Hugging Face's logo

Suprit
/
Zhongjing-LLaMA-lora

Model card Files Community
Suprit commited on
Commit
9ac6599
•
1 Parent(s): eb3bdc6

Upload 13 files

Browse files
Files changed (13) hide show
  1. zhongjing_7_13/.DS_Store +0 -0
  2. zhongjing_7_13/adapter_config.json +17 -0
  3. zhongjing_7_13/adapter_model.bin +3 -0
  4. zhongjing_7_13/checkpoint-6000/.DS_Store +0 -0
  5. zhongjing_7_13/checkpoint-6000/adapter_config.json +17 -0
  6. zhongjing_7_13/checkpoint-6000/adapter_model.bin +3 -0
  7. zhongjing_7_13/checkpoint-6000/finetuning_args.json +13 -0
  8. zhongjing_7_13/checkpoint-6000/trainer_state.json +184 -0
  9. zhongjing_7_13/checkpoint-6000/training_args.bin +3 -0
  10. zhongjing_7_13/description.txt +1 -0
  11. zhongjing_7_13/finetuning_args.json +13 -0
  12. zhongjing_7_13/trainer_state.json +184 -0
  13. zhongjing_7_13/training_args.bin +3 -0
zhongjing_7_13/.DS_Store ADDED
Binary file (6.15 kB). View file
 
zhongjing_7_13/adapter_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model_name_or_path": "/hy-tmp/Ziya-LLaMA-13B-v1",
3
+ "bias": "none",
4
+ "fan_in_fan_out": false,
5
+ "inference_mode": true,
6
+ "init_lora_weights": true,
7
+ "lora_alpha": 32.0,
8
+ "lora_dropout": 0.1,
9
+ "modules_to_save": null,
10
+ "peft_type": "LORA",
11
+ "r": 16,
12
+ "target_modules": [
13
+ "q_proj",
14
+ "v_proj"
15
+ ],
16
+ "task_type": "CAUSAL_LM"
17
+ }
zhongjing_7_13/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f72614d7700f217837e96d903ae850a4d1e224627c615e23381d6a29b834940
3
+ size 26269837
zhongjing_7_13/checkpoint-6000/.DS_Store ADDED
Binary file (6.15 kB). View file
 
zhongjing_7_13/checkpoint-6000/adapter_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model_name_or_path": "/hy-tmp/Ziya-LLaMA-13B-v1",
3
+ "bias": "none",
4
+ "fan_in_fan_out": false,
5
+ "inference_mode": true,
6
+ "init_lora_weights": true,
7
+ "lora_alpha": 32.0,
8
+ "lora_dropout": 0.1,
9
+ "modules_to_save": null,
10
+ "peft_type": "LORA",
11
+ "r": 16,
12
+ "target_modules": [
13
+ "q_proj",
14
+ "v_proj"
15
+ ],
16
+ "task_type": "CAUSAL_LM"
17
+ }
zhongjing_7_13/checkpoint-6000/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f72614d7700f217837e96d903ae850a4d1e224627c615e23381d6a29b834940
3
+ size 26269837
zhongjing_7_13/checkpoint-6000/finetuning_args.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "finetuning_type": "lora",
3
+ "lora_alpha": 32.0,
4
+ "lora_dropout": 0.1,
5
+ "lora_rank": 16,
6
+ "lora_target": [
7
+ "q_proj",
8
+ "v_proj"
9
+ ],
10
+ "name_module_trainable": "mlp",
11
+ "num_hidden_layers": 32,
12
+ "num_layer_trainable": 3
13
+ }
zhongjing_7_13/checkpoint-6000/trainer_state.json ADDED
@@ -0,0 +1,184 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.6521381139755249,
3
+ "best_model_checkpoint": "/hy-tmp/checkpoints/zhongjing_7-13/checkpoint-6000",
4
+ "epoch": 2.271221728021198,
5
+ "global_step": 6000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.19,
12
+ "learning_rate": 5.944240397524184e-05,
13
