sedrickkeh commited on
Commit
52ff20f
·
verified ·
1 Parent(s): 5aaa70a

End of training

Browse files
README.md CHANGED
@@ -4,6 +4,7 @@ license: gemma
4
  base_model: google/gemma-2-9b
5
  tags:
6
  - llama-factory
 
7
  - generated_from_trainer
8
  model-index:
9
  - name: hp_ablations_gemma_bsz512_dcftv1.2
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  # hp_ablations_gemma_bsz512_dcftv1.2
17
 
18
- This model is a fine-tuned version of [google/gemma-2-9b](https://huggingface.co/google/gemma-2-9b) on an unknown dataset.
19
  It achieves the following results on the evaluation set:
20
  - Loss: 0.6347
21
 
 
4
  base_model: google/gemma-2-9b
5
  tags:
6
  - llama-factory
7
+ - full
8
  - generated_from_trainer
9
  model-index:
10
  - name: hp_ablations_gemma_bsz512_dcftv1.2
 
16
 
17
  # hp_ablations_gemma_bsz512_dcftv1.2
18
 
19
+ This model is a fine-tuned version of [google/gemma-2-9b](https://huggingface.co/google/gemma-2-9b) 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: 0.6347
22
 
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.999438727782975,
3
+ "eval_loss": 0.6347318887710571,
4
+ "eval_runtime": 516.1922,
5
+ "eval_samples_per_second": 17.437,
6
+ "eval_steps_per_second": 0.546,
7
+ "total_flos": 3818092983484416.0,
8
+ "train_loss": 0.5628350560298699,
9
+ "train_runtime": 90903.3937,
10
+ "train_samples_per_second": 5.644,
11
+ "train_steps_per_second": 0.011
12
+ }
eval_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.999438727782975,
3
+ "eval_loss": 0.6347318887710571,
4
+ "eval_runtime": 516.1922,
5
+ "eval_samples_per_second": 17.437,
6
+ "eval_steps_per_second": 0.546
7
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.999438727782975,
3
+ "total_flos": 3818092983484416.0,
4
+ "train_loss": 0.5628350560298699,
5
+ "train_runtime": 90903.3937,
6
+ "train_samples_per_second": 5.644,
7
+ "train_steps_per_second": 0.011
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,766 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.999438727782975,
5
+ "eval_steps": 500,
6
+ "global_step": 1002,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.029934518241347054,
13
+ "grad_norm": 1.3387344639766428,
14
+ "learning_rate": 5e-06,
15
+ "loss": 0.7726,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.05986903648269411,
20
+ "grad_norm": 0.7768397662301797,
21
+ "learning_rate": 5e-06,
22
+ "loss": 0.6941,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.08980355472404115,
27
+ "grad_norm": 0.707220098018675,
28
+ "learning_rate": 5e-06,
29
+ "loss": 0.669,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.11973807296538821,
34
+ "grad_norm": 0.6673639225008394,
35
+ "learning_rate": 5e-06,
36
+ "loss": 0.6651,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.14967259120673526,
41
