126,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 2.96,
90
+ "learning_rate": 0.00028016528925619834,
91
+ "loss": 1.2193,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 3.17,
96
+ "learning_rate": 0.0002752066115702479,
97
+ "loss": 1.3303,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 3.38,
102
+ "learning_rate": 0.0002702479338842975,
103
+ "loss": 1.3181,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 3.59,
108
+ "learning_rate": 0.0002652892561983471,
109
+ "loss": 1.3102,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 3.81,
114
+ "learning_rate": 0.00026033057851239665,
115
+ "loss": 1.2773,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 4.02,
120
+ "learning_rate": 0.00025537190082644627,
121
+ "loss": 1.1511,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 4.23,
126
+ "learning_rate": 0.00025041322314049583,
127
+ "loss": 1.1912,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 4.23,
132
+ "eval_loss": 1.4607141017913818,
133
+ "eval_runtime": 81.1682,
134
+ "eval_samples_per_second": 7.774,
135
+ "eval_steps_per_second": 0.973,
136
+ "step": 200
137
+ },
138
+ {
139
+ "epoch": 4.44,
140
+ "learning_rate": 0.00024545454545454545,
141
+ "loss": 1.278,
142
+ "step": 210
143
+ },
144
+ {
145
+ "epoch": 4.65,
146
+ "learning_rate": 0.000240495867768595,
147
+ "loss": 1.2741,
148
+ "step": 220
149
+ },
150
+ {
151
+ "epoch": 4.86,
152
+ "learning_rate": 0.00023553719008264463,
153
+ "loss": 1.2626,
154
+ "step": 230
155
+ },
156
+ {
157
+ "epoch": 5.07,
158
+ "learning_rate": 0.0002305785123966942,
159
+ "loss": 1.2519,
160
+ "step": 240
161
+ },
162
+ {
163
+ "epoch": 5.29,
164
+ "learning_rate": 0.00022561983471074378,
165
+ "loss": 1.1882,
166
+ "step": 250
167
+ },
168
+ {
169
+ "epoch": 5.5,
170
+ "learning_rate": 0.00022066115702479338,
171
+ "loss": 1.1867,
172
+ "step": 260
173
+ },
174
+ {
175
+ "epoch": 5.71,
176
+ "learning_rate": 0.00021570247933884297,
177
+ "loss": 1.1278,
178
+ "step": 270
179
+ },
180
+ {
181
+ "epoch": 5.92,
182
+ "learning_rate": 0.00021074380165289253,
183
+ "loss": 1.1719,
184
+ "step": 280
185
+ },
186
+ {
187
+ "epoch": 6.13,
188
+ "learning_rate": 0.00020578512396694215,
189
+ "loss": 1.3097,
190
+ "step": 290
191
+ },
192
+ {
193
+ "epoch": 6.34,
194
+ "learning_rate": 0.0002008264462809917,
195
+ "loss": 1.192,
196
+ "step": 300
197
+ },
198
+ {
199
+ "epoch": 6.55,
200
+ "learning_rate": 0.00019586776859504133,
201
+ "loss": 1.1772,
202
+ "step": 310
203
+ },
204
+ {
205
+ "epoch": 6.77,
206
+ "learning_rate": 0.0001909090909090909,
207
+ "loss": 1.1579,
208
+ "step": 320
209
+ },
210
+ {
211
+ "epoch": 6.98,
212
+ "learning_rate": 0.00018595041322314049,
213
+ "loss": 1.0586,
214
+ "step": 330
215
+ },
216
+ {
217
+ "epoch": 7.19,
218
+ "learning_rate": 0.00018099173553719008,
219
+ "loss": 1.1763,
220
+ "step": 340
221
+ },
222
+ {
223
+ "epoch": 7.4,
224
+ "learning_rate": 0.00017603305785123967,
225
+ "loss": 1.1715,
226
+ "step": 350
227
+ },
228
+ {
229
+ "epoch": 7.61,
230
+ "learning_rate": 0.00017107438016528923,
231
+ "loss": 1.1778,
232
+ "step": 360
233
+ },
234
+ {
235
+ "epoch": 7.82,
236
+ "learning_rate": 0.00016611570247933885,
237
+ "loss": 1.1716,
238
+ "step": 370
239
+ },
240
+ {
241
+ "epoch": 8.03,
242
+ "learning_rate": 0.00016115702479338841,
243
+ "loss": 1.1188,
244
+ "step": 380
245
+ },
246
+ {
247
+ "epoch": 8.25,
248
+ "learning_rate": 0.000156198347107438,
249
+ "loss": 1.0312,
250
+ "step": 390
251
+ },
252
+ {
253
+ "epoch": 8.46,
254
+ "learning_rate": 0.0001512396694214876,
