gsmyrnis commited on
Commit
9d13c68
1 Parent(s): ebd2a54

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-v3.1-llama-3.1-405b
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  # oh-dcft-v3.1-llama-3.1-405b
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: 0.3009
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-v3.1-llama-3.1-405b
 
16
 
17
  # oh-dcft-v3.1-llama-3.1-405b
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-v3.1-llama-3.1-405b dataset.
20
  It achieves the following results on the evaluation set:
21
  - Loss: 0.3009
22
 
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_loss": 0.30088362097740173,
4
+ "eval_runtime": 40.4677,
5
+ "eval_samples_per_second": 280.817,
6
+ "eval_steps_per_second": 1.112,
7
+ "total_flos": 2120597152727040.0,
8
+ "train_loss": 0.2836323574444317,
9
+ "train_runtime": 7986.5453,
10
+ "train_samples_per_second": 81.101,
11
+ "train_steps_per_second": 0.159
12
+ }
eval_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "eval_loss": 0.30088362097740173,
4
+ "eval_runtime": 40.4677,
5
+ "eval_samples_per_second": 280.817,
6
+ "eval_steps_per_second": 1.112
7
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.0,
3
+ "total_flos": 2120597152727040.0,
4
+ "train_loss": 0.2836323574444317,
5
+ "train_runtime": 7986.5453,
6
+ "train_samples_per_second": 81.101,
7
+ "train_steps_per_second": 0.159
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,948 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "eval_steps": 500,
6
+ "global_step": 1266,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.023696682464454975,
13
+ "grad_norm": 30.705133295132807,
14
+ "learning_rate": 5e-06,
15
+ "loss": 0.5563,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.04739336492890995,
20
+ "grad_norm": 2.043478075324726,
21
+ "learning_rate": 5e-06,
22
+ "loss": 0.4838,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.07109004739336493,
27
+ "grad_norm": 1.3358743818486332,
28
+ "learning_rate": 5e-06,
29
+ "loss": 0.4503,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.0947867298578199,
34
+ "grad_norm": 0.7538335151640748,
35
+ "learning_rate": 5e-06,
36
+ "loss": 0.429,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.11848341232227488,
41
+ "grad_norm": 0.7811530617140079,
42
+ "learning_rate": 5e-06,
43
+ "loss": 0.4108,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.14218009478672985,
48
+ "grad_norm": 0.9317427620124435,
49
+ "learning_rate": 5e-06,
50
+ "loss": 0.4018,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.16587677725118483,
55
+ "grad_norm": 0.6243367032731518,
56
+ "learning_rate": 5e-06,
57
+ "loss": 0.3908,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.1895734597156398,
62