+ "grad_norm": 0.931884409826821,
42
+ "learning_rate": 5e-06,
43
+ "loss": 0.6563,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.1796071094480823,
48
+ "grad_norm": 0.7248290568232586,
49
+ "learning_rate": 5e-06,
50
+ "loss": 0.6486,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.20954162768942938,
55
+ "grad_norm": 0.7317021024341339,
56
+ "learning_rate": 5e-06,
57
+ "loss": 0.6456,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.23947614593077643,
62
+ "grad_norm": 0.7878073380412045,
63
+ "learning_rate": 5e-06,
64
+ "loss": 0.647,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.2694106641721235,
69
+ "grad_norm": 0.9477711469098681,
70
+ "learning_rate": 5e-06,
71
+ "loss": 0.638,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.2993451824134705,
76
+ "grad_norm": 0.7824185434647374,
77
+ "learning_rate": 5e-06,
78
+ "loss": 0.6395,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.3292797006548176,
83
+ "grad_norm": 0.7691665877451451,
84
+ "learning_rate": 5e-06,
85
+ "loss": 0.6352,
86
+ "step": 110
87
+ },
88
+ {
89
+ "epoch": 0.3592142188961646,
90
+ "grad_norm": 0.85227640579011,
91
+ "learning_rate": 5e-06,
92
+ "loss": 0.6344,
93
+ "step": 120
94
+ },
95
+ {
96
+ "epoch": 0.3891487371375117,
97
+ "grad_norm": 0.7349060200879011,
98
+ "learning_rate": 5e-06,
99
+ "loss": 0.6343,
100
+ "step": 130
101
+ },
102
+ {
103
+ "epoch": 0.41908325537885877,
104
+ "grad_norm": 0.7948303412589806,
105
+ "learning_rate": 5e-06,
106
+ "loss": 0.628,
107
+ "step": 140
108
+ },
109
+ {
110
+ "epoch": 0.4490177736202058,
111
+ "grad_norm": 0.6403554998368226,
112
+ "learning_rate": 5e-06,
113
+ "loss": 0.6296,
114
+ "step": 150
115
+ },
116
+ {
117
+ "epoch": 0.47895229186155286,
118
+ "grad_norm": 0.6579875234047781,
119
+ "learning_rate": 5e-06,
120
+ "loss": 0.6286,
121
+ "step": 160
122
+ },
123
+ {
124
+ "epoch": 0.5088868101028999,
125
+ "grad_norm": 0.6635988202382727,
126
+ "learning_rate": 5e-06,
127
+ "loss": 0.6228,
128
+ "step": 170
129
+ },
130
+ {
131
+ "epoch": 0.538821328344247,
132
+ "grad_norm": 0.6766189913752628,
133
+ "learning_rate": 5e-06,
134
+ "loss": 0.6242,
135
+ "step": 180
136
+ },
137
+ {
138
+ "epoch": 0.568755846585594,
139
+ "grad_norm": 0.7350593528596518,
140
+ "learning_rate": 5e-06,
141
+ "loss": 0.6221,
142
+ "step": 190
143
+ },
144
+ {
145
+ "epoch": 0.598690364826941,
146
+ "grad_norm": 0.7042346638489535,
147
+ "learning_rate": 5e-06,
148
+ "loss": 0.6299,
149
+ "step": 200
150
+ },
151
+ {
152
+ "epoch": 0.6286248830682881,
153
+ "grad_norm": 0.7310808118757001,
154
+ "learning_rate": 5e-06,
155
+ "loss": 0.6238,
156
+ "step": 210
157
+ },
158
+ {
159
+ "epoch": 0.6585594013096352,
160
+ "grad_norm": 0.915027325723907,
161
+ "learning_rate": 5e-06,
162
+ "loss": 0.622,
163
+ "step": 220
164
+ },
165
+ {
