gsmyrnis commited on
Commit
0744ae0
·
verified ·
1 Parent(s): b86913a

End of training

Browse files
README.md CHANGED
@@ -4,6 +4,7 @@ license: llama3.1
4
  base_model: meta-llama/Meta-Llama-3.1-8B
5
  tags:
6
  - llama-factory
 
7
  - generated_from_trainer
8
  model-index:
9
  - name: oh-dcft-v1.2_no-curation_gpt-4o-mini_wo_unnatural_instructions
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  # oh-dcft-v1.2_no-curation_gpt-4o-mini_wo_unnatural_instructions
17
 
18
- This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B) on an unknown dataset.
19
  It achieves the following results on the evaluation set:
20
  - Loss: 0.6580
21
 
 
4
  base_model: meta-llama/Meta-Llama-3.1-8B
5
  tags:
6
  - llama-factory
7
+ - full
8
  - generated_from_trainer
9
  model-index:
10
  - name: oh-dcft-v1.2_no-curation_gpt-4o-mini_wo_unnatural_instructions
 
16
 
17
  # oh-dcft-v1.2_no-curation_gpt-4o-mini_wo_unnatural_instructions
18
 
19
+ This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B) on the mlfoundations-dev/oh-dcft-v1.2_no-curation_gpt-4o-mini_wo_unnatural_instructions dataset.
20
  It achieves the following results on the evaluation set:
21
  - Loss: 0.6580
22
 
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_loss": 0.6580318212509155,
4
+ "eval_runtime": 30.4158,
5
+ "eval_samples_per_second": 285.575,
6
+ "eval_steps_per_second": 1.118,
7
+ "total_flos": 1623111090831360.0,
8
+ "train_loss": 0.628350810984955,
9
+ "train_runtime": 6073.5579,
10
+ "train_samples_per_second": 81.516,
11
+ "train_steps_per_second": 0.16
12
+ }
eval_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_loss": 0.6580318212509155,
4
+ "eval_runtime": 30.4158,
5
+ "eval_samples_per_second": 285.575,
6
+ "eval_steps_per_second": 1.118
7
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "total_flos": 1623111090831360.0,
4
+ "train_loss": 0.628350810984955,
5
+ "train_runtime": 6073.5579,
6
+ "train_samples_per_second": 81.516,
7
+ "train_steps_per_second": 0.16
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,738 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "eval_steps": 500,
6
+ "global_step": 969,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.030959752321981424,
13
+ "grad_norm": 9.26792445696984,
14
+ "learning_rate": 5e-06,
15
+ "loss": 0.9296,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.06191950464396285,
20
+ "grad_norm": 1.9224039875777081,
21
+ "learning_rate": 5e-06,
22
+ "loss": 0.826,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.09287925696594428,
27
+ "grad_norm": 4.074240624021427,
28
+ "learning_rate": 5e-06,
29
+ "loss": 0.7818,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.1238390092879257,
34
+ "grad_norm": 1.5813292349658419,
35
+ "learning_rate": 5e-06,
36
+ "loss": 0.7624,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.15479876160990713,
41
+ "grad_norm": 0.8536213234663932,
42
+ "learning_rate": 5e-06,
43
+ "loss": 0.7313,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.18575851393188855,
48
+ "grad_norm": 1.087523888330213,
49
+ "learning_rate": 5e-06,
50
+ "loss": 0.7307,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.21671826625386997,
