sedrickkeh commited on
Commit
b9c5ddc
·
verified ·
1 Parent(s): 85734ba

End of training

Browse files
README.md CHANGED
@@ -4,6 +4,7 @@ license: apache-2.0
4
  base_model: mistralai/Mistral-7B-v0.1
5
  tags:
6
  - llama-factory
 
7
  - generated_from_trainer
8
  model-index:
9
  - name: hp_ablations_mistral_scheduler_cosine_warmup0.05_minlr1e-7_dcftv1.2
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  # hp_ablations_mistral_scheduler_cosine_warmup0.05_minlr1e-7_dcftv1.2
17
 
18
- This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on an unknown dataset.
19
  It achieves the following results on the evaluation set:
20
  - Loss: 0.0734
21
 
 
4
  base_model: mistralai/Mistral-7B-v0.1
5
  tags:
6
  - llama-factory
7
+ - full
8
  - generated_from_trainer
9
  model-index:
10
  - name: hp_ablations_mistral_scheduler_cosine_warmup0.05_minlr1e-7_dcftv1.2
 
16
 
17
  # hp_ablations_mistral_scheduler_cosine_warmup0.05_minlr1e-7_dcftv1.2
18
 
19
+ This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the mlfoundations-dev/oh-dcft-v1.2_no-curation_gpt-4o-mini dataset.
20
  It achieves the following results on the evaluation set:
21
  - Loss: 0.0734
22
 
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.99695843190267,
3
+ "eval_loss": 0.07343952357769012,
4
+ "eval_runtime": 380.6676,
5
+ "eval_samples_per_second": 26.178,
6
+ "eval_steps_per_second": 0.41,
7
+ "total_flos": 1854056851046400.0,
8
+ "train_loss": 0.4996785878489011,
9
+ "train_runtime": 63355.7173,
10
+ "train_samples_per_second": 8.965,
11
+ "train_steps_per_second": 0.017
12
+ }
eval_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.99695843190267,
3
+ "eval_loss": 0.07343952357769012,
4
+ "eval_runtime": 380.6676,
5
+ "eval_samples_per_second": 26.178,
6
+ "eval_steps_per_second": 0.41
7
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.99695843190267,
3
+ "total_flos": 1854056851046400.0,
4
+ "train_loss": 0.4996785878489011,
5
+ "train_runtime": 63355.7173,
6
+ "train_samples_per_second": 8.965,
7
+ "train_steps_per_second": 0.017
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,836 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.99695843190267,
5
+ "eval_steps": 500,
6
+ "global_step": 1107,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.027036160865157147,
13
+ "grad_norm": 5.9832664062841445,
14
+ "learning_rate": 8.928571428571429e-07,
15
+ "loss": 0.8856,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.054072321730314295,
20
+ "grad_norm": 2.422574640898435,
21
+ "learning_rate": 1.7857142857142859e-06,
22
+ "loss": 0.7436,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.08110848259547145,
27
+ "grad_norm": 2.306483898812809,
28
+ "learning_rate": 2.6785714285714285e-06,
29
+ "loss": 0.6792,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.10814464346062859,
34
+ "grad_norm": 1.737436000738722,
35
+ "learning_rate": 3.5714285714285718e-06,