+ "grad_norm": 0.6176921128719449,
63
+ "learning_rate": 5e-06,
64
+ "loss": 0.3833,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.2132701421800948,
69
+ "grad_norm": 0.4314255330259147,
70
+ "learning_rate": 5e-06,
71
+ "loss": 0.3777,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.23696682464454977,
76
+ "grad_norm": 0.5197854521683744,
77
+ "learning_rate": 5e-06,
78
+ "loss": 0.3704,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.26066350710900477,
83
+ "grad_norm": 1.002573805206526,
84
+ "learning_rate": 5e-06,
85
+ "loss": 0.3662,
86
+ "step": 110
87
+ },
88
+ {
89
+ "epoch": 0.2843601895734597,
90
+ "grad_norm": 0.5881621358357682,
91
+ "learning_rate": 5e-06,
92
+ "loss": 0.3658,
93
+ "step": 120
94
+ },
95
+ {
96
+ "epoch": 0.3080568720379147,
97
+ "grad_norm": 0.5217282145946449,
98
+ "learning_rate": 5e-06,
99
+ "loss": 0.3629,
100
+ "step": 130
101
+ },
102
+ {
103
+ "epoch": 0.33175355450236965,
104
+ "grad_norm": 0.6091442268483355,
105
+ "learning_rate": 5e-06,
106
+ "loss": 0.3652,
107
+ "step": 140
108
+ },
109
+ {
110
+ "epoch": 0.35545023696682465,
111
+ "grad_norm": 0.8005863736511483,
112
+ "learning_rate": 5e-06,
113
+ "loss": 0.3531,
114
+ "step": 150
115
+ },
116
+ {
117
+ "epoch": 0.3791469194312796,
118
+ "grad_norm": 0.6150361006740718,
119
+ "learning_rate": 5e-06,
120
+ "loss": 0.3498,
121
+ "step": 160
122
+ },
123
+ {
124
+ "epoch": 0.4028436018957346,
125
+ "grad_norm": 0.6278490509821739,
126
+ "learning_rate": 5e-06,
127
+ "loss": 0.3507,
128
+ "step": 170
129
+ },
130
+ {
131
+ "epoch": 0.4265402843601896,
132
+ "grad_norm": 0.8241804964154705,
133
+ "learning_rate": 5e-06,
134
+ "loss": 0.348,
135
+ "step": 180
136
+ },
137
+ {
138
+ "epoch": 0.45023696682464454,
139
+ "grad_norm": 0.6555844743575341,
140
+ "learning_rate": 5e-06,
141
+ "loss": 0.3476,
142
+ "step": 190
143
+ },
144
+ {
145
+ "epoch": 0.47393364928909953,
146
+ "grad_norm": 0.5422180070121905,
147
+ "learning_rate": 5e-06,
148
+ "loss": 0.3427,
149
+ "step": 200
150
+ },
151
+ {
152
+ "epoch": 0.4976303317535545,
153
+ "grad_norm": 0.5207523671441107,
154
+ "learning_rate": 5e-06,
155
+ "loss": 0.3447,
156
+ "step": 210
157
+ },
158
+ {
159
+ "epoch": 0.5213270142180095,
160
+ "grad_norm": 0.45313363674774493,
161
+ "learning_rate": 5e-06,
162
+ "loss": 0.3408,
163
+ "step": 220
164
+ },
165
+ {
166
+ "epoch": 0.5450236966824644,
167
+ "grad_norm": 0.5162603087798662,
168
+ "learning_rate": 5e-06,
169
+ "loss": 0.341,
170
+ "step": 230
171
+ },
172
+ {
173
+ "epoch": 0.5687203791469194,
174
+ "grad_norm": 0.46580112576483534,
175
+ "learning_rate": 5e-06,
176
+ "loss": 0.3335,
177
+ "step": 240
178
+ },
179
+ {
180
+ "epoch": 0.5924170616113744,
181
