GaetanMichelet commited on
Commit
9d7c150
1 Parent(s): d7b304b

Model save

Browse files
README.md ADDED
@@ -0,0 +1,76 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: mistralai/Mistral-7B-Instruct-v0.3
3
+ library_name: peft
4
+ license: apache-2.0
5
+ tags:
6
+ - trl
7
+ - sft
8
+ - generated_from_trainer
9
+ model-index:
10
+ - name: Mistral-7B_task-3_120-samples_config-1
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # Mistral-7B_task-3_120-samples_config-1
18
+
19
+ This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.3](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3) on the None dataset.
20
+ It achieves the following results on the evaluation set:
21
+ - Loss: 0.7278
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 0.0001
41
+ - train_batch_size: 1
42
+ - eval_batch_size: 1
43
+ - seed: 42
44
+ - distributed_type: multi-GPU
45
+ - gradient_accumulation_steps: 8
46
+ - total_train_batch_size: 8
47
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
48
+ - lr_scheduler_type: cosine
49
+ - lr_scheduler_warmup_ratio: 0.1
50
+ - num_epochs: 50
51
+
52
+ ### Training results
53
+
54
+ | Training Loss | Epoch | Step | Validation Loss |
55
+ |:-------------:|:-----:|:----:|:---------------:|
56
+ | 2.2415 | 1.0 | 11 | 1.8293 |
57
+ | 0.3793 | 2.0 | 22 | 0.5602 |
58
+ | 0.2382 | 3.0 | 33 | 0.4231 |
59
+ | 0.2918 | 4.0 | 44 | 0.3916 |
60
+ | 0.1887 | 5.0 | 55 | 0.3913 |
61
+ | 0.0868 | 6.0 | 66 | 0.4560 |
62
+ | 0.0333 | 7.0 | 77 | 0.5414 |
63
+ | 0.0184 | 8.0 | 88 | 0.5703 |
64
+ | 0.0298 | 9.0 | 99 | 0.5941 |
65
+ | 0.0026 | 10.0 | 110 | 0.6328 |
66
+ | 0.0007 | 11.0 | 121 | 0.6887 |
67
+ | 0.0086 | 12.0 | 132 | 0.7278 |
68
+
69
+
70
+ ### Framework versions
71
+
72
+ - PEFT 0.12.0
73
+ - Transformers 4.44.0
74
+ - Pytorch 2.1.2+cu121
75
+ - Datasets 2.20.0
76
+ - Tokenizers 0.19.1
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e72392565bee0a67d796b48342d75d52e84c9d9c4f3ba04076293cdb3ac50420
3
  size 167832240
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:07a2aca854f48abefca6d1130a67ebd0028bd07a05f46fbd59e21e3595fb8e41
3
  size 167832240
all_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 12.0,
3
+ "total_flos": 1.0256775620486758e+17,
4
+ "train_loss": 0.4348593218530754,
5
+ "train_runtime": 3817.9089,
6
+ "train_samples": 88,
7
+ "train_samples_per_second": 1.152,
8
+ "train_steps_per_second": 0.144
9
+ }
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 12.0,
3
+ "total_flos": 1.0256775620486758e+17,
4
+ "train_loss": 0.4348593218530754,
5
+ "train_runtime": 3817.9089,
6
+ "train_samples": 88,
7
+ "train_samples_per_second": 1.152,
8
+ "train_steps_per_second": 0.144
9
+ }
trainer_state.json ADDED
@@ -0,0 +1,616 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.39128661155700684,
3
+ "best_model_checkpoint": "data/Mistral-7B_task-3_120-samples_config-1/checkpoint-55",
4
+ "epoch": 12.0,
5
+ "eval_steps": 500,
6
+ "global_step": 132,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.09090909090909091,