36
+ "loss": 0.655,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.13518080432578575,
41
+ "grad_norm": 1.823323318308892,
42
+ "learning_rate": 4.464285714285715e-06,
43
+ "loss": 0.6334,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.1622169651909429,
48
+ "grad_norm": 3.6157460204989493,
49
+ "learning_rate": 4.999824876184517e-06,
50
+ "loss": 0.6258,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.18925312605610004,
55
+ "grad_norm": 2.21572570092998,
56
+ "learning_rate": 4.997855020764802e-06,
57
+ "loss": 0.62,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.21628928692125718,
62
+ "grad_norm": 2.4335006932811805,
63
+ "learning_rate": 4.99369817095752e-06,
64
+ "loss": 0.6104,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.24332544778641432,
69
+ "grad_norm": 3.0492932227575755,
70
+ "learning_rate": 4.98735804062935e-06,
71
+ "loss": 0.6067,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.2703616086515715,
76
+ "grad_norm": 2.148281590957587,
77
+ "learning_rate": 4.978840294261777e-06,
78
+ "loss": 0.6084,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.29739776951672864,
83
+ "grad_norm": 1.7664347921328134,
84
+ "learning_rate": 4.968152541890256e-06,
85
+ "loss": 0.6054,
86
+ "step": 110
87
+ },
88
+ {
89
+ "epoch": 0.3244339303818858,
90
+ "grad_norm": 1.8560219332271684,
91
+ "learning_rate": 4.955304332305161e-06,
92
+ "loss": 0.6012,
93
+ "step": 120
94
+ },
95
+ {
96
+ "epoch": 0.3514700912470429,
97
+ "grad_norm": 1.8079805894465493,
98
+ "learning_rate": 4.9403071445205725e-06,
99
+ "loss": 0.6028,
100
+ "step": 130
101
+ },
102
+ {
103
+ "epoch": 0.37850625211220007,
104
+ "grad_norm": 1.8246347697698218,
105
+ "learning_rate": 4.923174377518552e-06,
106
+ "loss": 0.6014,
107
+ "step": 140
108
+ },
109
+ {
110
+ "epoch": 0.4055424129773572,
111
+ "grad_norm": 2.749005509317375,
112
+ "learning_rate": 4.903921338278051e-06,
113
+ "loss": 0.5932,
114
+ "step": 150
115
+ },
116
+ {
117
+ "epoch": 0.43257857384251436,
118
+ "grad_norm": 2.244874423316534,
119
+ "learning_rate": 4.882565228099148e-06,
120
+ "loss": 0.5958,
121
+ "step": 160
122
+ },
123
+ {
124
+ "epoch": 0.4596147347076715,
125
+ "grad_norm": 1.8552147310211213,
126
+ "learning_rate": 4.859125127234842e-06,
127
+ "loss": 0.5946,
128
+ "step": 170
129
+ },
130
+ {
131
+ "epoch": 0.48665089557282865,
132
+ "grad_norm": 2.058882586288013,
133
+ "learning_rate": 4.833621977844127e-06,
134
+ "loss": 0.5931,
135
+ "step": 180
136
+ },
137
+ {
138
+ "epoch": 0.5136870564379858,
139
+ "grad_norm": 1.7633755533968514,
140
+ "learning_rate": 4.80607856528157e-06,
141
+ "loss": 0.5905,
142
+ "step": 190
143
+ },
144
+ {
145
+ "epoch": 0.540723217303143,
146
+ "grad_norm": 1.5062385913475653,
147
+ "learning_rate": 4.776519497740133e-06,
148
+ "loss": 0.5881,
149
+ "step": 200
150
+ },
151
+ {
152
+ "epoch": 0.5677593781683001,