+ "loss": 0.7616,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.19,
18
+ "eval_loss": 0.7090576887130737,
19
+ "eval_runtime": 1700.4285,
20
+ "eval_samples_per_second": 20.932,
21
+ "eval_steps_per_second": 0.328,
22
+ "step": 500
23
+ },
24
+ {
25
+ "epoch": 0.38,
26
+ "learning_rate": 5.7731724820150744e-05,
27
+ "loss": 0.6972,
28
+ "step": 1000
29
+ },
30
+ {
31
+ "epoch": 0.38,
32
+ "eval_loss": 0.6910683512687683,
33
+ "eval_runtime": 1700.3181,
34
+ "eval_samples_per_second": 20.933,
35
+ "eval_steps_per_second": 0.328,
36
+ "step": 1000
37
+ },
38
+ {
39
+ "epoch": 0.57,
40
+ "learning_rate": 5.4934584214999246e-05,
41
+ "loss": 0.6845,
42
+ "step": 1500
43
+ },
44
+ {
45
+ "epoch": 0.57,
46
+ "eval_loss": 0.6805794835090637,
47
+ "eval_runtime": 1698.9819,
48
+ "eval_samples_per_second": 20.95,
49
+ "eval_steps_per_second": 0.328,
50
+ "step": 1500
51
+ },
52
+ {
53
+ "epoch": 0.76,
54
+ "learning_rate": 5.116056731749404e-05,
55
+ "loss": 0.6764,
56
+ "step": 2000
57
+ },
58
+ {
59
+ "epoch": 0.76,
60
+ "eval_loss": 0.6737232804298401,
61
+ "eval_runtime": 1702.5596,
62
+ "eval_samples_per_second": 20.906,
63
+ "eval_steps_per_second": 0.327,
64
+ "step": 2000
65
+ },
66
+ {
67
+ "epoch": 0.95,
68
+ "learning_rate": 4.6557530919724635e-05,
69
+ "loss": 0.6701,
70
+ "step": 2500
71
+ },
72
+ {
73
+ "epoch": 0.95,
74
+ "eval_loss": 0.6684596538543701,
75
+ "eval_runtime": 1702.1448,
76
+ "eval_samples_per_second": 20.911,
77
+ "eval_steps_per_second": 0.327,
78
+ "step": 2500
79
+ },
80
+ {
81
+ "epoch": 1.14,
82
+ "learning_rate": 4.131682830461389e-05,
83
+ "loss": 0.663,
84
+ "step": 3000
85
+ },
86
+ {
87
+ "epoch": 1.14,
88
+ "eval_loss": 0.6644229292869568,
89
+ "eval_runtime": 1699.1079,
90
+ "eval_samples_per_second": 20.948,
91
+ "eval_steps_per_second": 0.328,
92
+ "step": 3000
93
+ },
94
+ {
95
+ "epoch": 1.32,
96
+ "learning_rate": 3.5622841602995877e-05,
97
+ "loss": 0.6594,
98
+ "step": 3500
99
+ },
100
+ {
101
+ "epoch": 1.32,
102
+ "eval_loss": 0.661072313785553,
103
+ "eval_runtime": 1696.6207,
104
+ "eval_samples_per_second": 20.979,
105
+ "eval_steps_per_second": 0.328,
106
+ "step": 3500
107
+ },
108
+ {
109
+ "epoch": 1.51,
110
+ "learning_rate": 2.9720460586930557e-05,
111
+ "loss": 0.6561,
112
+ "step": 4000
113
+ },
114
+ {
115
+ "epoch": 1.51,
116
+ "eval_loss": 0.6583240032196045,
117
+ "eval_runtime": 1699.0104,
118
+ "eval_samples_per_second": 20.949,
119
+ "eval_steps_per_second": 0.328,
120
+ "step": 4000
121
+ },
122
+ {
123
+ "epoch": 1.7,
124
+ "learning_rate": 2.381734702787557e-05,
125
+ "loss": 0.6538,
126
+ "step": 4500
127
+ },
128
+ {
129
+ "epoch": 1.7,
130
+ "eval_loss": 0.6560451984405518,
131
+ "eval_runtime": 1701.2915,
132
+ "eval_samples_per_second": 20.921,
133
+ "eval_steps_per_second": 0.327,
134
+ "step": 4500
135
+ },
136
+ {
137
+ "epoch": 1.89,
138
+ "learning_rate": 1.815645475714211e-05,
139
+ "loss": 0.6525,
140
+ "step": 5000
141
+ },
142
+ {
143
+ "epoch": 1.89,
144
+ "eval_loss": 0.654275119304657,
145
+ "eval_runtime": 1698.3722,