13
+ "grad_norm": 26.581195831298828,
14
+ "learning_rate": 1.818181818181818e-06,
15
+ "loss": 5.0961,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.18181818181818182,
20
+ "grad_norm": 17.186176300048828,
21
+ "learning_rate": 3.636363636363636e-06,
22
+ "loss": 3.7834,
23
+ "step": 2
24
+ },
25
+ {
26
+ "epoch": 0.36363636363636365,
27
+ "grad_norm": 16.22742462158203,
28
+ "learning_rate": 7.272727272727272e-06,
29
+ "loss": 3.7902,
30
+ "step": 4
31
+ },
32
+ {
33
+ "epoch": 0.5454545454545454,
34
+ "grad_norm": 13.555403709411621,
35
+ "learning_rate": 1.0909090909090909e-05,
36
+ "loss": 3.3208,
37
+ "step": 6
38
+ },
39
+ {
40
+ "epoch": 0.7272727272727273,
41
+ "grad_norm": 19.296531677246094,
42
+ "learning_rate": 1.4545454545454545e-05,
43
+ "loss": 2.8761,
44
+ "step": 8
45
+ },
46
+ {
47
+ "epoch": 0.9090909090909091,
48
+ "grad_norm": 10.272661209106445,
49
+ "learning_rate": 1.8181818181818182e-05,
50
+ "loss": 2.2415,
51
+ "step": 10
52
+ },
53
+ {
54
+ "epoch": 1.0,
55
+ "eval_loss": 1.829300880432129,
56
+ "eval_runtime": 25.7663,
57
+ "eval_samples_per_second": 0.931,
58
+ "eval_steps_per_second": 0.931,
59
+ "step": 11
60
+ },
61
+ {
62
+ "epoch": 1.0909090909090908,
63
+ "grad_norm": 7.439228057861328,
64
+ "learning_rate": 2.1818181818181818e-05,
65
+ "loss": 1.8772,
66
+ "step": 12
67
+ },
68
+ {
69
+ "epoch": 1.2727272727272727,
70
+ "grad_norm": 5.770608901977539,
71
+ "learning_rate": 2.5454545454545454e-05,
72
+ "loss": 1.4832,
73
+ "step": 14
74
+ },
75
+ {
76
+ "epoch": 1.4545454545454546,
77
+ "grad_norm": 4.287290573120117,
78
+ "learning_rate": 2.909090909090909e-05,
79
+ "loss": 1.0633,
80
+ "step": 16
81
+ },
82
+ {
83
+ "epoch": 1.6363636363636362,
84
+ "grad_norm": 2.3973379135131836,
85
+ "learning_rate": 3.272727272727273e-05,
86
+ "loss": 0.8615,
87
+ "step": 18
88
+ },
89
+ {
90
+ "epoch": 1.8181818181818183,
91
+ "grad_norm": 2.0681209564208984,
92
+ "learning_rate": 3.6363636363636364e-05,
93
+ "loss": 0.6223,
94
+ "step": 20
95
+ },
96
+ {
97
+ "epoch": 2.0,
98
+ "grad_norm": 1.4477177858352661,
99
+ "learning_rate": 4e-05,
100
+ "loss": 0.3793,
101
+ "step": 22
102
+ },
103
+ {
104
+ "epoch": 2.0,
105
+ "eval_loss": 0.5601645112037659,
106
+ "eval_runtime": 25.7673,
107
+ "eval_samples_per_second": 0.931,
108
+ "eval_steps_per_second": 0.931,
109
+ "step": 22
110
+ },
111
+ {
112
+ "epoch": 2.1818181818181817,
113
+ "grad_norm": 1.0610381364822388,
114
+ "learning_rate": 4.3636363636363636e-05,
115
+ "loss": 0.4043,
116
+ "step": 24
117
+ },
118
+ {
119
+ "epoch": 2.3636363636363638,
120
+ "grad_norm": 1.2741479873657227,
121
+ "learning_rate": 4.7272727272727275e-05,
122
+ "loss": 0.5729,
123
+ "step": 26
124
+ },
125
+ {
126
+ "epoch": 2.5454545454545454,
127
+ "grad_norm": 0.8593302369117737,
128
+ "learning_rate": 5.090909090909091e-05,
129
+ "loss": 0.3782,
130
+ "step": 28
131
+ },
132
+ {
133
+ "epoch": 2.7272727272727275,
134
+ "grad_norm": 1.1046785116195679,
135