153
+ "grad_norm": 1.649775437241844,
154
+ "learning_rate": 4.744971184265394e-06,
155
+ "loss": 0.5898,
156
+ "step": 210
157
+ },
158
+ {
159
+ "epoch": 0.5947955390334573,
160
+ "grad_norm": 1.4372383743256916,
161
+ "learning_rate": 4.711461811160839e-06,
162
+ "loss": 0.5852,
163
+ "step": 220
164
+ },
165
+ {
166
+ "epoch": 0.6218316998986144,
167
+ "grad_norm": 1.6374054025057205,
168
+ "learning_rate": 4.676021316805291e-06,
169
+ "loss": 0.5864,
170
+ "step": 230
171
+ },
172
+ {
173
+ "epoch": 0.6488678607637716,
174
+ "grad_norm": 1.3705672684359966,
175
+ "learning_rate": 4.638681364904967e-06,
176
+ "loss": 0.5828,
177
+ "step": 240
178
+ },
179
+ {
180
+ "epoch": 0.6759040216289287,
181
+ "grad_norm": 1.4497378255708682,
182
+ "learning_rate": 4.599475316204093e-06,
183
+ "loss": 0.576,
184
+ "step": 250
185
+ },
186
+ {
187
+ "epoch": 0.7029401824940859,
188
+ "grad_norm": 1.4901529664738118,
189
+ "learning_rate": 4.558438198679298e-06,
190
+ "loss": 0.5789,
191
+ "step": 260
192
+ },
193
+ {
194
+ "epoch": 0.729976343359243,
195
+ "grad_norm": 1.509487695379131,
196
+ "learning_rate": 4.515606676244479e-06,
197
+ "loss": 0.5823,
198
+ "step": 270
199
+ },
200
+ {
201
+ "epoch": 0.7570125042244001,
202
+ "grad_norm": 1.4142815759753038,
203
+ "learning_rate": 4.471019015994042e-06,
204
+ "loss": 0.5763,
205
+ "step": 280
206
+ },
207
+ {
208
+ "epoch": 0.7840486650895573,
209
+ "grad_norm": 1.391721023396763,
210
+ "learning_rate": 4.424715054013825e-06,
211
+ "loss": 0.5763,
212
+ "step": 290
213
+ },
214
+ {
215
+ "epoch": 0.8110848259547144,
216
+ "grad_norm": 2.036811255977957,
217
+ "learning_rate": 4.376736159790221e-06,
218
+ "loss": 0.5768,
219
+ "step": 300
220
+ },
221
+ {
222
+ "epoch": 0.8381209868198716,
223
+ "grad_norm": 1.8743584001202067,
224
+ "learning_rate": 4.327125199249313e-06,
225
+ "loss": 0.5742,
226
+ "step": 310
227
+ },
228
+ {
229
+ "epoch": 0.8651571476850287,
230
+ "grad_norm": 1.7592149703550843,
231
+ "learning_rate": 4.275926496459052e-06,
232
+ "loss": 0.5721,
233
+ "step": 320
234
+ },
235
+ {
236
+ "epoch": 0.8921933085501859,
237
+ "grad_norm": 1.1705725093085595,
238
+ "learning_rate": 4.223185794028659e-06,
239
+ "loss": 0.5711,
240
+ "step": 330
241
+ },
242
+ {
243
+ "epoch": 0.919229469415343,
244
+ "grad_norm": 1.9309874457184006,
245
+ "learning_rate": 4.168950212240682e-06,
246
+ "loss": 0.5782,
247
+ "step": 340
248
+ },
249
+ {
250
+ "epoch": 0.9462656302805001,
251
+ "grad_norm": 2.440179989180274,
252
+ "learning_rate": 4.113268206952177e-06,
253
+ "loss": 0.57,
254
+ "step": 350
255
+ },
256
+ {
257
+ "epoch": 0.9733017911456573,
258
+ "grad_norm": 1.7341492741288718,
259
+ "learning_rate": 4.056189526302645e-06,
260
+ "loss": 0.5697,
261
+ "step": 360
262
+ },
263
+ {
264
+ "epoch": 0.9976343359242987,
265
+ "eval_loss": 0.07097312808036804,
266
+ "eval_runtime": 379.122,
267