255
+ "loss": 1.0545,
256
+ "step": 400
257
+ },
258
+ {
259
+ "epoch": 8.46,
260
+ "eval_loss": 1.4225733280181885,
261
+ "eval_runtime": 81.2375,
262
+ "eval_samples_per_second": 7.767,
263
+ "eval_steps_per_second": 0.972,
264
+ "step": 400
265
+ }
266
+ ],
267
+ "max_steps": 705,
268
+ "num_train_epochs": 15,
269
+ "total_flos": 2.84040605501227e+17,
270
+ "trial_name": null,
271
+ "trial_params": null
272
+ }
checkpoint-400/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a46f4da27d88d972b2396250eeef005295a27631879666feecc45de066a6d6c0
3
+ size 3963
checkpoint-600/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: True
9
+ - load_in_4bit: False
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: fp4
15
+ - bnb_4bit_use_double_quant: False
16
+ - bnb_4bit_compute_dtype: float32
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0.dev0
checkpoint-600/adapter_config.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model_name_or_path": "yahma/llama-7b-hf",
3
+ "bias": "none",
4
+ "fan_in_fan_out": false,
5
+ "inference_mode": true,
6
+ "init_lora_weights": true,
7
+ "layers_pattern": null,
8
+ "layers_to_transform": null,
9
+ "lora_alpha": 16,
10
+ "lora_dropout": 0.05,
11
+ "modules_to_save": null,
12
+ "peft_type": "LORA",
13
+ "r": 16,
14
+ "revision": null,
15
+ "target_modules": [
16
+ "q_proj",
17
+ "k_proj",
18
+ "v_proj",
19
+ "o_proj"
20
+ ],
21
+ "task_type": "CAUSAL_LM"
22
+ }
checkpoint-600/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f0a8c6aa44f6ee4f436d5eb16124b7291b40797359cb730d9fb877dc53053b63
3
+ size 67201357
checkpoint-600/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:11dc788304ace1d486a22be656eade755118f83dce4d7e70720bb85451ceb0e0
3
+ size 134433093
checkpoint-600/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:951fe833863d1da2daea5714e8ce50a0d33136c9f37c89f3d4ab27fb6e82f3d5
3
+ size 14575
checkpoint-600/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f1a7ecb136708877b1a2fcea65d3301032811012ba0e827c1cd4060c7ccf2e8
3
+ size 627
checkpoint-600/trainer_state.json ADDED
@@ -0,0 +1,400 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 1.4200056791305542,
3
+ "best_model_checkpoint": "./lora-alpaca-1/checkpoint-600",
4
+ "epoch": 12.684989429175475,
5
+ "global_step": 600,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.21,
12
+ "learning_rate": 2.9999999999999997e-05,
13
+ "loss": 2.9303,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.42,
18
+ "learning_rate": 5.9999999999999995e-05,
19
+ "loss": 2.8354,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.63,
24
+ "learning_rate": 8.999999999999999e-05,
25
+ "loss": 2.4967,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.85,
30
+ "learning_rate": 0.00011999999999999999,
31
+ "loss": 2.0483,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 1.06,
36
+ "learning_rate": 0.00015,
37
+ "loss": 1.6662,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 1.27,
42
+ "learning_rate": 0.00017999999999999998,
43
+ "loss": 1.4603,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 1.48,
48
+ "learning_rate": 0.00020999999999999998,
49
+ "loss": 1.3619,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 1.69,
54
+ "learning_rate": 0.00023999999999999998,
55
+ "loss": 1.3654,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 1.9,
60
+ "learning_rate": 0.00027,
61
+ "loss": 1.4322,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 2.11,
66
+ "learning_rate": 0.0003,
67
+ "loss": 1.4571,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 2.33,
72
