Hugging Face's logo

klcsp
/
mistral-7b-0.3-gpt4o_100k_classification-k

PEFT
TensorBoard
Safetensors
mistral
alignment-handbook
trl
sft
Generated from Trainer
8-bit precision
bitsandbytes
Model card Files Metrics Community
chansung commited on
Commit
d01ff6c
•
1 Parent(s): 645c953

Model save

Browse files
Files changed (6) hide show
  1. README.md +69 -0
  2. adapter_model.safetensors +1 -1
  3. all_results.json +9 -0
  4. runs/Sep29_07-15-40_82b7fdadb1f8/events.out.tfevents.1727594387.82b7fdadb1f8.455084.0 +2 -2
  5. train_results.json +9 -0
  6. trainer_state.json +428 -0
README.md ADDED
@@ -0,0 +1,69 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: mistralai/Mistral-7B-v0.3
3
+ datasets:
4
+ - generator
5
+ library_name: peft
6
+ license: apache-2.0
7
+ tags:
8
+ - trl
9
+ - sft
10
+ - generated_from_trainer
11
+ model-index:
12
+ - name: mistral-7b-0.3-gpt4o_100k_classification-k
13
+ results: []
14
+ ---
15
+
16
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
17
+ should probably proofread and complete it, then remove this comment. -->
18
+
19
+ # mistral-7b-0.3-gpt4o_100k_classification-k
20
+
21
+ This model is a fine-tuned version of [mistralai/Mistral-7B-v0.3](https://huggingface.co/mistralai/Mistral-7B-v0.3) on the generator dataset.
22
+ It achieves the following results on the evaluation set:
23
+ - Loss: 1.5775
24
+
25
+ ## Model description
26
+
27
+ More information needed
28
+
29
+ ## Intended uses & limitations
30
+
31
+ More information needed
32
+
33
+ ## Training and evaluation data
34
+
35
+ More information needed
36
+
37
+ ## Training procedure
38
+
39
+ ### Training hyperparameters
40
+
41
+ The following hyperparameters were used during training:
42
+ - learning_rate: 0.0002
43
+ - train_batch_size: 2
44
+ - eval_batch_size: 2
45
+ - seed: 42
46
+ - distributed_type: multi-GPU
47
+ - num_devices: 8
48
+ - gradient_accumulation_steps: 2
49
+ - total_train_batch_size: 32
50
+ - total_eval_batch_size: 16
51
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
52
+ - lr_scheduler_type: cosine
53
+ - lr_scheduler_warmup_ratio: 0.1
54
+ - num_epochs: 1
55
+
56
+ ### Training results
57
+
58
+ | Training Loss | Epoch | Step | Validation Loss |
59
+ |:-------------:|:-----:|:----:|:---------------:|
60
+ | 0.8854 | 1.0 | 268 | 1.5775 |
61
+
62
+
63
+ ### Framework versions
64
+
65
+ - PEFT 0.13.1.dev0
66
+ - Transformers 4.45.1
67
+ - Pytorch 2.4.1+cu121
68
+ - Datasets 3.0.1
69
+ - Tokenizers 0.20.0
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ff73f265f8883ead03a7df40798b053fe62971b91a8400b9b1004c0d167dfda8
3
  size 1087481480
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09e5ecff665e26d3bad6ffa70decd8ae5d55b5e205ac04da83c2df58a48c8c69
3
  size 1087481480
all_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "total_flos": 3.750133557582889e+17,
4