55
+ "grad_norm": 0.8453603539631231,
56
+ "learning_rate": 5e-06,
57
+ "loss": 0.7126,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.2476780185758514,
62
+ "grad_norm": 0.9824688751534248,
63
+ "learning_rate": 5e-06,
64
+ "loss": 0.7113,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.2786377708978328,
69
+ "grad_norm": 0.9326537359571924,
70
+ "learning_rate": 5e-06,
71
+ "loss": 0.7037,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.30959752321981426,
76
+ "grad_norm": 0.548855695013045,
77
+ "learning_rate": 5e-06,
78
+ "loss": 0.6988,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.34055727554179566,
83
+ "grad_norm": 0.5127345076323682,
84
+ "learning_rate": 5e-06,
85
+ "loss": 0.6969,
86
+ "step": 110
87
+ },
88
+ {
89
+ "epoch": 0.3715170278637771,
90
+ "grad_norm": 0.5346034637159197,
91
+ "learning_rate": 5e-06,
92
+ "loss": 0.6932,
93
+ "step": 120
94
+ },
95
+ {
96
+ "epoch": 0.4024767801857585,
97
+ "grad_norm": 0.6366003373202664,
98
+ "learning_rate": 5e-06,
99
+ "loss": 0.6862,
100
+ "step": 130
101
+ },
102
+ {
103
+ "epoch": 0.43343653250773995,
104
+ "grad_norm": 0.659232587562787,
105
+ "learning_rate": 5e-06,
106
+ "loss": 0.6861,
107
+ "step": 140
108
+ },
109
+ {
110
+ "epoch": 0.46439628482972134,
111
+ "grad_norm": 0.7626299453869156,
112
+ "learning_rate": 5e-06,
113
+ "loss": 0.6867,
114
+ "step": 150
115
+ },
116
+ {
117
+ "epoch": 0.4953560371517028,
118
+ "grad_norm": 0.600108061207952,
119
+ "learning_rate": 5e-06,
120
+ "loss": 0.6761,
121
+ "step": 160
122
+ },
123
+ {
124
+ "epoch": 0.5263157894736842,
125
+ "grad_norm": 0.6562920403508146,
126
+ "learning_rate": 5e-06,
127
+ "loss": 0.6822,
128
+ "step": 170
129
+ },
130
+ {
131
+ "epoch": 0.5572755417956656,
132
+ "grad_norm": 0.554500335402875,
133
+ "learning_rate": 5e-06,
134
+ "loss": 0.6771,
135
+ "step": 180
136
+ },
137
+ {
138
+ "epoch": 0.5882352941176471,
139
+ "grad_norm": 0.7448258243980852,
140
+ "learning_rate": 5e-06,
141
+ "loss": 0.6725,
142
+ "step": 190
143
+ },
144
+ {
145
+ "epoch": 0.6191950464396285,
146
+ "grad_norm": 0.7212261495226227,
147
+ "learning_rate": 5e-06,
148
+ "loss": 0.6704,
149
+ "step": 200
150
+ },
151
+ {
152
+ "epoch": 0.6501547987616099,
153
+ "grad_norm": 0.650918538222657,
154
+ "learning_rate": 5e-06,
155
+ "loss": 0.6747,
156
+ "step": 210
157
+ },
158
+ {
159
+ "epoch": 0.6811145510835913,
160
+ "grad_norm": 0.5979672431899267,
161
+ "learning_rate": 5e-06,
162
+ "loss": 0.6721,
163
+ "step": 220
164
+ },
165
+ {
166
+ "epoch": 0.7120743034055728,
167
+ "grad_norm": 0.5587950559181737,
168
+ "learning_rate": 5e-06,
169
+ "loss": 0.6775,
170
+ "step": 230
171
+ },
172
+ {
173
+ "epoch": 0.7430340557275542,
174
+ "grad_norm": 0.7504748626653679,
175
+ "learning_rate": 5e-06,
176
+ "loss": 0.6772,
177
+ "step": 240
178
+ },
179
+ {
180
+ "epoch": 0.7739938080495357,
181
+ "grad_norm": 0.5672124597726503,
182
+ "learning_rate": 5e-06,
183
+ "loss": 0.6649,
184
+ "step": 250
185
+ },
186
+ {
187
+ "epoch": 0.804953560371517,
188
+ "grad_norm": 0.6667059518291187,
189
+ "learning_rate": 5e-06,
190
+ "loss": 0.6628,
191
+ "step": 260
192
+ },
193
+ {
194