+ "learning_rate": 5.4545454545454546e-05,
136
+ "loss": 0.3238,
137
+ "step": 30
138
+ },
139
+ {
140
+ "epoch": 2.909090909090909,
141
+ "grad_norm": 0.7316509485244751,
142
+ "learning_rate": 5.818181818181818e-05,
143
+ "loss": 0.2382,
144
+ "step": 32
145
+ },
146
+ {
147
+ "epoch": 3.0,
148
+ "eval_loss": 0.42310047149658203,
149
+ "eval_runtime": 25.7592,
150
+ "eval_samples_per_second": 0.932,
151
+ "eval_steps_per_second": 0.932,
152
+ "step": 33
153
+ },
154
+ {
155
+ "epoch": 3.090909090909091,
156
+ "grad_norm": 1.131074070930481,
157
+ "learning_rate": 6.181818181818182e-05,
158
+ "loss": 0.4555,
159
+ "step": 34
160
+ },
161
+ {
162
+ "epoch": 3.2727272727272725,
163
+ "grad_norm": 0.38236701488494873,
164
+ "learning_rate": 6.545454545454546e-05,
165
+ "loss": 0.1055,
166
+ "step": 36
167
+ },
168
+ {
169
+ "epoch": 3.4545454545454546,
170
+ "grad_norm": 4.889406681060791,
171
+ "learning_rate": 6.90909090909091e-05,
172
+ "loss": 0.2955,
173
+ "step": 38
174
+ },
175
+ {
176
+ "epoch": 3.6363636363636362,
177
+ "grad_norm": 0.8969003558158875,
178
+ "learning_rate": 7.272727272727273e-05,
179
+ "loss": 0.281,
180
+ "step": 40
181
+ },
182
+ {
183
+ "epoch": 3.8181818181818183,
184
+ "grad_norm": 1.034072995185852,
185
+ "learning_rate": 7.636363636363637e-05,
186
+ "loss": 0.3261,
187
+ "step": 42
188
+ },
189
+ {
190
+ "epoch": 4.0,
191
+ "grad_norm": 2.0129034519195557,
192
+ "learning_rate": 8e-05,
193
+ "loss": 0.2918,
194
+ "step": 44
195
+ },
196
+ {
197
+ "epoch": 4.0,
198
+ "eval_loss": 0.39155831933021545,
199
+ "eval_runtime": 25.7607,
200
+ "eval_samples_per_second": 0.932,
201
+ "eval_steps_per_second": 0.932,
202
+ "step": 44
203
+ },
204
+ {
205
+ "epoch": 4.181818181818182,
206
+ "grad_norm": 0.6423981785774231,
207
+ "learning_rate": 8.363636363636364e-05,
208
+ "loss": 0.1505,
209
+ "step": 46
210
+ },
211
+ {
212
+ "epoch": 4.363636363636363,
213
+ "grad_norm": 1.1609551906585693,
214
+ "learning_rate": 8.727272727272727e-05,
215
+ "loss": 0.2499,
216
+ "step": 48
217
+ },
218
+ {
219
+ "epoch": 4.545454545454545,
220
+ "grad_norm": 0.9071617722511292,
221
+ "learning_rate": 9.090909090909092e-05,
222
+ "loss": 0.1717,
223
+ "step": 50
224
+ },
225
+ {
226
+ "epoch": 4.7272727272727275,
227
+ "grad_norm": 2.986572027206421,
228
+ "learning_rate": 9.454545454545455e-05,
229
+ "loss": 0.1974,
230
+ "step": 52
231
+ },
232
+ {
233
+ "epoch": 4.909090909090909,
234
+ "grad_norm": 1.1126000881195068,
235
+ "learning_rate": 9.818181818181818e-05,
236
+ "loss": 0.1887,
237
+ "step": 54
238
+ },
239
+ {
240
+ "epoch": 5.0,
241
+ "eval_loss": 0.39128661155700684,
242
+ "eval_runtime": 25.7604,
243
+ "eval_samples_per_second": 0.932,
244
+ "eval_steps_per_second": 0.932,
245
+ "step": 55
246
+ },
247
+ {
248
+ "epoch": 5.090909090909091,
249
+ "grad_norm": 0.8015605807304382,
250
+ "learning_rate": 9.999899300364532e-05,
251
+ "loss": 0.1774,
252
+ "step": 56
253
+ },
254
+ {
255
+ "epoch": 5.2727272727272725,
256
+ "grad_norm": 0.818981409072876,