+ "train_loss": 0.9775012651486183,
5
+ "train_runtime": 1907.2727,
6
+ "train_samples": 92634,
7
+ "train_samples_per_second": 4.491,
8
+ "train_steps_per_second": 0.141
9
+ }
runs/Sep29_07-15-40_82b7fdadb1f8/events.out.tfevents.1727594387.82b7fdadb1f8.455084.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:08e2c5b42c2317f1e4984ba1446aa699a0e50083adf1b09dda09bd635322fd55
3
- size 14576
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f4f76c3ec11b83829d0bd3502d0eb85bf443a87c7959ed89f05240a713416f8b
3
+ size 17944
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "total_flos": 3.750133557582889e+17,
4
+ "train_loss": 0.9775012651486183,
5
+ "train_runtime": 1907.2727,
6
+ "train_samples": 92634,
7
+ "train_samples_per_second": 4.491,
8
+ "train_steps_per_second": 0.141
9
+ }
trainer_state.json ADDED
@@ -0,0 +1,428 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "eval_steps": 500,
6
+ "global_step": 268,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0037313432835820895,
13
+ "grad_norm": 7.512597560882568,
14
+ "learning_rate": 7.4074074074074075e-06,
15
+ "loss": 2.0542,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.018656716417910446,
20
+ "grad_norm": 7.686585426330566,
21
+ "learning_rate": 3.7037037037037037e-05,
22
+ "loss": 2.0294,
23
+ "step": 5
24
+ },
25
+ {
26
+ "epoch": 0.03731343283582089,
27
+ "grad_norm": 4.031171798706055,
28
+ "learning_rate": 7.407407407407407e-05,
29
+ "loss": 1.7759,
30
+ "step": 10
31
+ },
32
+ {
33
+ "epoch": 0.055970149253731345,
34
+ "grad_norm": 2.5195655822753906,
35
+ "learning_rate": 0.00011111111111111112,
36
+ "loss": 1.4858,
37
+ "step": 15
38
+ },
39
+ {
40
+ "epoch": 0.07462686567164178,
41
+ "grad_norm": 1.1103084087371826,
42
+ "learning_rate": 0.00014814814814814815,
43
+ "loss": 1.3199,
44
+ "step": 20
45
+ },
46
+ {
47
+ "epoch": 0.09328358208955224,
48
+ "grad_norm": 0.894225001335144,
49
+ "learning_rate": 0.0001851851851851852,
50
+ "loss": 1.213,
51
+ "step": 25
52
+ },
53
+ {
54
+ "epoch": 0.11194029850746269,
55
+ "grad_norm": 1.1811108589172363,
56
+ "learning_rate": 0.00019992354201925428,
57
+ "loss": 1.1141,
58
+ "step": 30
59
+ },
60
+ {
61
+ "epoch": 0.13059701492537312,
62
+ "grad_norm": 0.6414879560470581,
63
+ "learning_rate": 0.0001994567221375987,
64
+ "loss": 1.064,
65
+ "step": 35
66
+ },
67
+ {
68
+ "epoch": 0.14925373134328357,
69
+ "grad_norm": 0.7366268634796143,
70
+ "learning_rate": 0.00019856753906964686,
71
+ "loss": 1.0217,
72
+ "step": 40
73
+ },
74
+ {
75
+ "epoch": 0.16791044776119404,
76
+ "grad_norm": 0.6300023198127747,
77
+ "learning_rate": 0.00019725976891203376,
78
+ "loss": 0.9985,
79
+ "step": 45
80
+ },
81
+ {
82
+ "epoch": 0.1865671641791045,
83
+ "grad_norm": 0.5019010305404663,
84
+ "learning_rate": 0.00019553896537655318,
85
+ "loss": 0.9718,
86
+ "step": 50
87
+ },
88
+ {
89
+ "epoch": 0.20522388059701493,