166
+ "epoch": 0.6884939195509823,
167
+ "grad_norm": 0.7146070049162939,
168
+ "learning_rate": 5e-06,
169
+ "loss": 0.6245,
170
+ "step": 230
171
+ },
172
+ {
173
+ "epoch": 0.7184284377923292,
174
+ "grad_norm": 0.6682750595418552,
175
+ "learning_rate": 5e-06,
176
+ "loss": 0.6234,
177
+ "step": 240
178
+ },
179
+ {
180
+ "epoch": 0.7483629560336763,
181
+ "grad_norm": 0.6075185838844738,
182
+ "learning_rate": 5e-06,
183
+ "loss": 0.6199,
184
+ "step": 250
185
+ },
186
+ {
187
+ "epoch": 0.7782974742750234,
188
+ "grad_norm": 0.6614005815451841,
189
+ "learning_rate": 5e-06,
190
+ "loss": 0.6186,
191
+ "step": 260
192
+ },
193
+ {
194
+ "epoch": 0.8082319925163705,
195
+ "grad_norm": 0.7431404889886319,
196
+ "learning_rate": 5e-06,
197
+ "loss": 0.6261,
198
+ "step": 270
199
+ },
200
+ {
201
+ "epoch": 0.8381665107577175,
202
+ "grad_norm": 0.6659051677024307,
203
+ "learning_rate": 5e-06,
204
+ "loss": 0.6163,
205
+ "step": 280
206
+ },
207
+ {
208
+ "epoch": 0.8681010289990645,
209
+ "grad_norm": 0.761848043492867,
210
+ "learning_rate": 5e-06,
211
+ "loss": 0.6131,
212
+ "step": 290
213
+ },
214
+ {
215
+ "epoch": 0.8980355472404116,
216
+ "grad_norm": 0.7790121000018401,
217
+ "learning_rate": 5e-06,
218
+ "loss": 0.6145,
219
+ "step": 300
220
+ },
221
+ {
222
+ "epoch": 0.9279700654817586,
223
+ "grad_norm": 0.700122187157476,
224
+ "learning_rate": 5e-06,
225
+ "loss": 0.6131,
226
+ "step": 310
227
+ },
228
+ {
229
+ "epoch": 0.9579045837231057,
230
+ "grad_norm": 0.6439594670749554,
231
+ "learning_rate": 5e-06,
232
+ "loss": 0.6152,
233
+ "step": 320
234
+ },
235
+ {
236
+ "epoch": 0.9878391019644528,
237
+ "grad_norm": 0.9505341757424101,
238
+ "learning_rate": 5e-06,
239
+ "loss": 0.6082,
240
+ "step": 330
241
+ },
242
+ {
243
+ "epoch": 0.9998129092609915,
244
+ "eval_loss": 0.619162917137146,
245
+ "eval_runtime": 516.7234,
246
+ "eval_samples_per_second": 17.419,
247
+ "eval_steps_per_second": 0.546,
248
+ "step": 334
249
+ },
250
+ {
251
+ "epoch": 1.0177736202057999,
252
+ "grad_norm": 1.0078817159156914,
253
+ "learning_rate": 5e-06,
254
+ "loss": 0.6383,
255
+ "step": 340
256
+ },
257
+ {
258
+ "epoch": 1.047708138447147,
259
+ "grad_norm": 0.9933360913511419,
260
+ "learning_rate": 5e-06,
261
+ "loss": 0.5523,
262
+ "step": 350
263
+ },
264
+ {
265
+ "epoch": 1.077642656688494,
266
+ "grad_norm": 0.7284101974086702,
267
+ "learning_rate": 5e-06,
268
+ "loss": 0.5478,
269
+ "step": 360
270
+ },
271
+ {
272
+ "epoch": 1.1075771749298409,
273
+ "grad_norm": 0.8279225421843056,
274
+ "learning_rate": 5e-06,
275
+ "loss": 0.5481,
276
+ "step": 370
277
+ },
278
+ {
279
+ "epoch": 1.137511693171188,
280
+ "grad_norm": 0.7632668730408979,
281
+ "learning_rate": 5e-06,
282
+ "loss": 0.5484,
283
+ "step": 380
284
+ },
285
+ {
286