257
+ "learning_rate": 9.99909372761763e-05,
258
+ "loss": 0.1105,
259
+ "step": 58
260
+ },
261
+ {
262
+ "epoch": 5.454545454545454,
263
+ "grad_norm": 0.3991180658340454,
264
+ "learning_rate": 9.997482711915927e-05,
265
+ "loss": 0.0658,
266
+ "step": 60
267
+ },
268
+ {
269
+ "epoch": 5.636363636363637,
270
+ "grad_norm": 1.6436625719070435,
271
+ "learning_rate": 9.99506651282272e-05,
272
+ "loss": 0.1566,
273
+ "step": 62
274
+ },
275
+ {
276
+ "epoch": 5.818181818181818,
277
+ "grad_norm": 1.0331753492355347,
278
+ "learning_rate": 9.991845519630678e-05,
279
+ "loss": 0.0605,
280
+ "step": 64
281
+ },
282
+ {
283
+ "epoch": 6.0,
284
+ "grad_norm": 0.9555497169494629,
285
+ "learning_rate": 9.987820251299122e-05,
286
+ "loss": 0.0868,
287
+ "step": 66
288
+ },
289
+ {
290
+ "epoch": 6.0,
291
+ "eval_loss": 0.45599040389060974,
292
+ "eval_runtime": 25.7595,
293
+ "eval_samples_per_second": 0.932,
294
+ "eval_steps_per_second": 0.932,
295
+ "step": 66
296
+ },
297
+ {
298
+ "epoch": 6.181818181818182,
299
+ "grad_norm": 0.46245571970939636,
300
+ "learning_rate": 9.982991356370404e-05,
301
+ "loss": 0.0335,
302
+ "step": 68
303
+ },
304
+ {
305
+ "epoch": 6.363636363636363,
306
+ "grad_norm": 0.4522000253200531,
307
+ "learning_rate": 9.977359612865423e-05,
308
+ "loss": 0.0224,
309
+ "step": 70
310
+ },
311
+ {
312
+ "epoch": 6.545454545454545,
313
+ "grad_norm": 1.2461261749267578,
314
+ "learning_rate": 9.970925928158274e-05,
315
+ "loss": 0.0728,
316
+ "step": 72
317
+ },
318
+ {
319
+ "epoch": 6.7272727272727275,
320
+ "grad_norm": 0.866970419883728,
321
+ "learning_rate": 9.963691338830044e-05,
322
+ "loss": 0.0479,
323
+ "step": 74
324
+ },
325
+ {
326
+ "epoch": 6.909090909090909,
327
+ "grad_norm": 1.0041601657867432,
328
+ "learning_rate": 9.955657010501806e-05,
329
+ "loss": 0.0333,
330
+ "step": 76
331
+ },
332
+ {
333
+ "epoch": 7.0,
334
+ "eval_loss": 0.5413926243782043,
335
+ "eval_runtime": 25.7595,
336
+ "eval_samples_per_second": 0.932,
337
+ "eval_steps_per_second": 0.932,
338
+ "step": 77
339
+ },
340
+ {
341
+ "epoch": 7.090909090909091,
342
+ "grad_norm": 0.7442712783813477,
343
+ "learning_rate": 9.946824237646824e-05,
344
+ "loss": 0.0094,
345
+ "step": 78
346
+ },
347
+ {
348
+ "epoch": 7.2727272727272725,
349
+ "grad_norm": 1.4385738372802734,
350
+ "learning_rate": 9.937194443381972e-05,
351
+ "loss": 0.023,
352
+ "step": 80
353
+ },
354
+ {
355
+ "epoch": 7.454545454545454,
356
+ "grad_norm": 0.7076303958892822,
357
+ "learning_rate": 9.926769179238466e-05,
358
+ "loss": 0.0098,
359
+ "step": 82
360
+ },
361
+ {
362
+ "epoch": 7.636363636363637,
363
+ "grad_norm": 0.49934569001197815,
364
+ "learning_rate": 9.915550124911866e-05,
365
+ "loss": 0.018,
366
+ "step": 84
367
+ },
368
+ {
369
+ "epoch": 7.818181818181818,
370
+ "grad_norm": 0.6519784331321716,
371
+ "learning_rate": 9.903539087991462e-05,
372
+ "loss": 0.032,
373
+ "step": 86
374
+ },
375
+ {
376
+ "epoch": 8.0,
377
+ "grad_norm": 1.025206446647644,
