Hugging Face's logo

Chinese-Vicuna
/
Chinese-Vicuna-lora-7b-chatv1

Chinese
English
Model card Files Community
lu-vae commited on
Commit
803d440
•
1 Parent(s): c8d8767

Upload 7 files

Browse files
Files changed (7) hide show
  1. train_4800_args/optimizer.pt +3 -0
  2. train_4800_args/rng_state.pth +3 -0
  3. train_4800_args/scaler.pt +3 -0
  4. train_4800_args/scheduler.pt +3 -0
  5. train_4800_args/session.log +27 -0
  6. train_4800_args/trainer_state.json +310 -0
  7. train_4800_args/training_args.bin +3 -0
train_4800_args/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:57bfd4ae66d033f1ae073199b0afa94c79c696f2ad07d5a135646811773a9cf3
3
+ size 160172869
train_4800_args/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d3127aea20eec7db2fcb4d7b7505d6d917ad24359bba61a883f63a4c9e51c4df
3
+ size 14575
train_4800_args/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5bffa96704cd3337e68ccbf054eb33eac1544bc3a84cd3e4d9fbd9ddc4240d2b
3
+ size 557
train_4800_args/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:34f37b96fea3116c16b4bc64ed869707f21766c8cccc2a4aed3dbe4d40b65a1f
3
+ size 627
train_4800_args/session.log ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ args = (
2
+ wandb=False
3
+ prompt_type='chat'
4
+ data_path='instruct.zh.jsonl'
5
+ model_path='llama_7b'
6
+ micro_batch=3
7
+ total_batch=32
8
+ log_steps=100
9
+ eval_steps=0
10
+ save_steps=200
11
+ warmup_ratio=0.01
12
+ test_size=0
13
+ resume_from_checkpoint=None
14
+ ignore_data_skip=False
15
+ )
16
+
17
+ >>> trainable params: 19988480 || all params: 6758404096 || trainable%: 0.2957573965106688
18
+ ***** Running training *****
19
+ Num examples = 51,584
20
+ Num Epochs = 3
21
+ Instantaneous batch size per device = 3
22
+ Total train batch size (w. parallel, distributed & accumulation) = 30
23
+ Gradient Accumulation steps = 10
24
+ Total optimization steps = 4,836
25
+ Number of trainable parameters = 19,988,480
26
+ 4836/4836 [41:14:59<00:00, 30.71s/it]
27
+ {'train_runtime': 148499.2028, 'train_samples_per_second': 0.977, 'train_steps_per_second': 0.033, 'train_loss': 0.7752850797671341, 'epoch': 2.81}
train_4800_args/trainer_state.json ADDED
@@ -0,0 +1,310 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.79150915963943,
5
+ "global_step": 4800,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 6.122448979591835e-06,
13
+ "loss": 1.1171,
14
+ "step": 1
15
+ },
16
+ {
17
+ "epoch": 0.06,
18
+ "learning_rate": 0.00029680384374347184,
19
+ "loss": 1.0109,
20
+ "step": 100
21
+ },
22
+ {
23
+ "epoch": 0.12,
24
+ "learning_rate": 0.000290536870691456,
25
+ "loss": 0.9896,
26
+ "step": 200
27
+ },
28
+ {
29
+ "epoch": 0.17,
30
+ "learning_rate": 0.0002842698976394401,
31
+ "loss": 0.9585,
32
+ "step": 300
33
+ },
34
+ {
35
+ "epoch": 0.23,
36
+ "learning_rate": 0.0002780029245874242,
37
+ "loss": 0.9377,
38
+ "step": 400
39
+ },
40
+ {