+ "grad_norm": 0.608794869219161,
182
+ "learning_rate": 5e-06,
183
+ "loss": 0.3345,
184
+ "step": 250
185
+ },
186
+ {
187
+ "epoch": 0.6161137440758294,
188
+ "grad_norm": 0.523854370493108,
189
+ "learning_rate": 5e-06,
190
+ "loss": 0.3308,
191
+ "step": 260
192
+ },
193
+ {
194
+ "epoch": 0.6398104265402843,
195
+ "grad_norm": 0.5706943490018692,
196
+ "learning_rate": 5e-06,
197
+ "loss": 0.3361,
198
+ "step": 270
199
+ },
200
+ {
201
+ "epoch": 0.6635071090047393,
202
+ "grad_norm": 0.45300138137727003,
203
+ "learning_rate": 5e-06,
204
+ "loss": 0.3326,
205
+ "step": 280
206
+ },
207
+ {
208
+ "epoch": 0.6872037914691943,
209
+ "grad_norm": 0.45102429427517104,
210
+ "learning_rate": 5e-06,
211
+ "loss": 0.3302,
212
+ "step": 290
213
+ },
214
+ {
215
+ "epoch": 0.7109004739336493,
216
+ "grad_norm": 0.5137700452326517,
217
+ "learning_rate": 5e-06,
218
+ "loss": 0.3289,
219
+ "step": 300
220
+ },
221
+ {
222
+ "epoch": 0.7345971563981043,
223
+ "grad_norm": 0.6127249666946868,
224
+ "learning_rate": 5e-06,
225
+ "loss": 0.3316,
226
+ "step": 310
227
+ },
228
+ {
229
+ "epoch": 0.7582938388625592,
230
+ "grad_norm": 0.44932409996775663,
231
+ "learning_rate": 5e-06,
232
+ "loss": 0.326,
233
+ "step": 320
234
+ },
235
+ {
236
+ "epoch": 0.7819905213270142,
237
+ "grad_norm": 0.5532587149282668,
238
+ "learning_rate": 5e-06,
239
+ "loss": 0.3296,
240
+ "step": 330
241
+ },
242
+ {
243
+ "epoch": 0.8056872037914692,
244
+ "grad_norm": 0.43366234423399613,
245
+ "learning_rate": 5e-06,
246
+ "loss": 0.3292,
247
+ "step": 340
248
+ },
249
+ {
250
+ "epoch": 0.8293838862559242,
251
+ "grad_norm": 0.6322058991355771,
252
+ "learning_rate": 5e-06,
253
+ "loss": 0.3294,
254
+ "step": 350
255
+ },
256
+ {
257
+ "epoch": 0.8530805687203792,
258
+ "grad_norm": 0.46695206519297977,
259
+ "learning_rate": 5e-06,
260
+ "loss": 0.3302,
261
+ "step": 360
262
+ },
263
+ {
264
+ "epoch": 0.8767772511848341,
265
+ "grad_norm": 0.5812611546801564,
266
+ "learning_rate": 5e-06,
267
+ "loss": 0.3273,
268
+ "step": 370
269
+ },
270
+ {
271
+ "epoch": 0.9004739336492891,
272
+ "grad_norm": 0.5621804321411954,
273
+ "learning_rate": 5e-06,
274
+ "loss": 0.3192,
275
+ "step": 380
276
+ },
277
+ {
278
+ "epoch": 0.9241706161137441,
279
+ "grad_norm": 0.4679602140385812,
280
+ "learning_rate": 5e-06,
281
+ "loss": 0.3214,
282
+ "step": 390
283
+ },
284
+ {
285
+ "epoch": 0.9478672985781991,
286
+ "grad_norm": 0.5333407829184633,
287
+ "learning_rate": 5e-06,
288
+ "loss": 0.3204,
289
+ "step": 400
290
+ },
291
+ {
292
+ "epoch": 0.9715639810426541,
293
+ "grad_norm": 0.43987965909268334,
294
+ "learning_rate": 5e-06,
295
+ "loss": 0.3214,
296
+ "step": 410
297
+ },
298
+ {
299
+ "epoch": 0.995260663507109,
300