+ "epoch": 0.8359133126934984,
195
+ "grad_norm": 0.5557591111686754,
196
+ "learning_rate": 5e-06,
197
+ "loss": 0.666,
198
+ "step": 270
199
+ },
200
+ {
201
+ "epoch": 0.8668730650154799,
202
+ "grad_norm": 0.7400262670967761,
203
+ "learning_rate": 5e-06,
204
+ "loss": 0.6608,
205
+ "step": 280
206
+ },
207
+ {
208
+ "epoch": 0.8978328173374613,
209
+ "grad_norm": 0.5053622688793449,
210
+ "learning_rate": 5e-06,
211
+ "loss": 0.6657,
212
+ "step": 290
213
+ },
214
+ {
215
+ "epoch": 0.9287925696594427,
216
+ "grad_norm": 0.611108366609724,
217
+ "learning_rate": 5e-06,
218
+ "loss": 0.6612,
219
+ "step": 300
220
+ },
221
+ {
222
+ "epoch": 0.9597523219814241,
223
+ "grad_norm": 0.5223978557328981,
224
+ "learning_rate": 5e-06,
225
+ "loss": 0.6585,
226
+ "step": 310
227
+ },
228
+ {
229
+ "epoch": 0.9907120743034056,
230
+ "grad_norm": 0.5084066168996301,
231
+ "learning_rate": 5e-06,
232
+ "loss": 0.6582,
233
+ "step": 320
234
+ },
235
+ {
236
+ "epoch": 1.0,
237
+ "eval_loss": 0.6614387631416321,
238
+ "eval_runtime": 31.0197,
239
+ "eval_samples_per_second": 280.015,
240
+ "eval_steps_per_second": 1.096,
241
+ "step": 323
242
+ },
243
+ {
244
+ "epoch": 1.021671826625387,
245
+ "grad_norm": 0.8185084818131129,
246
+ "learning_rate": 5e-06,
247
+ "loss": 0.6318,
248
+ "step": 330
249
+ },
250
+ {
251
+ "epoch": 1.0526315789473684,
252
+ "grad_norm": 0.7247142810746562,
253
+ "learning_rate": 5e-06,
254
+ "loss": 0.6153,
255
+ "step": 340
256
+ },
257
+ {
258
+ "epoch": 1.08359133126935,
259
+ "grad_norm": 0.574027398851242,
260
+ "learning_rate": 5e-06,
261
+ "loss": 0.6182,
262
+ "step": 350
263
+ },
264
+ {
265
+ "epoch": 1.1145510835913313,
266
+ "grad_norm": 0.5449783832911268,
267
+ "learning_rate": 5e-06,
268
+ "loss": 0.6155,
269
+ "step": 360
270
+ },
271
+ {
272
+ "epoch": 1.1455108359133126,
273
+ "grad_norm": 0.5025194777117987,
274
+ "learning_rate": 5e-06,
275
+ "loss": 0.6181,
276
+ "step": 370
277
+ },
278
+ {
279
+ "epoch": 1.1764705882352942,
280
+ "grad_norm": 0.5884698444538693,
281
+ "learning_rate": 5e-06,
282
+ "loss": 0.6145,
283
+ "step": 380
284
+ },
285
+ {
286
+ "epoch": 1.2074303405572755,
287
+ "grad_norm": 0.6402773293709152,
288
+ "learning_rate": 5e-06,
289
+ "loss": 0.6182,
290
+ "step": 390
291
+ },
292
+ {
293
+ "epoch": 1.238390092879257,
294
+ "grad_norm": 0.7376501065513348,
295
+ "learning_rate": 5e-06,
296
+ "loss": 0.6078,
297
+ "step": 400
298
+ },
299
+ {
300
+ "epoch": 1.2693498452012384,
301
+ "grad_norm": 0.5604731743792086,
302
+ "learning_rate": 5e-06,
303
+ "loss": 0.6178,
304
+ "step": 410
305
+ },
306
+ {
307
+ "epoch": 1.3003095975232197,
308
+ "grad_norm": 0.5374880157799365,
309
+ "learning_rate": 5e-06,
310
+ "loss": 0.6118,
311
+ "step": 420
312
+ },
313
+ {
314
+ "epoch": 1.3312693498452013,
315
+ "grad_norm": 0.5183594439426741,
316
+ "learning_rate": 5e-06,
317
+ "loss": 0.6152,
318
+ "step": 430
319
+ },
320
+ {
321
+ "epoch": 1.3622291021671826,
322
+ "grad_norm": 0.7429617462855133,
323
+ "learning_rate": 5e-06,
324
+ "loss": 0.6189,
325
+ "step": 440
326
+ },
327
+ {
328
+ "epoch": 1.3931888544891642,
329
+ "grad_norm": 0.5637315725076051,
330