146
+ "eval_samples_per_second": 20.957,
147
+ "eval_steps_per_second": 0.328,
148
+ "step": 5000
149
+ },
150
+ {
151
+ "epoch": 2.08,
152
+ "learning_rate": 1.2969355277673462e-05,
153
+ "loss": 0.6488,
154
+ "step": 5500
155
+ },
156
+ {
157
+ "epoch": 2.08,
158
+ "eval_loss": 0.6530821919441223,
159
+ "eval_runtime": 1698.6748,
160
+ "eval_samples_per_second": 20.953,
161
+ "eval_steps_per_second": 0.328,
162
+ "step": 5500
163
+ },
164
+ {
165
+ "epoch": 2.27,
166
+ "learning_rate": 8.438544816619625e-06,
167
+ "loss": 0.6478,
168
+ "step": 6000
169
+ },
170
+ {
171
+ "epoch": 2.27,
172
+ "eval_loss": 0.6521381139755249,
173
+ "eval_runtime": 1697.953,
174
+ "eval_samples_per_second": 20.962,
175
+ "eval_steps_per_second": 0.328,
176
+ "step": 6000
177
+ }
178
+ ],
179
+ "max_steps": 7923,
180
+ "num_train_epochs": 3,
181
+ "total_flos": 9.200694994360966e+19,
182
+ "trial_name": null,
183
+ "trial_params": null
184
+ }
zhongjing_7_13/checkpoint-6000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c4db80be0af32cfab744d9ec8084f5cfa7fa46570ab2bfcbdfdd0cf72d853bd
3
+ size 3348
zhongjing_7_13/description.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ 用于完成医学对话的训练
zhongjing_7_13/finetuning_args.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "finetuning_type": "lora",
3
+ "lora_alpha": 32.0,
4
+ "lora_dropout": 0.1,
5
+ "lora_rank": 16,
6
+ "lora_target": [
7
+ "q_proj",
8
+ "v_proj"
9
+ ],
10
+ "name_module_trainable": "mlp",
11
+ "num_hidden_layers": 32,
12
+ "num_layer_trainable": 3
13
+ }
zhongjing_7_13/trainer_state.json ADDED
@@ -0,0 +1,184 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.6521381139755249,
3
+ "best_model_checkpoint": "/hy-tmp/checkpoints/zhongjing_7-13/checkpoint-6000",
4
+ "epoch": 2.271221728021198,
5
+ "global_step": 6000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.19,
12
+ "learning_rate": 5.944240397524184e-05,
13
+ "loss": 0.7616,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.19,
18
+ "eval_loss": 0.7090576887130737,
19
+ "eval_runtime": 1700.4285,
20
+ "eval_samples_per_second": 20.932,
21
+ "eval_steps_per_second": 0.328,
22
+ "step": 500
23
+ },
24
+ {
25
+ "epoch": 0.38,
26
+ "learning_rate": 5.7731724820150744e-05,
27
+ "loss": 0.6972,
28
+ "step": 1000
29
+ },
30
+ {
31
+ "epoch": 0.38,
32
+ "eval_loss": 0.6910683512687683,
33
+ "eval_runtime": 1700.3181,
34
+ "eval_samples_per_second": 20.933,
35
+ "eval_steps_per_second": 0.328,
36
+ "step": 1000
37
+ },
38
+ {
39
+ "epoch": 0.57,
40
+ "learning_rate": 5.4934584214999246e-05,
41
+ "loss": 0.6845,
42
+ "step": 1500
43
+ },
44
+ {
45
+ "epoch": 0.57,
46
+ "eval_loss": 0.6805794835090637,
47
+ "eval_runtime": 1698.9819,
48
+ "eval_samples_per_second": 20.95,
49
+ "eval_steps_per_second": 0.328,
50
+ "step": 1500
51
+ },
52
+ {
53
+ "epoch": 0.76,
54
+ "learning_rate": 5.116056731749404e-05,
55
+ "loss": 0.6764,
56
+ "step": 2000
57
+ },
58
+ {
59
+ "epoch": 0.76,
60
+ "eval_loss": 0.6737232804298401,
61
+ "eval_runtime": 1702.5596,
62