+ "learning_rate": 0.0002950413223140496,
73
+ "loss": 1.3471,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 2.54,
78
+ "learning_rate": 0.00029008264462809916,
79
+ "loss": 1.3187,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 2.75,
84
+ "learning_rate": 0.0002851239669421488,
85
+ "loss": 1.3126,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 2.96,
90
+ "learning_rate": 0.00028016528925619834,
91
+ "loss": 1.2193,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 3.17,
96
+ "learning_rate": 0.0002752066115702479,
97
+ "loss": 1.3303,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 3.38,
102
+ "learning_rate": 0.0002702479338842975,
103
+ "loss": 1.3181,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 3.59,
108
+ "learning_rate": 0.0002652892561983471,
109
+ "loss": 1.3102,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 3.81,
114
+ "learning_rate": 0.00026033057851239665,
115
+ "loss": 1.2773,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 4.02,
120
+ "learning_rate": 0.00025537190082644627,
121
+ "loss": 1.1511,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 4.23,
126
+ "learning_rate": 0.00025041322314049583,
127
+ "loss": 1.1912,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 4.23,
132
+ "eval_loss": 1.4607141017913818,
133
+ "eval_runtime": 81.1682,
134
+ "eval_samples_per_second": 7.774,
135
+ "eval_steps_per_second": 0.973,
136
+ "step": 200
137
+ },
138
+ {
139
+ "epoch": 4.44,
140
+ "learning_rate": 0.00024545454545454545,
141
+ "loss": 1.278,
142
+ "step": 210
143
+ },
144
+ {
145
+ "epoch": 4.65,
146
+ "learning_rate": 0.000240495867768595,
147
+ "loss": 1.2741,
148
+ "step": 220
149
+ },
150
+ {
151
+ "epoch": 4.86,
152
+ "learning_rate": 0.00023553719008264463,
153
+ "loss": 1.2626,
154
+ "step": 230
155
+ },
156
+ {
157
+ "epoch": 5.07,
158
+ "learning_rate": 0.0002305785123966942,
159
+ "loss": 1.2519,
160
+ "step": 240
161
+ },
162
+ {
163
+ "epoch": 5.29,
164
+ "learning_rate": 0.00022561983471074378,
165
+ "loss": 1.1882,
166
+ "step": 250
167
+ },
168
+ {
169
+ "epoch": 5.5,
170
+ "learning_rate": 0.00022066115702479338,
171
+ "loss": 1.1867,
172
+ "step": 260
173
+ },
174
+ {
175
+ "epoch": 5.71,
176
+ "learning_rate": 0.00021570247933884297,
177
+ "loss": 1.1278,
178
+ "step": 270
179
+ },
180
+ {
181
+ "epoch": 5.92,
182
+ "learning_rate": 0.00021074380165289253,
183
+ "loss": 1.1719,
184
+ "step": 280
185
+ },
186
+ {
187
+ "epoch": 6.13,
188
+ "learning_rate": 0.00020578512396694215,
189
+ "loss": 1.3097,
190
+ "step": 290
191
+ },
192
+ {
193
+ "epoch": 6.34,
194
+ "learning_rate": 0.0002008264462809917,
195
+ "loss": 1.192,
196
+ "step": 300
197
+ },
198
+ {
199
+ "epoch": 6.55,
200
+ "learning_rate": 0.00019586776859504133,
201
+ "loss": 1.1772,
202
+ "step": 310
203
+ },
204
+ {
205
+ "epoch": 6.77,
206
+ "learning_rate": 0.0001909090909090909,
207
+ "loss": 1.1579,
208
+ "step": 320
209
+ },
210
+ {
211
+ "epoch": 6.98,
212
+ "learning_rate": 0.00018595041322314049,
213
+ "loss": 1.0586,
214
+ "step": 330
215
+ },
216
+ {
217
+ "epoch": 7.19,
218
+ "learning_rate": 0.00018099173553719008,
219
+ "loss": 1.1763,
220
+ "step": 340
221
+ },
222
+ {
223
+ "epoch": 7.4,
224
+ "learning_rate": 0.00017603305785123967,
225
+ "loss": 1.1715,
226
+ "step": 350
227
+ },
228
+ {
229
+ "epoch": 7.61,
230
+ "learning_rate": 0.00017107438016528923,
231
+ "loss": 1.1778,
232
+ "step": 360
233
+ },
234
+ {
235
+ "epoch": 7.82,
236
+ "learning_rate": 0.00016611570247933885,
237
+ "loss": 1.1716,
238
+ "step": 370
239
+ },
240
+ {
241
+ "epoch": 8.03,