378
+ "learning_rate": 9.890738003669029e-05,
379
+ "loss": 0.0184,
380
+ "step": 88
381
+ },
382
+ {
383
+ "epoch": 8.0,
384
+ "eval_loss": 0.5703032612800598,
385
+ "eval_runtime": 25.7606,
386
+ "eval_samples_per_second": 0.932,
387
+ "eval_steps_per_second": 0.932,
388
+ "step": 88
389
+ },
390
+ {
391
+ "epoch": 8.181818181818182,
392
+ "grad_norm": 0.32475417852401733,
393
+ "learning_rate": 9.877148934427037e-05,
394
+ "loss": 0.0117,
395
+ "step": 90
396
+ },
397
+ {
398
+ "epoch": 8.363636363636363,
399
+ "grad_norm": 0.3134045898914337,
400
+ "learning_rate": 9.862774069706346e-05,
401
+ "loss": 0.0058,
402
+ "step": 92
403
+ },
404
+ {
405
+ "epoch": 8.545454545454545,
406
+ "grad_norm": 0.2756688892841339,
407
+ "learning_rate": 9.847615725553456e-05,
408
+ "loss": 0.0025,
409
+ "step": 94
410
+ },
411
+ {
412
+ "epoch": 8.727272727272727,
413
+ "grad_norm": 0.28128984570503235,
414
+ "learning_rate": 9.831676344247342e-05,
415
+ "loss": 0.0189,
416
+ "step": 96
417
+ },
418
+ {
419
+ "epoch": 8.909090909090908,
420
+ "grad_norm": 1.5065057277679443,
421
+ "learning_rate": 9.814958493905963e-05,
422
+ "loss": 0.0298,
423
+ "step": 98
424
+ },
425
+ {
426
+ "epoch": 9.0,
427
+ "eval_loss": 0.5940575003623962,
428
+ "eval_runtime": 25.767,
429
+ "eval_samples_per_second": 0.931,
430
+ "eval_steps_per_second": 0.931,
431
+ "step": 99
432
+ },
433
+ {
434
+ "epoch": 9.090909090909092,
435
+ "grad_norm": 0.26251956820487976,
436
+ "learning_rate": 9.797464868072488e-05,
437
+ "loss": 0.0124,
438
+ "step": 100
439
+ },
440
+ {
441
+ "epoch": 9.272727272727273,
442
+ "grad_norm": 0.011677834205329418,
443
+ "learning_rate": 9.779198285281325e-05,
444
+ "loss": 0.0009,
445
+ "step": 102
446
+ },
447
+ {
448
+ "epoch": 9.454545454545455,
449
+ "grad_norm": 0.3504819869995117,
450
+ "learning_rate": 9.760161688604008e-05,
451
+ "loss": 0.01,
452
+ "step": 104
453
+ },
454
+ {
455
+ "epoch": 9.636363636363637,
456
+ "grad_norm": 0.32781001925468445,
457
+ "learning_rate": 9.740358145174998e-05,
458
+ "loss": 0.0029,
459
+ "step": 106
460
+ },
461
+ {
462
+ "epoch": 9.818181818181818,
463
+ "grad_norm": 1.1787352561950684,
464
+ "learning_rate": 9.719790845697533e-05,
465
+ "loss": 0.0088,
466
+ "step": 108
467
+ },
468
+ {
469
+ "epoch": 10.0,
470
+ "grad_norm": 0.2411821335554123,
471
+ "learning_rate": 9.698463103929542e-05,
472
+ "loss": 0.0026,
473
+ "step": 110
474
+ },
475
+ {
476
+ "epoch": 10.0,
477
+ "eval_loss": 0.632849395275116,
478
+ "eval_runtime": 25.7634,
479
+ "eval_samples_per_second": 0.932,
480
+ "eval_steps_per_second": 0.932,
481
+ "step": 110
482
+ },
483
+ {
484
+ "epoch": 10.181818181818182,
485
+ "grad_norm": 0.5745018124580383,
486
+ "learning_rate": 9.676378356149734e-05,
487
+ "loss": 0.0022,
488
+ "step": 112
489
+ },
490
+ {
491
+ "epoch": 10.363636363636363,
492
+ "grad_norm": 0.12452679872512817,
493
+ "learning_rate": 9.653540160603956e-05,
494
+ "loss": 0.002,
495
+ "step": 114
496
+ },
497
+ {
498