41
+ "epoch": 0.29,
42
+ "learning_rate": 0.00027173595153540835,
43
+ "loss": 0.9123,
44
+ "step": 500
45
+ },
46
+ {
47
+ "epoch": 0.35,
48
+ "learning_rate": 0.0002654689784833925,
49
+ "loss": 0.9131,
50
+ "step": 600
51
+ },
52
+ {
53
+ "epoch": 0.41,
54
+ "learning_rate": 0.00025920200543137664,
55
+ "loss": 0.9059,
56
+ "step": 700
57
+ },
58
+ {
59
+ "epoch": 0.47,
60
+ "learning_rate": 0.0002529350323793608,
61
+ "loss": 0.9045,
62
+ "step": 800
63
+ },
64
+ {
65
+ "epoch": 0.52,
66
+ "learning_rate": 0.00024666805932734487,
67
+ "loss": 0.8838,
68
+ "step": 900
69
+ },
70
+ {
71
+ "epoch": 0.58,
72
+ "learning_rate": 0.000240401086275329,
73
+ "loss": 0.8889,
74
+ "step": 1000
75
+ },
76
+ {
77
+ "epoch": 0.64,
78
+ "learning_rate": 0.00023413411322331312,
79
+ "loss": 0.8715,
80
+ "step": 1100
81
+ },
82
+ {
83
+ "epoch": 0.7,
84
+ "learning_rate": 0.00022786714017129724,
85
+ "loss": 0.8569,
86
+ "step": 1200
87
+ },
88
+ {
89
+ "epoch": 0.76,
90
+ "learning_rate": 0.00022160016711928138,
91
+ "loss": 0.8623,
92
+ "step": 1300
93
+ },
94
+ {
95
+ "epoch": 0.81,
96
+ "learning_rate": 0.0002153331940672655,
97
+ "loss": 0.8612,
98
+ "step": 1400
99
+ },
100
+ {
101
+ "epoch": 0.87,
102
+ "learning_rate": 0.0002090662210152496,
103
+ "loss": 0.8409,
104
+ "step": 1500
105
+ },
106
+ {
107
+ "epoch": 0.93,
108
+ "learning_rate": 0.00020279924796323375,
109
+ "loss": 0.846,
110
+ "step": 1600
111
+ },
112
+ {
113
+ "epoch": 0.99,
114
+ "learning_rate": 0.00019653227491121786,
115
+ "loss": 0.8467,
116
+ "step": 1700
117
+ },
118
+ {
119
+ "epoch": 1.05,
120
+ "learning_rate": 0.00019026530185920198,
121
+ "loss": 0.7872,
122
+ "step": 1800
123
+ },
124
+ {
125
+ "epoch": 1.1,
126
+ "learning_rate": 0.0001839983288071861,
127
+ "loss": 0.7711,
128
+ "step": 1900
129
+ },
130
+ {
131
+ "epoch": 1.16,
132
+ "learning_rate": 0.00017773135575517023,
133
+ "loss": 0.7451,
134
+ "step": 2000
135
+ },
136
+ {
137
+ "epoch": 1.22,
138
+ "learning_rate": 0.00017146438270315435,
139
+ "loss": 0.754,
140
+ "step": 2100
141
+ },
142
+ {
143
+ "epoch": 1.28,
144
+ "learning_rate": 0.0001651974096511385,
145
+ "loss": 0.7648,
146
+ "step": 2200
147
+ },
148
+ {
149
+ "epoch": 1.34,
150
+ "learning_rate": 0.00015893043659912263,
151
+ "loss": 0.7585,
152
+ "step": 2300
153
+ },
154
+ {
155
+ "epoch": 1.4,
156
+ "learning_rate": 0.00015266346354710674,
157
+ "loss": 0.7654,
158
+ "step": 2400
159
+ },
160
+ {
161
+ "epoch": 1.45,
162
+ "learning_rate": 0.00014639649049509086,
163
+ "loss": 0.7578,
164
+ "step": 2500
165
+ },
166
+ {
167
+ "epoch": 1.51,
168
+ "learning_rate": 0.000140129517443075,
169
+ "loss": 0.7324,
170
+ "step": 2600
171
+ },
172
+ {
173
+ "epoch": 1.57,
174
+ "learning_rate": 0.00013386254439105911,
175
+ "loss": 0.7439,
176
+ "step": 2700
177
+ },
178
+ {
179
+ "epoch": 1.63,
180
+ "learning_rate": 0.00012759557133904323,