+ "grad_norm": 0.6709211874509404,
301
+ "learning_rate": 5e-06,
302
+ "loss": 0.3182,
303
+ "step": 420
304
+ },
305
+ {
306
+ "epoch": 1.0,
307
+ "eval_loss": 0.3187981843948364,
308
+ "eval_runtime": 41.0925,
309
+ "eval_samples_per_second": 276.547,
310
+ "eval_steps_per_second": 1.095,
311
+ "step": 422
312
+ },
313
+ {
314
+ "epoch": 1.018957345971564,
315
+ "grad_norm": 0.5760447436307622,
316
+ "learning_rate": 5e-06,
317
+ "loss": 0.287,
318
+ "step": 430
319
+ },
320
+ {
321
+ "epoch": 1.042654028436019,
322
+ "grad_norm": 0.6762165069988788,
323
+ "learning_rate": 5e-06,
324
+ "loss": 0.2762,
325
+ "step": 440
326
+ },
327
+ {
328
+ "epoch": 1.066350710900474,
329
+ "grad_norm": 0.47402123924703626,
330
+ "learning_rate": 5e-06,
331
+ "loss": 0.2734,
332
+ "step": 450
333
+ },
334
+ {
335
+ "epoch": 1.0900473933649288,
336
+ "grad_norm": 0.5516286942729635,
337
+ "learning_rate": 5e-06,
338
+ "loss": 0.2769,
339
+ "step": 460
340
+ },
341
+ {
342
+ "epoch": 1.113744075829384,
343
+ "grad_norm": 0.5287265469429161,
344
+ "learning_rate": 5e-06,
345
+ "loss": 0.273,
346
+ "step": 470
347
+ },
348
+ {
349
+ "epoch": 1.1374407582938388,
350
+ "grad_norm": 0.6297802215905517,
351
+ "learning_rate": 5e-06,
352
+ "loss": 0.2763,
353
+ "step": 480
354
+ },
355
+ {
356
+ "epoch": 1.161137440758294,
357
+ "grad_norm": 0.5655517444516921,
358
+ "learning_rate": 5e-06,
359
+ "loss": 0.2779,
360
+ "step": 490
361
+ },
362
+ {
363
+ "epoch": 1.1848341232227488,
364
+ "grad_norm": 0.4960690379887489,
365
+ "learning_rate": 5e-06,
366
+ "loss": 0.2745,
367
+ "step": 500
368
+ },
369
+ {
370
+ "epoch": 1.2085308056872037,
371
+ "grad_norm": 0.4549491415898102,
372
+ "learning_rate": 5e-06,
373
+ "loss": 0.2754,
374
+ "step": 510
375
+ },
376
+ {
377
+ "epoch": 1.2322274881516588,
378
+ "grad_norm": 0.4752122430144407,
379
+ "learning_rate": 5e-06,
380
+ "loss": 0.2731,
381
+ "step": 520
382
+ },
383
+ {
384
+ "epoch": 1.2559241706161137,
385
+ "grad_norm": 0.4647993470821906,
386
+ "learning_rate": 5e-06,
387
+ "loss": 0.2739,
388
+ "step": 530
389
+ },
390
+ {
391
+ "epoch": 1.2796208530805688,
392
+ "grad_norm": 0.5109279201523487,
393
+ "learning_rate": 5e-06,
394
+ "loss": 0.2714,
395
+ "step": 540
396
+ },
397
+ {
398
+ "epoch": 1.3033175355450237,
399
+ "grad_norm": 0.5840908685437068,
400
+ "learning_rate": 5e-06,
401
+ "loss": 0.2737,
402
+ "step": 550
403
+ },
404
+ {
405
+ "epoch": 1.3270142180094786,
406
+ "grad_norm": 0.5246063574530767,
407
+ "learning_rate": 5e-06,
408
+ "loss": 0.2751,
409
+ "step": 560
410
+ },
411
+ {
412
+ "epoch": 1.3507109004739337,
413
+ "grad_norm": 0.44184294378239247,
414
+ "learning_rate": 5e-06,
415
+ "loss": 0.2745,
416
+ "step": 570
417
+ },
418
+ {
419