+ "learning_rate": 5e-06,
331
+ "loss": 0.6143,
332
+ "step": 450
333
+ },
334
+ {
335
+ "epoch": 1.4241486068111455,
336
+ "grad_norm": 0.47806274705798535,
337
+ "learning_rate": 5e-06,
338
+ "loss": 0.6184,
339
+ "step": 460
340
+ },
341
+ {
342
+ "epoch": 1.4551083591331269,
343
+ "grad_norm": 0.7013859457502354,
344
+ "learning_rate": 5e-06,
345
+ "loss": 0.6146,
346
+ "step": 470
347
+ },
348
+ {
349
+ "epoch": 1.4860681114551084,
350
+ "grad_norm": 0.7728057436850783,
351
+ "learning_rate": 5e-06,
352
+ "loss": 0.6235,
353
+ "step": 480
354
+ },
355
+ {
356
+ "epoch": 1.5170278637770898,
357
+ "grad_norm": 0.5540033221172351,
358
+ "learning_rate": 5e-06,
359
+ "loss": 0.6146,
360
+ "step": 490
361
+ },
362
+ {
363
+ "epoch": 1.5479876160990713,
364
+ "grad_norm": 0.4989101370327553,
365
+ "learning_rate": 5e-06,
366
+ "loss": 0.6109,
367
+ "step": 500
368
+ },
369
+ {
370
+ "epoch": 1.5789473684210527,
371
+ "grad_norm": 0.6201415740790441,
372
+ "learning_rate": 5e-06,
373
+ "loss": 0.6127,
374
+ "step": 510
375
+ },
376
+ {
377
+ "epoch": 1.609907120743034,
378
+ "grad_norm": 0.6182247247832351,
379
+ "learning_rate": 5e-06,
380
+ "loss": 0.6158,
381
+ "step": 520
382
+ },
383
+ {
384
+ "epoch": 1.6408668730650153,
385
+ "grad_norm": 0.5273934388255005,
386
+ "learning_rate": 5e-06,
387
+ "loss": 0.6152,
388
+ "step": 530
389
+ },
390
+ {
391
+ "epoch": 1.671826625386997,
392
+ "grad_norm": 0.5657886835824731,
393
+ "learning_rate": 5e-06,
394
+ "loss": 0.6135,
395
+ "step": 540
396
+ },
397
+ {
398
+ "epoch": 1.7027863777089784,
399
+ "grad_norm": 0.6873412871431165,
400
+ "learning_rate": 5e-06,
401
+ "loss": 0.615,
402
+ "step": 550
403
+ },
404
+ {
405
+ "epoch": 1.7337461300309598,
406
+ "grad_norm": 0.5877043924887657,
407
+ "learning_rate": 5e-06,
408
+ "loss": 0.6125,
409
+ "step": 560
410
+ },
411
+ {
412
+ "epoch": 1.7647058823529411,
413
+ "grad_norm": 0.6037122620569243,
414
+ "learning_rate": 5e-06,
415
+ "loss": 0.6118,
416
+ "step": 570
417
+ },
418
+ {
419
+ "epoch": 1.7956656346749225,
420
+ "grad_norm": 0.4753962140527953,
421
+ "learning_rate": 5e-06,
422
+ "loss": 0.6128,
423
+ "step": 580
424
+ },
425
+ {
426
+ "epoch": 1.826625386996904,
427
+ "grad_norm": 0.5268653628592879,
428
+ "learning_rate": 5e-06,
429
+ "loss": 0.6172,
430
+ "step": 590
431
+ },
432
+ {
433
+ "epoch": 1.8575851393188856,
434
+ "grad_norm": 1.002687895644436,
435
+ "learning_rate": 5e-06,
436
+ "loss": 0.615,
437
+ "step": 600
438
+ },
439
+ {
440
+ "epoch": 1.888544891640867,
441
+ "grad_norm": 0.5100009376935558,
442
+ "learning_rate": 5e-06,
443
+ "loss": 0.6182,
444
+ "step": 610
445
+ },
446
+ {
447
+ "epoch": 1.9195046439628483,
448
+ "grad_norm": 0.691746404073842,
449
+ "learning_rate": 5e-06,
450
+ "loss": 0.6131,
451
+ "step": 620
452
+ },
453
+ {
454
+ "epoch": 1.9504643962848296,
455
+ "grad_norm": 0.5303891186235742,
456
+ "learning_rate": 5e-06,
457
+ "loss": 0.6152,
458
+ "step": 630
459
+ },
460
+ {
461
+ "epoch": 1.9814241486068112,
462
+ "grad_norm": 0.5949259061449034,
463
+ "learning_rate": 5e-06,
464
+ "loss": 0.6073,
465
+ "step": 640
466
+ },
467
+ {
468
+ "epoch": 2.0,
469
+ "eval_loss": 0.6524380445480347,