181
+ "loss": 0.7515,
182
+ "step": 2800
183
+ },
184
+ {
185
+ "epoch": 1.69,
186
+ "learning_rate": 0.00012132859828702736,
187
+ "loss": 0.7484,
188
+ "step": 2900
189
+ },
190
+ {
191
+ "epoch": 1.74,
192
+ "learning_rate": 0.00011506162523501148,
193
+ "loss": 0.7466,
194
+ "step": 3000
195
+ },
196
+ {
197
+ "epoch": 1.8,
198
+ "learning_rate": 0.0001087946521829956,
199
+ "loss": 0.7242,
200
+ "step": 3100
201
+ },
202
+ {
203
+ "epoch": 1.86,
204
+ "learning_rate": 0.00010252767913097973,
205
+ "loss": 0.7357,
206
+ "step": 3200
207
+ },
208
+ {
209
+ "epoch": 1.92,
210
+ "learning_rate": 9.626070607896384e-05,
211
+ "loss": 0.7283,
212
+ "step": 3300
213
+ },
214
+ {
215
+ "epoch": 1.98,
216
+ "learning_rate": 8.999373302694798e-05,
217
+ "loss": 0.7375,
218
+ "step": 3400
219
+ },
220
+ {
221
+ "epoch": 2.04,
222
+ "learning_rate": 8.372675997493211e-05,
223
+ "loss": 0.6901,
224
+ "step": 3500
225
+ },
226
+ {
227
+ "epoch": 2.09,
228
+ "learning_rate": 7.745978692291623e-05,
229
+ "loss": 0.6553,
230
+ "step": 3600
231
+ },
232
+ {
233
+ "epoch": 2.15,
234
+ "learning_rate": 7.119281387090035e-05,
235
+ "loss": 0.6572,
236
+ "step": 3700
237
+ },
238
+ {
239
+ "epoch": 2.21,
240
+ "learning_rate": 6.492584081888447e-05,
241
+ "loss": 0.6609,
242
+ "step": 3800
243
+ },
244
+ {
245
+ "epoch": 2.27,
246
+ "learning_rate": 5.8658867766868595e-05,
247
+ "loss": 0.6704,
248
+ "step": 3900
249
+ },
250
+ {
251
+ "epoch": 2.33,
252
+ "learning_rate": 5.239189471485272e-05,
253
+ "loss": 0.6513,
254
+ "step": 4000
255
+ },
256
+ {
257
+ "epoch": 2.38,
258
+ "learning_rate": 4.6124921662836844e-05,
259
+ "loss": 0.6589,
260
+ "step": 4100
261
+ },
262
+ {
263
+ "epoch": 2.44,
264
+ "learning_rate": 3.985794861082097e-05,
265
+ "loss": 0.6755,
266
+ "step": 4200
267
+ },
268
+ {
269
+ "epoch": 2.5,
270
+ "learning_rate": 3.359097555880509e-05,
271
+ "loss": 0.6519,
272
+ "step": 4300
273
+ },
274
+ {
275
+ "epoch": 2.56,
276
+ "learning_rate": 2.7324002506789218e-05,
277
+ "loss": 0.6512,
278
+ "step": 4400
279
+ },
280
+ {
281
+ "epoch": 2.62,
282
+ "learning_rate": 2.1057029454773346e-05,
283
+ "loss": 0.6537,
284
+ "step": 4500
285
+ },
286
+ {
287
+ "epoch": 2.68,
288
+ "learning_rate": 1.4790056402757467e-05,
289
+ "loss": 0.6436,
290
+ "step": 4600
291
+ },
292
+ {
293
+ "epoch": 2.73,
294
+ "learning_rate": 8.523083350741591e-06,
295
+ "loss": 0.6488,
296
+ "step": 4700
297
+ },
298
+ {
299
+ "epoch": 2.79,
300
+ "learning_rate": 2.2561102987257156e-06,
301
+ "loss": 0.6393,
302
+ "step": 4800
303
+ }
304
+ ],
305
+ "max_steps": 4836,
306
+ "num_train_epochs": 3,
307
+ "total_flos": 8.901906445628056e+18,
308
+ "trial_name": null,
309
+ "trial_params": null
310
+ }
train_4800_args/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7beeffb92c704154ce6e6bac1e9bcea32cdad29a4af347eceebbd41b3742ce16
3
+ size 3579