+ "epoch": 1.3744075829383886,
420
+ "grad_norm": 0.5913058737082955,
421
+ "learning_rate": 5e-06,
422
+ "loss": 0.2734,
423
+ "step": 580
424
+ },
425
+ {
426
+ "epoch": 1.3981042654028437,
427
+ "grad_norm": 0.5008265601868412,
428
+ "learning_rate": 5e-06,
429
+ "loss": 0.2738,
430
+ "step": 590
431
+ },
432
+ {
433
+ "epoch": 1.4218009478672986,
434
+ "grad_norm": 0.49995889605138555,
435
+ "learning_rate": 5e-06,
436
+ "loss": 0.2744,
437
+ "step": 600
438
+ },
439
+ {
440
+ "epoch": 1.4454976303317535,
441
+ "grad_norm": 0.45557075455533524,
442
+ "learning_rate": 5e-06,
443
+ "loss": 0.2721,
444
+ "step": 610
445
+ },
446
+ {
447
+ "epoch": 1.4691943127962086,
448
+ "grad_norm": 0.6379507686940746,
449
+ "learning_rate": 5e-06,
450
+ "loss": 0.2732,
451
+ "step": 620
452
+ },
453
+ {
454
+ "epoch": 1.4928909952606635,
455
+ "grad_norm": 0.4583258359143169,
456
+ "learning_rate": 5e-06,
457
+ "loss": 0.2725,
458
+ "step": 630
459
+ },
460
+ {
461
+ "epoch": 1.5165876777251186,
462
+ "grad_norm": 0.5749182699938892,
463
+ "learning_rate": 5e-06,
464
+ "loss": 0.2665,
465
+ "step": 640
466
+ },
467
+ {
468
+ "epoch": 1.5402843601895735,
469
+ "grad_norm": 0.5326772009600166,
470
+ "learning_rate": 5e-06,
471
+ "loss": 0.2697,
472
+ "step": 650
473
+ },
474
+ {
475
+ "epoch": 1.5639810426540284,
476
+ "grad_norm": 0.5654654670246291,
477
+ "learning_rate": 5e-06,
478
+ "loss": 0.2673,
479
+ "step": 660
480
+ },
481
+ {
482
+ "epoch": 1.5876777251184833,
483
+ "grad_norm": 0.7750526195773435,
484
+ "learning_rate": 5e-06,
485
+ "loss": 0.274,
486
+ "step": 670
487
+ },
488
+ {
489
+ "epoch": 1.6113744075829384,
490
+ "grad_norm": 0.5315008234518865,
491
+ "learning_rate": 5e-06,
492
+ "loss": 0.2702,
493
+ "step": 680
494
+ },
495
+ {
496
+ "epoch": 1.6350710900473935,
497
+ "grad_norm": 0.5749838515161616,
498
+ "learning_rate": 5e-06,
499
+ "loss": 0.2695,
500
+ "step": 690
501
+ },
502
+ {
503
+ "epoch": 1.6587677725118484,
504
+ "grad_norm": 0.5001274305498948,
505
+ "learning_rate": 5e-06,
506
+ "loss": 0.2686,
507
+ "step": 700
508
+ },
509
+ {
510
+ "epoch": 1.6824644549763033,
511
+ "grad_norm": 0.5123878344164223,
512
+ "learning_rate": 5e-06,
513
+ "loss": 0.2707,
514
+ "step": 710
515
+ },
516
+ {
517
+ "epoch": 1.7061611374407581,
518
+ "grad_norm": 0.5614887710303913,
519
+ "learning_rate": 5e-06,
520
+ "loss": 0.2669,
521
+ "step": 720
522
+ },
523
+ {
524
+ "epoch": 1.7298578199052133,
525
+ "grad_norm": 0.549426396684614,
526
+ "learning_rate": 5e-06,
527
+ "loss": 0.2658,
528
+ "step": 730
529
+ },
530
+ {
531
+ "epoch": 1.7535545023696684,
532
+ "grad_norm": 0.6106476140869552,
533
+ "learning_rate": 5e-06,
534
+ "loss": 0.2687,
535
+ "step": 740
536
+ },
537
+ {
538