+ "eval_samples_per_second": 26.284,
268
+ "eval_steps_per_second": 0.411,
269
+ "step": 369
270
+ },
271
+ {
272
+ "epoch": 1.0023656640757013,
273
+ "grad_norm": 3.541884342240149,
274
+ "learning_rate": 3.997765166267408e-06,
275
+ "loss": 0.5655,
276
+ "step": 370
277
+ },
278
+ {
279
+ "epoch": 1.0294018249408583,
280
+ "grad_norm": 2.144287711298856,
281
+ "learning_rate": 3.9380473250961195e-06,
282
+ "loss": 0.4883,
283
+ "step": 380
284
+ },
285
+ {
286
+ "epoch": 1.0564379858060156,
287
+ "grad_norm": 1.6382074544301437,
288
+ "learning_rate": 3.877089356677135e-06,
289
+ "loss": 0.4839,
290
+ "step": 390
291
+ },
292
+ {
293
+ "epoch": 1.0834741466711728,
294
+ "grad_norm": 1.5188972286145213,
295
+ "learning_rate": 3.814945722869378e-06,
296
+ "loss": 0.4853,
297
+ "step": 400
298
+ },
299
+ {
300
+ "epoch": 1.1105103075363298,
301
+ "grad_norm": 1.3764586999097552,
302
+ "learning_rate": 3.7516719448443244e-06,
303
+ "loss": 0.485,
304
+ "step": 410
305
+ },
306
+ {
307
+ "epoch": 1.1375464684014869,
308
+ "grad_norm": 1.3804469321724957,
309
+ "learning_rate": 3.6873245534815626e-06,
310
+ "loss": 0.4873,
311
+ "step": 420
312
+ },
313
+ {
314
+ "epoch": 1.1645826292666441,
315
+ "grad_norm": 1.4074948363242845,
316
+ "learning_rate": 3.621961038862231e-06,
317
+ "loss": 0.495,
318
+ "step": 430
319
+ },
320
+ {
321
+ "epoch": 1.1916187901318014,
322
+ "grad_norm": 1.3672614070984725,
323
+ "learning_rate": 3.5556397989054944e-06,
324
+ "loss": 0.4861,
325
+ "step": 440
326
+ },
327
+ {
328
+ "epoch": 1.2186549509969584,
329
+ "grad_norm": 1.5256951985176181,
330
+ "learning_rate": 3.4884200871939088e-06,
331
+ "loss": 0.4904,
332
+ "step": 450
333
+ },
334
+ {
335
+ "epoch": 1.2456911118621157,
336
+ "grad_norm": 1.8390241289827158,
337
+ "learning_rate": 3.4203619600343226e-06,
338
+ "loss": 0.4923,
339
+ "step": 460
340
+ },
341
+ {
342
+ "epoch": 1.2727272727272727,
343
+ "grad_norm": 1.4516571737107622,
344
+ "learning_rate": 3.35152622280159e-06,
345
+ "loss": 0.4932,
346
+ "step": 470
347
+ },
348
+ {
349
+ "epoch": 1.29976343359243,
350
+ "grad_norm": 1.6389751646228208,
351
+ "learning_rate": 3.281974375613045e-06,
352
+ "loss": 0.4882,
353
+ "step": 480
354
+ },
355
+ {
356
+ "epoch": 1.326799594457587,
357
+ "grad_norm": 1.4400847542447044,
358
+ "learning_rate": 3.2117685583822672e-06,
359
+ "loss": 0.4877,
360
+ "step": 490
361
+ },
362
+ {
363
+ "epoch": 1.3538357553227442,
364
+ "grad_norm": 1.3579190420582792,
365
+ "learning_rate": 3.1409714953012355e-06,
366
+ "loss": 0.4897,
367
+ "step": 500
368
+ },
369
+ {
370
+ "epoch": 1.3808719161879013,
371
+ "grad_norm": 1.9092857680970101,
372
+ "learning_rate": 3.069646438800466e-06,
373
+ "loss": 0.4905,
374
+ "step": 510
375
+ },
376
+ {
377
+ "epoch": 1.4079080770530585,
378
+ "grad_norm": 1.3055916098302385,
379
+ "learning_rate": 2.9978571130372045e-06,
380
+ "loss": 0.4928,