470
+ "eval_runtime": 31.1666,
471
+ "eval_samples_per_second": 278.696,
472
+ "eval_steps_per_second": 1.091,
473
+ "step": 646
474
+ },
475
+ {
476
+ "epoch": 2.0123839009287927,
477
+ "grad_norm": 0.9446095499864162,
478
+ "learning_rate": 5e-06,
479
+ "loss": 0.591,
480
+ "step": 650
481
+ },
482
+ {
483
+ "epoch": 2.043343653250774,
484
+ "grad_norm": 0.6258477650873515,
485
+ "learning_rate": 5e-06,
486
+ "loss": 0.5646,
487
+ "step": 660
488
+ },
489
+ {
490
+ "epoch": 2.0743034055727554,
491
+ "grad_norm": 0.604477009985834,
492
+ "learning_rate": 5e-06,
493
+ "loss": 0.5693,
494
+ "step": 670
495
+ },
496
+ {
497
+ "epoch": 2.1052631578947367,
498
+ "grad_norm": 0.5940857975430127,
499
+ "learning_rate": 5e-06,
500
+ "loss": 0.5621,
501
+ "step": 680
502
+ },
503
+ {
504
+ "epoch": 2.136222910216718,
505
+ "grad_norm": 0.6310773073567004,
506
+ "learning_rate": 5e-06,
507
+ "loss": 0.5664,
508
+ "step": 690
509
+ },
510
+ {
511
+ "epoch": 2.1671826625387,
512
+ "grad_norm": 0.6735784589983905,
513
+ "learning_rate": 5e-06,
514
+ "loss": 0.5659,
515
+ "step": 700
516
+ },
517
+ {
518
+ "epoch": 2.198142414860681,
519
+ "grad_norm": 0.6737551186617048,
520
+ "learning_rate": 5e-06,
521
+ "loss": 0.5744,
522
+ "step": 710
523
+ },
524
+ {
525
+ "epoch": 2.2291021671826625,
526
+ "grad_norm": 0.538363932283787,
527
+ "learning_rate": 5e-06,
528
+ "loss": 0.5634,
529
+ "step": 720
530
+ },
531
+ {
532
+ "epoch": 2.260061919504644,
533
+ "grad_norm": 0.6696764784705668,
534
+ "learning_rate": 5e-06,
535
+ "loss": 0.5633,
536
+ "step": 730
537
+ },
538
+ {
539
+ "epoch": 2.291021671826625,
540
+ "grad_norm": 0.5541664073886028,
541
+ "learning_rate": 5e-06,
542
+ "loss": 0.572,
543
+ "step": 740
544
+ },
545
+ {
546
+ "epoch": 2.321981424148607,
547
+ "grad_norm": 0.7344897597565605,
548
+ "learning_rate": 5e-06,
549
+ "loss": 0.5686,
550
+ "step": 750
551
+ },
552
+ {
553
+ "epoch": 2.3529411764705883,
554
+ "grad_norm": 0.5507126279939225,
555
+ "learning_rate": 5e-06,
556
+ "loss": 0.5729,
557
+ "step": 760
558
+ },
559
+ {
560
+ "epoch": 2.3839009287925697,
561
+ "grad_norm": 0.6344649027391833,
562
+ "learning_rate": 5e-06,
563
+ "loss": 0.5665,
564
+ "step": 770
565
+ },
566
+ {
567
+ "epoch": 2.414860681114551,
568
+ "grad_norm": 0.6106920196828644,
569
+ "learning_rate": 5e-06,
570
+ "loss": 0.5702,
571
+ "step": 780
572
+ },
573
+ {
574
+ "epoch": 2.4458204334365323,
575
+ "grad_norm": 0.6238727356689014,
576
+ "learning_rate": 5e-06,
577
+ "loss": 0.5706,
578
+ "step": 790
579
+ },
580
+ {
581
+ "epoch": 2.476780185758514,
582
+ "grad_norm": 0.6326270787372308,
583
+ "learning_rate": 5e-06,
584
+ "loss": 0.5739,
585
+ "step": 800
586
+ },
587
+ {
588
+ "epoch": 2.5077399380804954,
589
+ "grad_norm": 0.6392551743906308,
590
+ "learning_rate": 5e-06,
591
+ "loss": 0.5707,
592
+ "step": 810
593
+ },
594
+ {
595
+ "epoch": 2.538699690402477,
596
+ "grad_norm": 0.5293919880888138,
597
+ "learning_rate": 5e-06,
598
+ "loss": 0.5694,
599
+ "step": 820
600
+ },
601
+ {
602
+ "epoch": 2.569659442724458,
603
+ "grad_norm": 0.5315004573871298,
604
+ "learning_rate": 5e-06,
605
+ "loss": 0.5715,
606