+ "epoch": 1.7772511848341233,
539
+ "grad_norm": 0.44864012484374616,
540
+ "learning_rate": 5e-06,
541
+ "loss": 0.2704,
542
+ "step": 750
543
+ },
544
+ {
545
+ "epoch": 1.8009478672985781,
546
+ "grad_norm": 0.6259517201369036,
547
+ "learning_rate": 5e-06,
548
+ "loss": 0.2673,
549
+ "step": 760
550
+ },
551
+ {
552
+ "epoch": 1.824644549763033,
553
+ "grad_norm": 0.5150026298786189,
554
+ "learning_rate": 5e-06,
555
+ "loss": 0.2693,
556
+ "step": 770
557
+ },
558
+ {
559
+ "epoch": 1.8483412322274881,
560
+ "grad_norm": 0.5457767160105244,
561
+ "learning_rate": 5e-06,
562
+ "loss": 0.2652,
563
+ "step": 780
564
+ },
565
+ {
566
+ "epoch": 1.8720379146919433,
567
+ "grad_norm": 0.4576520211880369,
568
+ "learning_rate": 5e-06,
569
+ "loss": 0.264,
570
+ "step": 790
571
+ },
572
+ {
573
+ "epoch": 1.8957345971563981,
574
+ "grad_norm": 0.42780791505078525,
575
+ "learning_rate": 5e-06,
576
+ "loss": 0.2707,
577
+ "step": 800
578
+ },
579
+ {
580
+ "epoch": 1.919431279620853,
581
+ "grad_norm": 0.48992225159350095,
582
+ "learning_rate": 5e-06,
583
+ "loss": 0.2645,
584
+ "step": 810
585
+ },
586
+ {
587
+ "epoch": 1.943127962085308,
588
+ "grad_norm": 0.5410708029490111,
589
+ "learning_rate": 5e-06,
590
+ "loss": 0.2627,
591
+ "step": 820
592
+ },
593
+ {
594
+ "epoch": 1.966824644549763,
595
+ "grad_norm": 0.4336435751671359,
596
+ "learning_rate": 5e-06,
597
+ "loss": 0.2663,
598
+ "step": 830
599
+ },
600
+ {
601
+ "epoch": 1.9905213270142181,
602
+ "grad_norm": 0.43877727937635,
603
+ "learning_rate": 5e-06,
604
+ "loss": 0.2676,
605
+ "step": 840
606
+ },
607
+ {
608
+ "epoch": 2.0,
609
+ "eval_loss": 0.30145883560180664,
610
+ "eval_runtime": 41.7379,
611
+ "eval_samples_per_second": 272.27,
612
+ "eval_steps_per_second": 1.078,
613
+ "step": 844
614
+ },
615
+ {
616
+ "epoch": 2.014218009478673,
617
+ "grad_norm": 0.6524222083837387,
618
+ "learning_rate": 5e-06,
619
+ "loss": 0.2369,
620
+ "step": 850
621
+ },
622
+ {
623
+ "epoch": 2.037914691943128,
624
+ "grad_norm": 0.5787525251465172,
625
+ "learning_rate": 5e-06,
626
+ "loss": 0.2195,
627
+ "step": 860
628
+ },
629
+ {
630
+ "epoch": 2.061611374407583,
631
+ "grad_norm": 0.5756101918178816,
632
+ "learning_rate": 5e-06,
633
+ "loss": 0.2199,
634
+ "step": 870
635
+ },
636
+ {
637
+ "epoch": 2.085308056872038,
638
+ "grad_norm": 0.6236522012957664,
639
+ "learning_rate": 5e-06,
640
+ "loss": 0.2204,
641
+ "step": 880
642
+ },
643
+ {
644
+ "epoch": 2.109004739336493,
645
+ "grad_norm": 0.6419600675102759,
646
+ "learning_rate": 5e-06,
647
+ "loss": 0.2207,
648
+ "step": 890
649
+ },
650
+ {
651
+ "epoch": 2.132701421800948,
652
+ "grad_norm": 0.6691668192668253,
653
+ "learning_rate": 5e-06,
654
+ "loss": 0.2205,
655