381
+ "step": 520
382
+ },
383
+ {
384
+ "epoch": 1.4349442379182156,
385
+ "grad_norm": 1.4382290860035074,
386
+ "learning_rate": 2.925667656962165e-06,
387
+ "loss": 0.485,
388
+ "step": 530
389
+ },
390
+ {
391
+ "epoch": 1.4619803987833728,
392
+ "grad_norm": 1.522057711526301,
393
+ "learning_rate": 2.8531425670156767e-06,
394
+ "loss": 0.4897,
395
+ "step": 540
396
+ },
397
+ {
398
+ "epoch": 1.4890165596485299,
399
+ "grad_norm": 1.4339512710367517,
400
+ "learning_rate": 2.7803466395044337e-06,
401
+ "loss": 0.4876,
402
+ "step": 550
403
+ },
404
+ {
405
+ "epoch": 1.5160527205136871,
406
+ "grad_norm": 1.3073643288950498,
407
+ "learning_rate": 2.707344912710342e-06,
408
+ "loss": 0.4891,
409
+ "step": 560
410
+ },
411
+ {
412
+ "epoch": 1.5430888813788441,
413
+ "grad_norm": 1.369774353076571,
414
+ "learning_rate": 2.634202608783171e-06,
415
+ "loss": 0.4845,
416
+ "step": 570
417
+ },
418
+ {
419
+ "epoch": 1.5701250422440014,
420
+ "grad_norm": 1.2255590052602647,
421
+ "learning_rate": 2.5609850754689393e-06,
422
+ "loss": 0.486,
423
+ "step": 580
424
+ },
425
+ {
426
+ "epoch": 1.5971612031091587,
427
+ "grad_norm": 1.282149554684781,
428
+ "learning_rate": 2.4877577277260773e-06,
429
+ "loss": 0.489,
430
+ "step": 590
431
+ },
432
+ {
433
+ "epoch": 1.6241973639743157,
434
+ "grad_norm": 1.2721046491548977,
435
+ "learning_rate": 2.4145859892815512e-06,
436
+ "loss": 0.4858,
437
+ "step": 600
438
+ },
439
+ {
440
+ "epoch": 1.6512335248394727,
441
+ "grad_norm": 1.3827702644219955,
442
+ "learning_rate": 2.3415352341791467e-06,
443
+ "loss": 0.4837,
444
+ "step": 610
445
+ },
446
+ {
447
+ "epoch": 1.67826968570463,
448
+ "grad_norm": 1.327575194327116,
449
+ "learning_rate": 2.2686707283721406e-06,
450
+ "loss": 0.4877,
451
+ "step": 620
452
+ },
453
+ {
454
+ "epoch": 1.7053058465697872,
455
+ "grad_norm": 1.3568828598993092,
456
+ "learning_rate": 2.19605757141255e-06,
457
+ "loss": 0.4926,
458
+ "step": 630
459
+ },
460
+ {
461
+ "epoch": 1.7323420074349443,
462
+ "grad_norm": 1.2509722338502092,
463
+ "learning_rate": 2.123760638289043e-06,
464
+ "loss": 0.4893,
465
+ "step": 640
466
+ },
467
+ {
468
+ "epoch": 1.7593781683001013,
469
+ "grad_norm": 1.621567247889154,
470
+ "learning_rate": 2.0518445214654866e-06,
471
+ "loss": 0.487,
472
+ "step": 650
473
+ },
474
+ {
475
+ "epoch": 1.7864143291652586,
476
+ "grad_norm": 1.269029986877882,
477
+ "learning_rate": 1.980373473171907e-06,
478
+ "loss": 0.4835,
479
+ "step": 660
480
+ },
481
+ {
482
+ "epoch": 1.8134504900304158,
483
+ "grad_norm": 1.1955677021994464,
484
+ "learning_rate": 1.9094113479994346e-06,
485
+ "loss": 0.4849,
486
+ "step": 670
487
+ },
488
+ {
489
+ "epoch": 1.8404866508955728,
490
+ "grad_norm": 1.2345433836895534,
491
+ "learning_rate": 1.839021545850499e-06,
492
+ "loss": 0.4875,
493
+ "step": 680
494
+ },
495
+ {
496
+ "epoch": 1.8675228117607299,