242
+ "learning_rate": 0.00016115702479338841,
243
+ "loss": 1.1188,
244
+ "step": 380
245
+ },
246
+ {
247
+ "epoch": 8.25,
248
+ "learning_rate": 0.000156198347107438,
249
+ "loss": 1.0312,
250
+ "step": 390
251
+ },
252
+ {
253
+ "epoch": 8.46,
254
+ "learning_rate": 0.0001512396694214876,
255
+ "loss": 1.0545,
256
+ "step": 400
257
+ },
258
+ {
259
+ "epoch": 8.46,
260
+ "eval_loss": 1.4225733280181885,
261
+ "eval_runtime": 81.2375,
262
+ "eval_samples_per_second": 7.767,
263
+ "eval_steps_per_second": 0.972,
264
+ "step": 400
265
+ },
266
+ {
267
+ "epoch": 8.67,
268
+ "learning_rate": 0.00014628099173553716,
269
+ "loss": 1.1773,
270
+ "step": 410
271
+ },
272
+ {
273
+ "epoch": 8.88,
274
+ "learning_rate": 0.00014132231404958675,
275
+ "loss": 1.1649,
276
+ "step": 420
277
+ },
278
+ {
279
+ "epoch": 9.09,
280
+ "learning_rate": 0.00013636363636363634,
281
+ "loss": 1.1801,
282
+ "step": 430
283
+ },
284
+ {
285
+ "epoch": 9.3,
286
+ "learning_rate": 0.00013140495867768593,
287
+ "loss": 1.0845,
288
+ "step": 440
289
+ },
290
+ {
291
+ "epoch": 9.51,
292
+ "learning_rate": 0.00012644628099173552,
293
+ "loss": 1.0754,
294
+ "step": 450
295
+ },
296
+ {
297
+ "epoch": 9.73,
298
+ "learning_rate": 0.00012148760330578511,
299
+ "loss": 1.0589,
300
+ "step": 460
301
+ },
302
+ {
303
+ "epoch": 9.94,
304
+ "learning_rate": 0.00011652892561983469,
305
+ "loss": 1.0123,
306
+ "step": 470
307
+ },
308
+ {
309
+ "epoch": 10.15,
310
+ "learning_rate": 0.00011157024793388428,
311
+ "loss": 1.1889,
312
+ "step": 480
313
+ },
314
+ {
315
+ "epoch": 10.36,
316
+ "learning_rate": 0.00010661157024793387,
317
+ "loss": 1.0788,
318
+ "step": 490
319
+ },
320
+ {
321
+ "epoch": 10.57,
322
+ "learning_rate": 0.00010165289256198345,
323
+ "loss": 1.0947,
324
+ "step": 500
325
+ },
326
+ {
327
+ "epoch": 10.78,
328
+ "learning_rate": 9.669421487603304e-05,
329
+ "loss": 1.0665,
330
+ "step": 510
331
+ },
332
+ {
333
+ "epoch": 10.99,
334
+ "learning_rate": 9.173553719008263e-05,
335
+ "loss": 0.9061,
336
+ "step": 520
337
+ },
338
+ {
339
+ "epoch": 11.21,
340
+ "learning_rate": 8.677685950413221e-05,
341
+ "loss": 1.0877,
342
+ "step": 530
343
+ },
344
+ {
345
+ "epoch": 11.42,
346
+ "learning_rate": 8.18181818181818e-05,
347
+ "loss": 1.0693,
348
+ "step": 540
349
+ },
350
+ {
351
+ "epoch": 11.63,
352
+ "learning_rate": 7.68595041322314e-05,
353
+ "loss": 1.0784,
354
+ "step": 550
355
+ },
356
+ {
357
+ "epoch": 11.84,
358
+ "learning_rate": 7.190082644628098e-05,
359
+ "loss": 1.0778,
360
+ "step": 560
361
+ },
362
+ {
363
+ "epoch": 12.05,
364
+ "learning_rate": 6.694214876033058e-05,
365
+ "loss": 1.0487,
366
+ "step": 570
367
+ },
368
+ {
369
+ "epoch": 12.26,
370
+ "learning_rate": 6.198347107438017e-05,
371
+ "loss": 0.9808,
372
+ "step": 580
373
+ },
374
+ {
375
+ "epoch": 12.47,
376
+ "learning_rate": 5.702479338842975e-05,
377
+ "loss": 0.9327,
378
+ "step": 590
379
+ },
380
+ {
381
+ "epoch": 12.68,
382
+ "learning_rate": 5.2066115702479335e-05,
383
+ "loss": 1.0115,
384
+ "step": 600
385
+ },
386
+ {
387
+ "epoch": 12.68,
388
+ "eval_loss": 1.4200056791305542,
389
+ "eval_runtime": 81.1908,
390
+ "eval_samples_per_second": 7.772,
391
+ "eval_steps_per_second": 0.973,
392
+ "step": 600
393
+ }
394
+ ],
395
+ "max_steps": 705,
396
+ "num_train_epochs": 15,
397
+ "total_flos": 4.2640048718217216e+17,
398
+ "trial_name": null,
399
+ "trial_params": null
400
+ }
checkpoint-600/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a46f4da27d88d972b2396250eeef005295a27631879666feecc45de066a6d6c0
3
+ size 3963