+ "step": 830
607
+ },
608
+ {
609
+ "epoch": 2.6006191950464395,
610
+ "grad_norm": 0.6021834459451176,
611
+ "learning_rate": 5e-06,
612
+ "loss": 0.5686,
613
+ "step": 840
614
+ },
615
+ {
616
+ "epoch": 2.6315789473684212,
617
+ "grad_norm": 0.7264827387006807,
618
+ "learning_rate": 5e-06,
619
+ "loss": 0.5711,
620
+ "step": 850
621
+ },
622
+ {
623
+ "epoch": 2.6625386996904026,
624
+ "grad_norm": 0.5432440378468479,
625
+ "learning_rate": 5e-06,
626
+ "loss": 0.5682,
627
+ "step": 860
628
+ },
629
+ {
630
+ "epoch": 2.693498452012384,
631
+ "grad_norm": 0.4887763215481647,
632
+ "learning_rate": 5e-06,
633
+ "loss": 0.5727,
634
+ "step": 870
635
+ },
636
+ {
637
+ "epoch": 2.7244582043343653,
638
+ "grad_norm": 0.6219072904402111,
639
+ "learning_rate": 5e-06,
640
+ "loss": 0.5676,
641
+ "step": 880
642
+ },
643
+ {
644
+ "epoch": 2.7554179566563466,
645
+ "grad_norm": 0.5596606256063349,
646
+ "learning_rate": 5e-06,
647
+ "loss": 0.5748,
648
+ "step": 890
649
+ },
650
+ {
651
+ "epoch": 2.7863777089783284,
652
+ "grad_norm": 0.6091604993895532,
653
+ "learning_rate": 5e-06,
654
+ "loss": 0.5639,
655
+ "step": 900
656
+ },
657
+ {
658
+ "epoch": 2.8173374613003097,
659
+ "grad_norm": 0.5920477606163769,
660
+ "learning_rate": 5e-06,
661
+ "loss": 0.5717,
662
+ "step": 910
663
+ },
664
+ {
665
+ "epoch": 2.848297213622291,
666
+ "grad_norm": 0.6807900156580523,
667
+ "learning_rate": 5e-06,
668
+ "loss": 0.5721,
669
+ "step": 920
670
+ },
671
+ {
672
+ "epoch": 2.8792569659442724,
673
+ "grad_norm": 0.6050955818661157,
674
+ "learning_rate": 5e-06,
675
+ "loss": 0.5748,
676
+ "step": 930
677
+ },
678
+ {
679
+ "epoch": 2.9102167182662537,
680
+ "grad_norm": 0.6165235050612774,
681
+ "learning_rate": 5e-06,
682
+ "loss": 0.5732,
683
+ "step": 940
684
+ },
685
+ {
686
+ "epoch": 2.9411764705882355,
687
+ "grad_norm": 0.6346237356986199,
688
+ "learning_rate": 5e-06,
689
+ "loss": 0.5743,
690
+ "step": 950
691
+ },
692
+ {
693
+ "epoch": 2.972136222910217,
694
+ "grad_norm": 0.6125475070291018,
695
+ "learning_rate": 5e-06,
696
+ "loss": 0.5733,
697
+ "step": 960
698
+ },
699
+ {
700
+ "epoch": 3.0,
701
+ "eval_loss": 0.6580318212509155,
702
+ "eval_runtime": 31.0741,
703
+ "eval_samples_per_second": 279.526,
704
+ "eval_steps_per_second": 1.094,
705
+ "step": 969
706
+ },
707
+ {
708
+ "epoch": 3.0,
709
+ "step": 969,
710
+ "total_flos": 1623111090831360.0,
711
+ "train_loss": 0.628350810984955,
712
+ "train_runtime": 6073.5579,
713
+ "train_samples_per_second": 81.516,
714
+ "train_steps_per_second": 0.16
715
+ }
716
+ ],
717
+ "logging_steps": 10,
718
+ "max_steps": 969,
719
+ "num_input_tokens_seen": 0,
720
+ "num_train_epochs": 3,
721
+ "save_steps": 500,
722
+ "stateful_callbacks": {
723
+ "TrainerControl": {
724
+ "args": {
725
+ "should_epoch_stop": false,
726
+ "should_evaluate": false,
727
+ "should_log": false,
728
+ "should_save": true,
729
+ "should_training_stop": true
730
+ },
731
+ "attributes": {}
732
+ }
733
+ },
734
+ "total_flos": 1623111090831360.0,
735
+ "train_batch_size": 16,
736
+ "trial_name": null,
737
+ "trial_params": null
738
+ }
training_eval_loss.png ADDED
training_loss.png ADDED