497
+ "grad_norm": 1.2564382135038328,
498
+ "learning_rate": 1.7692669552952723e-06,
499
+ "loss": 0.4858,
500
+ "step": 690
501
+ },
502
+ {
503
+ "epoch": 1.8945589726258871,
504
+ "grad_norm": 1.256899424662725,
505
+ "learning_rate": 1.7002098973849416e-06,
506
+ "loss": 0.4872,
507
+ "step": 700
508
+ },
509
+ {
510
+ "epoch": 1.9215951334910444,
511
+ "grad_norm": 1.357653334253869,
512
+ "learning_rate": 1.631912069972033e-06,
513
+ "loss": 0.4916,
514
+ "step": 710
515
+ },
516
+ {
517
+ "epoch": 1.9486312943562014,
518
+ "grad_norm": 1.2314684983309747,
519
+ "learning_rate": 1.5644344925875097e-06,
520
+ "loss": 0.4836,
521
+ "step": 720
522
+ },
523
+ {
524
+ "epoch": 1.9756674552213584,
525
+ "grad_norm": 1.1478454732004184,
526
+ "learning_rate": 1.4978374519239131e-06,
527
+ "loss": 0.4788,
528
+ "step": 730
529
+ },
530
+ {
531
+ "epoch": 1.9972963839134843,
532
+ "eval_loss": 0.07031858712434769,
533
+ "eval_runtime": 378.0397,
534
+ "eval_samples_per_second": 26.36,
535
+ "eval_steps_per_second": 0.413,
536
+ "step": 738
537
+ },
538
+ {
539
+ "epoch": 2.0047313281514025,
540
+ "grad_norm": 3.2511739123964976,
541
+ "learning_rate": 1.432180447973249e-06,
542
+ "loss": 0.4716,
543
+ "step": 740
544
+ },
545
+ {
546
+ "epoch": 2.0317674890165596,
547
+ "grad_norm": 1.722212954142693,
548
+ "learning_rate": 1.3675221408677241e-06,
549
+ "loss": 0.4027,
550
+ "step": 750
551
+ },
552
+ {
553
+ "epoch": 2.0588036498817166,
554
+ "grad_norm": 1.5303304387352163,
555
+ "learning_rate": 1.303920298470856e-06,
556
+ "loss": 0.4068,
557
+ "step": 760
558
+ },
559
+ {
560
+ "epoch": 2.085839810746874,
561
+ "grad_norm": 1.3702186369869054,
562
+ "learning_rate": 1.2414317447657564e-06,
563
+ "loss": 0.4028,
564
+ "step": 770
565
+ },
566
+ {
567
+ "epoch": 2.112875971612031,
568
+ "grad_norm": 1.3334591253881896,
569
+ "learning_rate": 1.1801123090867129e-06,
570
+ "loss": 0.4024,
571
+ "step": 780
572
+ },
573
+ {
574
+ "epoch": 2.139912132477188,
575
+ "grad_norm": 1.370776472523979,
576
+ "learning_rate": 1.1200167762394187e-06,
577
+ "loss": 0.4016,
578
+ "step": 790
579
+ },
580
+ {
581
+ "epoch": 2.1669482933423456,
582
+ "grad_norm": 1.3101835210391692,
583
+ "learning_rate": 1.0611988375544214e-06,
584
+ "loss": 0.4072,
585
+ "step": 800
586
+ },
587
+ {
588
+ "epoch": 2.1939844542075027,
589
+ "grad_norm": 1.2890483237347024,
590
+ "learning_rate": 1.0037110429175214e-06,
591
+ "loss": 0.4043,
592
+ "step": 810
593
+ },
594
+ {
595
+ "epoch": 2.2210206150726597,
596
+ "grad_norm": 1.2394879407914554,
597
+ "learning_rate": 9.47604753819966e-07,
598
+ "loss": 0.4049,
599
+ "step": 820
600
+ },
601
+ {
602
+ "epoch": 2.2480567759378167,
603
+ "grad_norm": 1.2626807074989126,
604
+ "learning_rate": 8.929300974704016e-07,
605
+ "loss": 0.407,
606
+ "step": 830
607
+ },
608
+ {
609
+ "epoch": 2.2750929368029738,
610