+ "epoch": 1.167446211412535,
287
+ "grad_norm": 0.6939352572415957,
288
+ "learning_rate": 5e-06,
289
+ "loss": 0.5501,
290
+ "step": 390
291
+ },
292
+ {
293
+ "epoch": 1.197380729653882,
294
+ "grad_norm": 0.702668433385052,
295
+ "learning_rate": 5e-06,
296
+ "loss": 0.5501,
297
+ "step": 400
298
+ },
299
+ {
300
+ "epoch": 1.2273152478952292,
301
+ "grad_norm": 0.6478605176631643,
302
+ "learning_rate": 5e-06,
303
+ "loss": 0.5524,
304
+ "step": 410
305
+ },
306
+ {
307
+ "epoch": 1.2572497661365762,
308
+ "grad_norm": 0.7369685315266656,
309
+ "learning_rate": 5e-06,
310
+ "loss": 0.5567,
311
+ "step": 420
312
+ },
313
+ {
314
+ "epoch": 1.2871842843779233,
315
+ "grad_norm": 0.6753735200396593,
316
+ "learning_rate": 5e-06,
317
+ "loss": 0.5514,
318
+ "step": 430
319
+ },
320
+ {
321
+ "epoch": 1.3171188026192704,
322
+ "grad_norm": 0.6984955164701175,
323
+ "learning_rate": 5e-06,
324
+ "loss": 0.5534,
325
+ "step": 440
326
+ },
327
+ {
328
+ "epoch": 1.3470533208606175,
329
+ "grad_norm": 0.6720392032007318,
330
+ "learning_rate": 5e-06,
331
+ "loss": 0.5475,
332
+ "step": 450
333
+ },
334
+ {
335
+ "epoch": 1.3769878391019645,
336
+ "grad_norm": 0.6876544426761866,
337
+ "learning_rate": 5e-06,
338
+ "loss": 0.5581,
339
+ "step": 460
340
+ },
341
+ {
342
+ "epoch": 1.4069223573433116,
343
+ "grad_norm": 0.7559712718072091,
344
+ "learning_rate": 5e-06,
345
+ "loss": 0.5573,
346
+ "step": 470
347
+ },
348
+ {
349
+ "epoch": 1.4368568755846587,
350
+ "grad_norm": 0.7543483123276029,
351
+ "learning_rate": 5e-06,
352
+ "loss": 0.5592,
353
+ "step": 480
354
+ },
355
+ {
356
+ "epoch": 1.4667913938260055,
357
+ "grad_norm": 0.670230242807614,
358
+ "learning_rate": 5e-06,
359
+ "loss": 0.5587,
360
+ "step": 490
361
+ },
362
+ {
363
+ "epoch": 1.4967259120673526,
364
+ "grad_norm": 0.793302321148842,
365
+ "learning_rate": 5e-06,
366
+ "loss": 0.5549,
367
+ "step": 500
368
+ },
369
+ {
370
+ "epoch": 1.5266604303086997,
371
+ "grad_norm": 0.6636753856100917,
372
+ "learning_rate": 5e-06,
373
+ "loss": 0.5577,
374
+ "step": 510
375
+ },
376
+ {
377
+ "epoch": 1.5565949485500468,
378
+ "grad_norm": 0.6821579825224984,
379
+ "learning_rate": 5e-06,
380
+ "loss": 0.5557,
381
+ "step": 520
382
+ },
383
+ {
384
+ "epoch": 1.5865294667913938,
385
+ "grad_norm": 0.7552762189905546,
386
+ "learning_rate": 5e-06,
387
+ "loss": 0.5563,
388
+ "step": 530
389
+ },
390
+ {
391
+ "epoch": 1.616463985032741,
392
+ "grad_norm": 0.6729733301722461,
393
+ "learning_rate": 5e-06,
394
+ "loss": 0.5599,
395
+ "step": 540
396
+ },
397
+ {
398
+ "epoch": 1.646398503274088,
399
+ "grad_norm": 0.631104770625896,
400
+ "learning_rate": 5e-06,
401
+ "loss": 0.5593,
402
+ "step": 550
403
+ },
404
+ {
405
+ "epoch": 1.6763330215154348,
406
+ "grad_norm": 0.6474672113131026,
407
+ "learning_rate": 5e-06,