90
+ "grad_norm": 0.5189563035964966,
91
+ "learning_rate": 0.0001934124362051919,
92
+ "loss": 0.9591,
93
+ "step": 55
94
+ },
95
+ {
96
+ "epoch": 0.22388059701492538,
97
+ "grad_norm": 0.5040785074234009,
98
+ "learning_rate": 0.000190889212136318,
99
+ "loss": 0.9711,
100
+ "step": 60
101
+ },
102
+ {
103
+ "epoch": 0.24253731343283583,
104
+ "grad_norm": 0.9470528364181519,
105
+ "learning_rate": 0.0001879800085538147,
106
+ "loss": 0.9512,
107
+ "step": 65
108
+ },
109
+ {
110
+ "epoch": 0.26119402985074625,
111
+ "grad_norm": 0.45922908186912537,
112
+ "learning_rate": 0.00018469717998202462,
113
+ "loss": 0.9433,
114
+ "step": 70
115
+ },
116
+ {
117
+ "epoch": 0.2798507462686567,
118
+ "grad_norm": 0.48557597398757935,
119
+ "learning_rate": 0.00018105466761975109,
120
+ "loss": 0.9228,
121
+ "step": 75
122
+ },
123
+ {
124
+ "epoch": 0.29850746268656714,
125
+ "grad_norm": 0.5197058916091919,
126
+ "learning_rate": 0.00017706794013612364,
127
+ "loss": 0.9366,
128
+ "step": 80
129
+ },
130
+ {
131
+ "epoch": 0.31716417910447764,
132
+ "grad_norm": 0.7142367362976074,
133
+ "learning_rate": 0.00017275392797975032,
134
+ "loss": 0.9464,
135
+ "step": 85
136
+ },
137
+ {
138
+ "epoch": 0.3358208955223881,
139
+ "grad_norm": 0.4473322927951813,
140
+ "learning_rate": 0.0001681309514801265,
141
+ "loss": 0.9397,
142
+ "step": 90
143
+ },
144
+ {
145
+ "epoch": 0.35447761194029853,
146
+ "grad_norm": 0.45314615964889526,
147
+ "learning_rate": 0.00016321864304663173,
148
+ "loss": 0.9241,
149
+ "step": 95
150
+ },
151
+ {
152
+ "epoch": 0.373134328358209,
153
+ "grad_norm": 0.5742659568786621,
154
+ "learning_rate": 0.0001580378637955128,
155
+ "loss": 0.9316,
156
+ "step": 100
157
+ },
158
+ {
159
+ "epoch": 0.3917910447761194,
160
+ "grad_norm": 0.7061174511909485,
161
+ "learning_rate": 0.00015261061495891345,
162
+ "loss": 0.9249,
163
+ "step": 105
164
+ },
165
+ {
166
+ "epoch": 0.41044776119402987,
167
+ "grad_norm": 0.4923759996891022,
168
+ "learning_rate": 0.00014695994445216985,
169
+ "loss": 0.9111,
170
+ "step": 110
171
+ },
172
+ {
173
+ "epoch": 0.4291044776119403,
174
+ "grad_norm": 0.7759695649147034,
175
+ "learning_rate": 0.00014110984899615367,
176
+ "loss": 0.9068,
177
+ "step": 115
178
+ },
179
+ {
180
+ "epoch": 0.44776119402985076,
181
+ "grad_norm": 1.0017133951187134,
182
+ "learning_rate": 0.000135085172210319,
183
+ "loss": 0.9151,
184
+ "step": 120
185
+ },
186
+ {
187
+ "epoch": 0.4664179104477612,
188
+ "grad_norm": 0.564063310623169,
189
+ "learning_rate": 0.00012891149910922267,
190
+ "loss": 0.9197,
191
+ "step": 125
192
+ },
193
+ {
194
+ "epoch": 0.48507462686567165,
195
+ "grad_norm": 0.635895311832428,
196
+ "learning_rate": 0.00012261504745055964,
197
+ "loss": 0.9042,
198
+ "step": 130
199
+ },
200
+ {
201
+ "epoch": 0.503731343283582,
202
+ "grad_norm": 0.5336510539054871,