+ "epoch": 10.545454545454545,
499
+ "grad_norm": 0.06492336839437485,
500
+ "learning_rate": 9.629952196931901e-05,
501
+ "loss": 0.0015,
502
+ "step": 116
503
+ },
504
+ {
505
+ "epoch": 10.727272727272727,
506
+ "grad_norm": 0.6480187773704529,
507
+ "learning_rate": 9.60561826557425e-05,
508
+ "loss": 0.0057,
509
+ "step": 118
510
+ },
511
+ {
512
+ "epoch": 10.909090909090908,
513
+ "grad_norm": 0.02387073077261448,
514
+ "learning_rate": 9.580542287160348e-05,
515
+ "loss": 0.0007,
516
+ "step": 120
517
+ },
518
+ {
519
+ "epoch": 11.0,
520
+ "eval_loss": 0.6887062191963196,
521
+ "eval_runtime": 25.7643,
522
+ "eval_samples_per_second": 0.932,
523
+ "eval_steps_per_second": 0.932,
524
+ "step": 121
525
+ },
526
+ {
527
+ "epoch": 11.090909090909092,
528
+ "grad_norm": 0.021298233419656754,
529
+ "learning_rate": 9.554728301876526e-05,
530
+ "loss": 0.0037,
531
+ "step": 122
532
+ },
533
+ {
534
+ "epoch": 11.272727272727273,
535
+ "grad_norm": 0.160748690366745,
536
+ "learning_rate": 9.528180468815155e-05,
537
+ "loss": 0.0009,
538
+ "step": 124
539
+ },
540
+ {
541
+ "epoch": 11.454545454545455,
542
+ "grad_norm": 0.018241755664348602,
543
+ "learning_rate": 9.50090306530454e-05,
544
+ "loss": 0.0009,
545
+ "step": 126
546
+ },
547
+ {
548
+ "epoch": 11.636363636363637,
549
+ "grad_norm": 0.5487923622131348,
550
+ "learning_rate": 9.472900486219769e-05,
551
+ "loss": 0.0031,
552
+ "step": 128
553
+ },
554
+ {
555
+ "epoch": 11.818181818181818,
556
+ "grad_norm": 0.021489502862095833,
557
+ "learning_rate": 9.444177243274618e-05,
558
+ "loss": 0.0008,
559
+ "step": 130
560
+ },
561
+ {
562
+ "epoch": 12.0,
563
+ "grad_norm": 0.4083148241043091,
564
+ "learning_rate": 9.414737964294636e-05,
565
+ "loss": 0.0086,
566
+ "step": 132
567
+ },
568
+ {
569
+ "epoch": 12.0,
570
+ "eval_loss": 0.7278346419334412,
571
+ "eval_runtime": 25.7635,
572
+ "eval_samples_per_second": 0.932,
573
+ "eval_steps_per_second": 0.932,
574
+ "step": 132
575
+ },
576
+ {
577
+ "epoch": 12.0,
578
+ "step": 132,
579
+ "total_flos": 1.0256775620486758e+17,
580
+ "train_loss": 0.4348593218530754,
581
+ "train_runtime": 3817.9089,
582
+ "train_samples_per_second": 1.152,
583
+ "train_steps_per_second": 0.144
584
+ }
585
+ ],
586
+ "logging_steps": 2,
587
+ "max_steps": 550,
588
+ "num_input_tokens_seen": 0,
589
+ "num_train_epochs": 50,
590
+ "save_steps": 25,
591
+ "stateful_callbacks": {
592
+ "EarlyStoppingCallback": {
593
+ "args": {
594
+ "early_stopping_patience": 7,
595
+ "early_stopping_threshold": 0.0
596
+ },
597
+ "attributes": {
598
+ "early_stopping_patience_counter": 0
599
+ }
600
+ },
601
+ "TrainerControl": {
602
+ "args": {
603
+ "should_epoch_stop": false,
604
+ "should_evaluate": false,
605
+ "should_log": false,
606
+ "should_save": true,
607
+ "should_training_stop": true
608
+ },
609
+ "attributes": {}
610
+ }
611
+ },
612
+ "total_flos": 1.0256775620486758e+17,
613
+ "train_batch_size": 1,
614
+ "trial_name": null,
615
+ "trial_params": null
616
+ }