408
+ "loss": 0.5499,
409
+ "step": 560
410
+ },
411
+ {
412
+ "epoch": 1.706267539756782,
413
+ "grad_norm": 0.8000532776371828,
414
+ "learning_rate": 5e-06,
415
+ "loss": 0.5517,
416
+ "step": 570
417
+ },
418
+ {
419
+ "epoch": 1.736202057998129,
420
+ "grad_norm": 0.7694340841619645,
421
+ "learning_rate": 5e-06,
422
+ "loss": 0.5529,
423
+ "step": 580
424
+ },
425
+ {
426
+ "epoch": 1.766136576239476,
427
+ "grad_norm": 0.6621277045994409,
428
+ "learning_rate": 5e-06,
429
+ "loss": 0.5558,
430
+ "step": 590
431
+ },
432
+ {
433
+ "epoch": 1.7960710944808231,
434
+ "grad_norm": 0.715383547152723,
435
+ "learning_rate": 5e-06,
436
+ "loss": 0.5565,
437
+ "step": 600
438
+ },
439
+ {
440
+ "epoch": 1.8260056127221702,
441
+ "grad_norm": 0.625130733885464,
442
+ "learning_rate": 5e-06,
443
+ "loss": 0.5542,
444
+ "step": 610
445
+ },
446
+ {
447
+ "epoch": 1.8559401309635173,
448
+ "grad_norm": 0.6823486917509712,
449
+ "learning_rate": 5e-06,
450
+ "loss": 0.5524,
451
+ "step": 620
452
+ },
453
+ {
454
+ "epoch": 1.8858746492048644,
455
+ "grad_norm": 0.6881561256441558,
456
+ "learning_rate": 5e-06,
457
+ "loss": 0.5542,
458
+ "step": 630
459
+ },
460
+ {
461
+ "epoch": 1.9158091674462114,
462
+ "grad_norm": 0.7622942086873891,
463
+ "learning_rate": 5e-06,
464
+ "loss": 0.5557,
465
+ "step": 640
466
+ },
467
+ {
468
+ "epoch": 1.9457436856875585,
469
+ "grad_norm": 0.665050257763723,
470
+ "learning_rate": 5e-06,
471
+ "loss": 0.5539,
472
+ "step": 650
473
+ },
474
+ {
475
+ "epoch": 1.9756782039289056,
476
+ "grad_norm": 0.7758483897762515,
477
+ "learning_rate": 5e-06,
478
+ "loss": 0.5587,
479
+ "step": 660
480
+ },
481
+ {
482
+ "epoch": 1.999625818521983,
483
+ "eval_loss": 0.6155834197998047,
484
+ "eval_runtime": 516.9866,
485
+ "eval_samples_per_second": 17.411,
486
+ "eval_steps_per_second": 0.545,
487
+ "step": 668
488
+ },
489
+ {
490
+ "epoch": 2.0056127221702527,
491
+ "grad_norm": 1.1722347519510599,
492
+ "learning_rate": 5e-06,
493
+ "loss": 0.6006,
494
+ "step": 670
495
+ },
496
+ {
497
+ "epoch": 2.0355472404115997,
498
+ "grad_norm": 0.8414624124662551,
499
+ "learning_rate": 5e-06,
500
+ "loss": 0.4846,
501
+ "step": 680
502
+ },
503
+ {
504
+ "epoch": 2.065481758652947,
505
+ "grad_norm": 0.6931245660591268,
506
+ "learning_rate": 5e-06,
507
+ "loss": 0.4808,
508
+ "step": 690
509
+ },
510
+ {
511
+ "epoch": 2.095416276894294,
512
+ "grad_norm": 0.7117529793838588,
513
+ "learning_rate": 5e-06,
514
+ "loss": 0.4855,
515
+ "step": 700
516
+ },
517
+ {
518
+ "epoch": 2.125350795135641,
519
+ "grad_norm": 0.7331820893440362,
520
+ "learning_rate": 5e-06,
521
+ "loss": 0.4858,
522
+ "step": 710
523
+ },
524
+ {
525
+ "epoch": 2.155285313376988,
526
+ "grad_norm": 0.8754972222725035,
527
+ "learning_rate": 5e-06,
528