+ "step": 900
656
+ },
657
+ {
658
+ "epoch": 2.156398104265403,
659
+ "grad_norm": 0.5662613639218179,
660
+ "learning_rate": 5e-06,
661
+ "loss": 0.2202,
662
+ "step": 910
663
+ },
664
+ {
665
+ "epoch": 2.1800947867298577,
666
+ "grad_norm": 0.5285961633074623,
667
+ "learning_rate": 5e-06,
668
+ "loss": 0.2213,
669
+ "step": 920
670
+ },
671
+ {
672
+ "epoch": 2.2037914691943126,
673
+ "grad_norm": 0.5182873734926402,
674
+ "learning_rate": 5e-06,
675
+ "loss": 0.2237,
676
+ "step": 930
677
+ },
678
+ {
679
+ "epoch": 2.227488151658768,
680
+ "grad_norm": 0.455467292575364,
681
+ "learning_rate": 5e-06,
682
+ "loss": 0.2196,
683
+ "step": 940
684
+ },
685
+ {
686
+ "epoch": 2.251184834123223,
687
+ "grad_norm": 0.5910920449175188,
688
+ "learning_rate": 5e-06,
689
+ "loss": 0.2213,
690
+ "step": 950
691
+ },
692
+ {
693
+ "epoch": 2.2748815165876777,
694
+ "grad_norm": 0.48913238274264176,
695
+ "learning_rate": 5e-06,
696
+ "loss": 0.22,
697
+ "step": 960
698
+ },
699
+ {
700
+ "epoch": 2.2985781990521326,
701
+ "grad_norm": 0.48506465561903783,
702
+ "learning_rate": 5e-06,
703
+ "loss": 0.2209,
704
+ "step": 970
705
+ },
706
+ {
707
+ "epoch": 2.322274881516588,
708
+ "grad_norm": 0.4777608016950058,
709
+ "learning_rate": 5e-06,
710
+ "loss": 0.225,
711
+ "step": 980
712
+ },
713
+ {
714
+ "epoch": 2.345971563981043,
715
+ "grad_norm": 0.5663901860164571,
716
+ "learning_rate": 5e-06,
717
+ "loss": 0.2236,
718
+ "step": 990
719
+ },
720
+ {
721
+ "epoch": 2.3696682464454977,
722
+ "grad_norm": 0.5218376695680845,
723
+ "learning_rate": 5e-06,
724
+ "loss": 0.2216,
725
+ "step": 1000
726
+ },
727
+ {
728
+ "epoch": 2.3933649289099526,
729
+ "grad_norm": 0.5699143314581089,
730
+ "learning_rate": 5e-06,
731
+ "loss": 0.2233,
732
+ "step": 1010
733
+ },
734
+ {
735
+ "epoch": 2.4170616113744074,
736
+ "grad_norm": 0.4742020161888442,
737
+ "learning_rate": 5e-06,
738
+ "loss": 0.2228,
739
+ "step": 1020
740
+ },
741
+ {
742
+ "epoch": 2.4407582938388623,
743
+ "grad_norm": 0.5183880143591837,
744
+ "learning_rate": 5e-06,
745
+ "loss": 0.2226,
746
+ "step": 1030
747
+ },
748
+ {
749
+ "epoch": 2.4644549763033177,
750
+ "grad_norm": 0.6480574727941906,
751
+ "learning_rate": 5e-06,
752
+ "loss": 0.2211,
753
+ "step": 1040
754
+ },
755
+ {
756
+ "epoch": 2.4881516587677726,
757
+ "grad_norm": 0.5046467998532,
758
+ "learning_rate": 5e-06,
759
+ "loss": 0.2225,
760
+ "step": 1050
761
+ },
762
+ {
763
+ "epoch": 2.5118483412322274,
764
+ "grad_norm": 0.6234765711742589,
765
+ "learning_rate": 5e-06,
766
+ "loss": 0.2207,
767
+ "step": 1060
768
+ },
769
+ {
770
+ "epoch": 2.5355450236966823,
771
+ "grad_norm": 0.5463355131695685,
772
+ "learning_rate": 5e-06,
773
+ "loss": 0.2224,
774