+ "grad_norm": 1.2526670795727008,
611
+ "learning_rate": 8.397359220095739e-07,
612
+ "loss": 0.407,
613
+ "step": 840
614
+ },
615
+ {
616
+ "epoch": 2.3021290976681312,
617
+ "grad_norm": 1.2912703090742628,
618
+ "learning_rate": 7.88069752867787e-07,
619
+ "loss": 0.4012,
620
+ "step": 850
621
+ },
622
+ {
623
+ "epoch": 2.3291652585332883,
624
+ "grad_norm": 1.2248414426266379,
625
+ "learning_rate": 7.379777503041166e-07,
626
+ "loss": 0.4042,
627
+ "step": 860
628
+ },
629
+ {
630
+ "epoch": 2.3562014193984453,
631
+ "grad_norm": 1.245734138527781,
632
+ "learning_rate": 6.89504668165316e-07,
633
+ "loss": 0.4083,
634
+ "step": 870
635
+ },
636
+ {
637
+ "epoch": 2.3832375802636028,
638
+ "grad_norm": 1.2703147157094865,
639
+ "learning_rate": 6.426938139012476e-07,
640
+ "loss": 0.4068,
641
+ "step": 880
642
+ },
643
+ {
644
+ "epoch": 2.41027374112876,
645
+ "grad_norm": 1.26460233866951,
646
+ "learning_rate": 5.975870098725839e-07,
647
+ "loss": 0.4052,
648
+ "step": 890
649
+ },
650
+ {
651
+ "epoch": 2.437309901993917,
652
+ "grad_norm": 1.203035775818215,
653
+ "learning_rate": 5.542245559853314e-07,
654
+ "loss": 0.4043,
655
+ "step": 900
656
+ },
657
+ {
658
+ "epoch": 2.464346062859074,
659
+ "grad_norm": 1.2801464316323534,
660
+ "learning_rate": 5.12645193685563e-07,
661
+ "loss": 0.4026,
662
+ "step": 910
663
+ },
664
+ {
665
+ "epoch": 2.4913822237242313,
666
+ "grad_norm": 1.207211407952489,
667
+ "learning_rate": 4.728860713465348e-07,
668
+ "loss": 0.4018,
669
+ "step": 920
670
+ },
671
+ {
672
+ "epoch": 2.5184183845893884,
673
+ "grad_norm": 1.2978323617108884,
674
+ "learning_rate": 4.3498271107910423e-07,
675
+ "loss": 0.4006,
676
+ "step": 930
677
+ },
678
+ {
679
+ "epoch": 2.5454545454545454,
680
+ "grad_norm": 1.2419638574449388,
681
+ "learning_rate": 3.989689769951037e-07,
682
+ "loss": 0.4072,
683
+ "step": 940
684
+ },
685
+ {
686
+ "epoch": 2.5724907063197024,
687
+ "grad_norm": 1.2184809628304476,
688
+ "learning_rate": 3.6487704495203003e-07,
689
+ "loss": 0.4012,
690
+ "step": 950
691
+ },
692
+ {
693
+ "epoch": 2.59952686718486,
694
+ "grad_norm": 1.2408761979548522,
695
+ "learning_rate": 3.3273737380606715e-07,
696
+ "loss": 0.407,
697
+ "step": 960
698
+ },
699
+ {
700
+ "epoch": 2.626563028050017,
701
+ "grad_norm": 1.2318951856879983,
702
+ "learning_rate": 3.02578678199144e-07,
703
+ "loss": 0.4042,
704
+ "step": 970
705
+ },
706
+ {
707
+ "epoch": 2.653599188915174,
708
+ "grad_norm": 1.2341509325385753,
709
+ "learning_rate": 2.74427902904325e-07,
710
+ "loss": 0.4074,
711
+ "step": 980
712
+ },
713
+ {
714
+ "epoch": 2.6806353497803315,
715
+ "grad_norm": 1.2021221404178184,
716
+ "learning_rate": 2.483101987524612e-07,
717
+ "loss": 0.4048,
718
+ "step": 990
719
+ },
720
+ {
721
+ "epoch": 2.7076715106454885,
722
+ "grad_norm": 1.2738693365179738,
723
+ "learning_rate": 2.2424890016160911e-07,
724
+ "loss": 0.3986,