+ "loss": 0.4893,
529
+ "step": 720
530
+ },
531
+ {
532
+ "epoch": 2.185219831618335,
533
+ "grad_norm": 0.7192523205414377,
534
+ "learning_rate": 5e-06,
535
+ "loss": 0.4894,
536
+ "step": 730
537
+ },
538
+ {
539
+ "epoch": 2.2151543498596817,
540
+ "grad_norm": 0.7815473273248343,
541
+ "learning_rate": 5e-06,
542
+ "loss": 0.49,
543
+ "step": 740
544
+ },
545
+ {
546
+ "epoch": 2.245088868101029,
547
+ "grad_norm": 0.7224517891086425,
548
+ "learning_rate": 5e-06,
549
+ "loss": 0.4925,
550
+ "step": 750
551
+ },
552
+ {
553
+ "epoch": 2.275023386342376,
554
+ "grad_norm": 0.6941146001457271,
555
+ "learning_rate": 5e-06,
556
+ "loss": 0.4882,
557
+ "step": 760
558
+ },
559
+ {
560
+ "epoch": 2.304957904583723,
561
+ "grad_norm": 0.7516859140116606,
562
+ "learning_rate": 5e-06,
563
+ "loss": 0.4935,
564
+ "step": 770
565
+ },
566
+ {
567
+ "epoch": 2.33489242282507,
568
+ "grad_norm": 0.6995328412478509,
569
+ "learning_rate": 5e-06,
570
+ "loss": 0.4892,
571
+ "step": 780
572
+ },
573
+ {
574
+ "epoch": 2.364826941066417,
575
+ "grad_norm": 0.6450052032271214,
576
+ "learning_rate": 5e-06,
577
+ "loss": 0.4909,
578
+ "step": 790
579
+ },
580
+ {
581
+ "epoch": 2.394761459307764,
582
+ "grad_norm": 0.7383031235734733,
583
+ "learning_rate": 5e-06,
584
+ "loss": 0.4936,
585
+ "step": 800
586
+ },
587
+ {
588
+ "epoch": 2.4246959775491113,
589
+ "grad_norm": 0.9052311314704138,
590
+ "learning_rate": 5e-06,
591
+ "loss": 0.5006,
592
+ "step": 810
593
+ },
594
+ {
595
+ "epoch": 2.4546304957904583,
596
+ "grad_norm": 0.744817569114051,
597
+ "learning_rate": 5e-06,
598
+ "loss": 0.4978,
599
+ "step": 820
600
+ },
601
+ {
602
+ "epoch": 2.4845650140318054,
603
+ "grad_norm": 0.8333667937918693,
604
+ "learning_rate": 5e-06,
605
+ "loss": 0.4976,
606
+ "step": 830
607
+ },
608
+ {
609
+ "epoch": 2.5144995322731525,
610
+ "grad_norm": 0.7032740125894698,
611
+ "learning_rate": 5e-06,
612
+ "loss": 0.4947,
613
+ "step": 840
614
+ },
615
+ {
616
+ "epoch": 2.5444340505144996,
617
+ "grad_norm": 0.7332627913564072,
618
+ "learning_rate": 5e-06,
619
+ "loss": 0.4984,
620
+ "step": 850
621
+ },
622
+ {
623
+ "epoch": 2.5743685687558466,
624
+ "grad_norm": 0.7267448308374836,
625
+ "learning_rate": 5e-06,
626
+ "loss": 0.4964,
627
+ "step": 860
628
+ },
629
+ {
630
+ "epoch": 2.6043030869971937,
631
+ "grad_norm": 0.6783409226329004,
632
+ "learning_rate": 5e-06,
633
+ "loss": 0.4941,
634
+ "step": 870
635
+ },
636
+ {
637
+ "epoch": 2.634237605238541,
638
+ "grad_norm": 0.7177369341898473,
639
+ "learning_rate": 5e-06,
640
+ "loss": 0.4922,
641
+ "step": 880
642
+ },
643
+ {
644
+ "epoch": 2.664172123479888,
645
+ "grad_norm": 0.7220986253572231,
646
+ "learning_rate": 5e-06,
647
+ "loss": 0.4983,
648
+ "step": 890
649
+ },
650
+ {
651
+ "epoch": 2.694106641721235,