203
+ "learning_rate": 0.00011622255639612554,
204
+ "loss": 0.8974,
205
+ "step": 135
206
+ },
207
+ {
208
+ "epoch": 0.5223880597014925,
209
+ "grad_norm": 0.5803085565567017,
210
+ "learning_rate": 0.00010976117295853154,
211
+ "loss": 0.9051,
212
+ "step": 140
213
+ },
214
+ {
215
+ "epoch": 0.5410447761194029,
216
+ "grad_norm": 0.6651114821434021,
217
+ "learning_rate": 0.00010325833671589687,
218
+ "loss": 0.9099,
219
+ "step": 145
220
+ },
221
+ {
222
+ "epoch": 0.5597014925373134,
223
+ "grad_norm": 0.6448351740837097,
224
+ "learning_rate": 9.674166328410318e-05,
225
+ "loss": 0.8999,
226
+ "step": 150
227
+ },
228
+ {
229
+ "epoch": 0.5783582089552238,
230
+ "grad_norm": 0.5589825510978699,
231
+ "learning_rate": 9.023882704146848e-05,
232
+ "loss": 0.9023,
233
+ "step": 155
234
+ },
235
+ {
236
+ "epoch": 0.5970149253731343,
237
+ "grad_norm": 0.5227958559989929,
238
+ "learning_rate": 8.377744360387447e-05,
239
+ "loss": 0.8896,
240
+ "step": 160
241
+ },
242
+ {
243
+ "epoch": 0.6156716417910447,
244
+ "grad_norm": 0.5533550977706909,
245
+ "learning_rate": 7.738495254944042e-05,
246
+ "loss": 0.8908,
247
+ "step": 165
248
+ },
249
+ {
250
+ "epoch": 0.6343283582089553,
251
+ "grad_norm": 0.7187372446060181,
252
+ "learning_rate": 7.108850089077735e-05,
253
+ "loss": 0.8977,
254
+ "step": 170
255
+ },
256
+ {
257
+ "epoch": 0.6529850746268657,
258
+ "grad_norm": 0.6411680579185486,
259
+ "learning_rate": 6.491482778968104e-05,
260
+ "loss": 0.8956,
261
+ "step": 175
262
+ },
263
+ {
264
+ "epoch": 0.6716417910447762,
265
+ "grad_norm": 0.6200590133666992,
266
+ "learning_rate": 5.889015100384636e-05,
267
+ "loss": 0.8675,
268
+ "step": 180
269
+ },
270
+ {
271
+ "epoch": 0.6902985074626866,
272
+ "grad_norm": 0.5138359069824219,
273
+ "learning_rate": 5.304005554783015e-05,
274
+ "loss": 0.8836,
275
+ "step": 185
276
+ },
277
+ {
278
+ "epoch": 0.7089552238805971,
279
+ "grad_norm": 0.5824055075645447,
280
+ "learning_rate": 4.738938504108659e-05,
281
+ "loss": 0.8845,
282
+ "step": 190
283
+ },
284
+ {
285
+ "epoch": 0.7276119402985075,
286
+ "grad_norm": 0.5803791880607605,
287
+ "learning_rate": 4.196213620448723e-05,
288
+ "loss": 0.8813,
289
+ "step": 195
290
+ },
291
+ {
292
+ "epoch": 0.746268656716418,
293
+ "grad_norm": 0.5962420701980591,
294
+ "learning_rate": 3.6781356953368284e-05,
295
+ "loss": 0.879,
296
+ "step": 200
297
+ },
298
+ {
299
+ "epoch": 0.7649253731343284,
300
+ "grad_norm": 0.5040367245674133,
301
+ "learning_rate": 3.186904851987351e-05,
302
+ "loss": 0.8835,
303
+ "step": 205
304
+ },
305
+ {
306
+ "epoch": 0.7835820895522388,
307
+ "grad_norm": 0.5313622355461121,
308
+ "learning_rate": 2.724607202024969e-05,
309
+ "loss": 0.8852,
310
+ "step": 210
311
+ },
312
+ {
313
+ "epoch": 0.8022388059701493,
314
+ "grad_norm": 0.9790322780609131,
315