725
+ "step": 1000
726
+ },
727
+ {
728
+ "epoch": 2.7347076715106455,
729
+ "grad_norm": 1.2692209511322345,
730
+ "learning_rate": 2.022655042892919e-07,
731
+ "loss": 0.4058,
732
+ "step": 1010
733
+ },
734
+ {
735
+ "epoch": 2.7617438323758026,
736
+ "grad_norm": 1.1771824503326054,
737
+ "learning_rate": 1.823796518262309e-07,
738
+ "loss": 0.4071,
739
+ "step": 1020
740
+ },
741
+ {
742
+ "epoch": 2.7887799932409596,
743
+ "grad_norm": 1.223679689500554,
744
+ "learning_rate": 1.6460910944870367e-07,
745
+ "loss": 0.4053,
746
+ "step": 1030
747
+ },
748
+ {
749
+ "epoch": 2.815816154106117,
750
+ "grad_norm": 1.1639416605117765,
751
+ "learning_rate": 1.489697539452106e-07,
752
+ "loss": 0.403,
753
+ "step": 1040
754
+ },
755
+ {
756
+ "epoch": 2.842852314971274,
757
+ "grad_norm": 1.191061950274548,
758
+ "learning_rate": 1.354755580316274e-07,
759
+ "loss": 0.4043,
760
+ "step": 1050
761
+ },
762
+ {
763
+ "epoch": 2.869888475836431,
764
+ "grad_norm": 1.1905738035063456,
765
+ "learning_rate": 1.2413857786752143e-07,
766
+ "loss": 0.399,
767
+ "step": 1060
768
+ },
769
+ {
770
+ "epoch": 2.8969246367015886,
771
+ "grad_norm": 1.1956160773000035,
772
+ "learning_rate": 1.1496894228478066e-07,
773
+ "loss": 0.4072,
774
+ "step": 1070
775
+ },
776
+ {
777
+ "epoch": 2.9239607975667457,
778
+ "grad_norm": 1.2066740723044425,
779
+ "learning_rate": 1.0797484373818065e-07,
780
+ "loss": 0.4014,
781
+ "step": 1080
782
+ },
783
+ {
784
+ "epoch": 2.9509969584319027,
785
+ "grad_norm": 1.1778515542586203,
786
+ "learning_rate": 1.0316253098597634e-07,
787
+ "loss": 0.4045,
788
+ "step": 1090
789
+ },
790
+ {
791
+ "epoch": 2.9780331192970597,
792
+ "grad_norm": 1.1602330819105369,
793
+ "learning_rate": 1.0053630350705497e-07,
794
+ "loss": 0.4034,
795
+ "step": 1100
796
+ },
797
+ {
798
+ "epoch": 2.99695843190267,
799
+ "eval_loss": 0.07343952357769012,
800
+ "eval_runtime": 380.9855,
801
+ "eval_samples_per_second": 26.156,
802
+ "eval_steps_per_second": 0.409,
803
+ "step": 1107
804
+ },
805
+ {
806
+ "epoch": 2.99695843190267,
807
+ "step": 1107,
808
+ "total_flos": 1854056851046400.0,
809
+ "train_loss": 0.4996785878489011,
810
+ "train_runtime": 63355.7173,
811
+ "train_samples_per_second": 8.965,
812
+ "train_steps_per_second": 0.017
813
+ }
814
+ ],
815
+ "logging_steps": 10,
816
+ "max_steps": 1107,
817
+ "num_input_tokens_seen": 0,
818
+ "num_train_epochs": 3,
819
+ "save_steps": 500,
820
+ "stateful_callbacks": {
821
+ "TrainerControl": {
822
+ "args": {
823
+ "should_epoch_stop": false,
824
+ "should_evaluate": false,
825
+ "should_log": false,
826
+ "should_save": true,
827
+ "should_training_stop": true
828
+ },
829
+ "attributes": {}
830
+ }
831
+ },
832
+ "total_flos": 1854056851046400.0,
833
+ "train_batch_size": 8,
834
+ "trial_name": null,
835
+ "trial_params": null
836
+ }
training_eval_loss.png ADDED
training_loss.png ADDED