+ "eval_samples_per_second": 20.906,
63
+ "eval_steps_per_second": 0.327,
64
+ "step": 2000
65
+ },
66
+ {
67
+ "epoch": 0.95,
68
+ "learning_rate": 4.6557530919724635e-05,
69
+ "loss": 0.6701,
70
+ "step": 2500
71
+ },
72
+ {
73
+ "epoch": 0.95,
74
+ "eval_loss": 0.6684596538543701,
75
+ "eval_runtime": 1702.1448,
76
+ "eval_samples_per_second": 20.911,
77
+ "eval_steps_per_second": 0.327,
78
+ "step": 2500
79
+ },
80
+ {
81
+ "epoch": 1.14,
82
+ "learning_rate": 4.131682830461389e-05,
83
+ "loss": 0.663,
84
+ "step": 3000
85
+ },
86
+ {
87
+ "epoch": 1.14,
88
+ "eval_loss": 0.6644229292869568,
89
+ "eval_runtime": 1699.1079,
90
+ "eval_samples_per_second": 20.948,
91
+ "eval_steps_per_second": 0.328,
92
+ "step": 3000
93
+ },
94
+ {
95
+ "epoch": 1.32,
96
+ "learning_rate": 3.5622841602995877e-05,
97
+ "loss": 0.6594,
98
+ "step": 3500
99
+ },
100
+ {
101
+ "epoch": 1.32,
102
+ "eval_loss": 0.661072313785553,
103
+ "eval_runtime": 1696.6207,
104
+ "eval_samples_per_second": 20.979,
105
+ "eval_steps_per_second": 0.328,
106
+ "step": 3500
107
+ },
108
+ {
109
+ "epoch": 1.51,
110
+ "learning_rate": 2.9720460586930557e-05,
111
+ "loss": 0.6561,
112
+ "step": 4000
113
+ },
114
+ {
115
+ "epoch": 1.51,
116
+ "eval_loss": 0.6583240032196045,
117
+ "eval_runtime": 1699.0104,
118
+ "eval_samples_per_second": 20.949,
119
+ "eval_steps_per_second": 0.328,
120
+ "step": 4000
121
+ },
122
+ {
123
+ "epoch": 1.7,
124
+ "learning_rate": 2.381734702787557e-05,
125
+ "loss": 0.6538,
126
+ "step": 4500
127
+ },
128
+ {
129
+ "epoch": 1.7,
130
+ "eval_loss": 0.6560451984405518,
131
+ "eval_runtime": 1701.2915,
132
+ "eval_samples_per_second": 20.921,
133
+ "eval_steps_per_second": 0.327,
134
+ "step": 4500
135
+ },
136
+ {
137
+ "epoch": 1.89,
138
+ "learning_rate": 1.815645475714211e-05,
139
+ "loss": 0.6525,
140
+ "step": 5000
141
+ },
142
+ {
143
+ "epoch": 1.89,
144
+ "eval_loss": 0.654275119304657,
145
+ "eval_runtime": 1698.3722,
146
+ "eval_samples_per_second": 20.957,
147
+ "eval_steps_per_second": 0.328,
148
+ "step": 5000
149
+ },
150
+ {
151
+ "epoch": 2.08,
152
+ "learning_rate": 1.2969355277673462e-05,
153
+ "loss": 0.6488,
154
+ "step": 5500
155
+ },
156
+ {
157
+ "epoch": 2.08,
158
+ "eval_loss": 0.6530821919441223,
159
+ "eval_runtime": 1698.6748,
160
+ "eval_samples_per_second": 20.953,
161
+ "eval_steps_per_second": 0.328,
162
+ "step": 5500
163
+ },
164
+ {
165
+ "epoch": 2.27,
166
+ "learning_rate": 8.438544816619625e-06,
167
+ "loss": 0.6478,
168
+ "step": 6000
169
+ },
170
+ {
171
+ "epoch": 2.27,
172
+ "eval_loss": 0.6521381139755249,
173
+ "eval_runtime": 1697.953,
174
+ "eval_samples_per_second": 20.962,
175
+ "eval_steps_per_second": 0.328,
176
+ "step": 6000
177
+ }
178
+ ],
179
+ "max_steps": 7923,
180
+ "num_train_epochs": 3,
181
+ "total_flos": 9.200694994360966e+19,
182
+ "trial_name": null,
183
+ "trial_params": null
184
+ }
zhongjing_7_13/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c4db80be0af32cfab744d9ec8084f5cfa7fa46570ab2bfcbdfdd0cf72d853bd
3
+ size 3348