652
+ "grad_norm": 0.6703622867889978,
653
+ "learning_rate": 5e-06,
654
+ "loss": 0.501,
655
+ "step": 900
656
+ },
657
+ {
658
+ "epoch": 2.724041159962582,
659
+ "grad_norm": 0.6964496953277525,
660
+ "learning_rate": 5e-06,
661
+ "loss": 0.4944,
662
+ "step": 910
663
+ },
664
+ {
665
+ "epoch": 2.753975678203929,
666
+ "grad_norm": 0.6590985545988255,
667
+ "learning_rate": 5e-06,
668
+ "loss": 0.4957,
669
+ "step": 920
670
+ },
671
+ {
672
+ "epoch": 2.7839101964452757,
673
+ "grad_norm": 0.6904019401527591,
674
+ "learning_rate": 5e-06,
675
+ "loss": 0.5015,
676
+ "step": 930
677
+ },
678
+ {
679
+ "epoch": 2.8138447146866232,
680
+ "grad_norm": 0.6886972345963736,
681
+ "learning_rate": 5e-06,
682
+ "loss": 0.4996,
683
+ "step": 940
684
+ },
685
+ {
686
+ "epoch": 2.84377923292797,
687
+ "grad_norm": 0.699334830976579,
688
+ "learning_rate": 5e-06,
689
+ "loss": 0.4991,
690
+ "step": 950
691
+ },
692
+ {
693
+ "epoch": 2.8737137511693174,
694
+ "grad_norm": 0.8229628219796593,
695
+ "learning_rate": 5e-06,
696
+ "loss": 0.5011,
697
+ "step": 960
698
+ },
699
+ {
700
+ "epoch": 2.903648269410664,
701
+ "grad_norm": 0.773910899786262,
702
+ "learning_rate": 5e-06,
703
+ "loss": 0.4967,
704
+ "step": 970
705
+ },
706
+ {
707
+ "epoch": 2.933582787652011,
708
+ "grad_norm": 0.7347626406072614,
709
+ "learning_rate": 5e-06,
710
+ "loss": 0.5024,
711
+ "step": 980
712
+ },
713
+ {
714
+ "epoch": 2.963517305893358,
715
+ "grad_norm": 0.7149658597566454,
716
+ "learning_rate": 5e-06,
717
+ "loss": 0.504,
718
+ "step": 990
719
+ },
720
+ {
721
+ "epoch": 2.9934518241347052,
722
+ "grad_norm": 0.6938231808502269,
723
+ "learning_rate": 5e-06,
724
+ "loss": 0.505,
725
+ "step": 1000
726
+ },
727
+ {
728
+ "epoch": 2.999438727782975,
729
+ "eval_loss": 0.6347318887710571,
730
+ "eval_runtime": 516.7784,
731
+ "eval_samples_per_second": 17.418,
732
+ "eval_steps_per_second": 0.546,
733
+ "step": 1002
734
+ },
735
+ {
736
+ "epoch": 2.999438727782975,
737
+ "step": 1002,
738
+ "total_flos": 3818092983484416.0,
739
+ "train_loss": 0.5628350560298699,
740
+ "train_runtime": 90903.3937,
741
+ "train_samples_per_second": 5.644,
742
+ "train_steps_per_second": 0.011
743
+ }
744
+ ],
745
+ "logging_steps": 10,
746
+ "max_steps": 1002,
747
+ "num_input_tokens_seen": 0,
748
+ "num_train_epochs": 3,
749
+ "save_steps": 500,
750
+ "stateful_callbacks": {
751
+ "TrainerControl": {
752
+ "args": {
753
+ "should_epoch_stop": false,
754
+ "should_evaluate": false,
755
+ "should_log": false,
756
+ "should_save": true,
757
+ "should_training_stop": true
758
+ },
759
+ "attributes": {}
760
+ }
761
+ },
762
+ "total_flos": 3818092983484416.0,
763
+ "train_batch_size": 4,
764
+ "trial_name": null,
765
+ "trial_params": null
766
+ }
training_eval_loss.png ADDED
training_loss.png ADDED