+ "learning_rate": 2.2932059863876365e-05,
316
+ "loss": 0.8931,
317
+ "step": 215
318
+ },
319
+ {
320
+ "epoch": 0.8208955223880597,
321
+ "grad_norm": 0.5015585422515869,
322
+ "learning_rate": 1.8945332380248913e-05,
323
+ "loss": 0.8808,
324
+ "step": 220
325
+ },
326
+ {
327
+ "epoch": 0.8395522388059702,
328
+ "grad_norm": 0.4975396990776062,
329
+ "learning_rate": 1.5302820017975394e-05,
330
+ "loss": 0.8867,
331
+ "step": 225
332
+ },
333
+ {
334
+ "epoch": 0.8582089552238806,
335
+ "grad_norm": 0.5071137547492981,
336
+ "learning_rate": 1.2019991446185309e-05,
337
+ "loss": 0.8919,
338
+ "step": 230
339
+ },
340
+ {
341
+ "epoch": 0.8768656716417911,
342
+ "grad_norm": 0.5157453417778015,
343
+ "learning_rate": 9.110787863682002e-06,
344
+ "loss": 0.8746,
345
+ "step": 235
346
+ },
347
+ {
348
+ "epoch": 0.8955223880597015,
349
+ "grad_norm": 0.5082049369812012,
350
+ "learning_rate": 6.587563794808127e-06,
351
+ "loss": 0.8678,
352
+ "step": 240
353
+ },
354
+ {
355
+ "epoch": 0.914179104477612,
356
+ "grad_norm": 0.5121915936470032,
357
+ "learning_rate": 4.461034623446847e-06,
358
+ "loss": 0.8979,
359
+ "step": 245
360
+ },
361
+ {
362
+ "epoch": 0.9328358208955224,
363
+ "grad_norm": 0.5356196761131287,
364
+ "learning_rate": 2.7402310879662497e-06,
365
+ "loss": 0.8849,
366
+ "step": 250
367
+ },
368
+ {
369
+ "epoch": 0.9514925373134329,
370
+ "grad_norm": 0.5412393808364868,
371
+ "learning_rate": 1.43246093035313e-06,
372
+ "loss": 0.8828,
373
+ "step": 255
374
+ },
375
+ {
376
+ "epoch": 0.9701492537313433,
377
+ "grad_norm": 0.5141103863716125,
378
+ "learning_rate": 5.432778624013257e-07,
379
+ "loss": 0.8721,
380
+ "step": 260
381
+ },
382
+ {
383
+ "epoch": 0.9888059701492538,
384
+ "grad_norm": 0.5164282917976379,
385
+ "learning_rate": 7.645798074572552e-08,
386
+ "loss": 0.8854,
387
+ "step": 265
388
+ },
389
+ {
390
+ "epoch": 1.0,
391
+ "eval_loss": 1.5774962902069092,
392
+ "eval_runtime": 0.9281,
393
+ "eval_samples_per_second": 2.155,
394
+ "eval_steps_per_second": 1.077,
395
+ "step": 268
396
+ },
397
+ {
398
+ "epoch": 1.0,
399
+ "step": 268,
400
+ "total_flos": 3.750133557582889e+17,
401
+ "train_loss": 0.9775012651486183,
402
+ "train_runtime": 1907.2727,
403
+ "train_samples_per_second": 4.491,
404
+ "train_steps_per_second": 0.141
405
+ }
406
+ ],
407
+ "logging_steps": 5,
408
+ "max_steps": 268,
409
+ "num_input_tokens_seen": 0,
410
+ "num_train_epochs": 1,
411
+ "save_steps": 100,
412
+ "stateful_callbacks": {
413
+ "TrainerControl": {
414
+ "args": {
415
+ "should_epoch_stop": false,
416
+ "should_evaluate": false,
417
+ "should_log": false,
418
+ "should_save": true,
419
+ "should_training_stop": true
420
+ },
421
+ "attributes": {}
422
+ }
423
+ },
424
+ "total_flos": 3.750133557582889e+17,
425
+ "train_batch_size": 2,
426
+ "trial_name": null,
427
+ "trial_params": null
428
+ }