+ "step": 1070
775
+ },
776
+ {
777
+ "epoch": 2.5592417061611377,
778
+ "grad_norm": 0.5045207519726893,
779
+ "learning_rate": 5e-06,
780
+ "loss": 0.2193,
781
+ "step": 1080
782
+ },
783
+ {
784
+ "epoch": 2.5829383886255926,
785
+ "grad_norm": 0.5385436231173875,
786
+ "learning_rate": 5e-06,
787
+ "loss": 0.2215,
788
+ "step": 1090
789
+ },
790
+ {
791
+ "epoch": 2.6066350710900474,
792
+ "grad_norm": 0.5203528739245922,
793
+ "learning_rate": 5e-06,
794
+ "loss": 0.2239,
795
+ "step": 1100
796
+ },
797
+ {
798
+ "epoch": 2.6303317535545023,
799
+ "grad_norm": 0.515758932015822,
800
+ "learning_rate": 5e-06,
801
+ "loss": 0.225,
802
+ "step": 1110
803
+ },
804
+ {
805
+ "epoch": 2.654028436018957,
806
+ "grad_norm": 0.5245283280931154,
807
+ "learning_rate": 5e-06,
808
+ "loss": 0.2217,
809
+ "step": 1120
810
+ },
811
+ {
812
+ "epoch": 2.677725118483412,
813
+ "grad_norm": 0.47961811011847694,
814
+ "learning_rate": 5e-06,
815
+ "loss": 0.2223,
816
+ "step": 1130
817
+ },
818
+ {
819
+ "epoch": 2.7014218009478674,
820
+ "grad_norm": 0.532667073838913,
821
+ "learning_rate": 5e-06,
822
+ "loss": 0.2202,
823
+ "step": 1140
824
+ },
825
+ {
826
+ "epoch": 2.7251184834123223,
827
+ "grad_norm": 0.4731691182426918,
828
+ "learning_rate": 5e-06,
829
+ "loss": 0.2203,
830
+ "step": 1150
831
+ },
832
+ {
833
+ "epoch": 2.748815165876777,
834
+ "grad_norm": 0.5085530590505501,
835
+ "learning_rate": 5e-06,
836
+ "loss": 0.2195,
837
+ "step": 1160
838
+ },
839
+ {
840
+ "epoch": 2.772511848341232,
841
+ "grad_norm": 0.5212827222451933,
842
+ "learning_rate": 5e-06,
843
+ "loss": 0.2201,
844
+ "step": 1170
845
+ },
846
+ {
847
+ "epoch": 2.7962085308056874,
848
+ "grad_norm": 0.6101309650612445,
849
+ "learning_rate": 5e-06,
850
+ "loss": 0.2241,
851
+ "step": 1180
852
+ },
853
+ {
854
+ "epoch": 2.8199052132701423,
855
+ "grad_norm": 0.524976392694722,
856
+ "learning_rate": 5e-06,
857
+ "loss": 0.2185,
858
+ "step": 1190
859
+ },
860
+ {
861
+ "epoch": 2.843601895734597,
862
+ "grad_norm": 0.5268018676437338,
863
+ "learning_rate": 5e-06,
864
+ "loss": 0.2218,
865
+ "step": 1200
866
+ },
867
+ {
868
+ "epoch": 2.867298578199052,
869
+ "grad_norm": 0.4815524564659738,
870
+ "learning_rate": 5e-06,
871
+ "loss": 0.2211,
872
+ "step": 1210
873
+ },
874
+ {
875
+ "epoch": 2.890995260663507,
876
+ "grad_norm": 0.5166341199930529,
877
+ "learning_rate": 5e-06,
878
+ "loss": 0.2194,
879
+ "step": 1220
880
+ },
881
+ {
882
+ "epoch": 2.914691943127962,
883
+ "grad_norm": 0.4522595493688504,
884
+ "learning_rate": 5e-06,
885
+ "loss": 0.2258,
886
+ "step": 1230
887
+ },
888
+ {
889
+ "epoch": 2.938388625592417,
890
+ "grad_norm": 0.5232915579904739,
891
+ "learning_rate": 5e-06,
892
+ "loss": 0.2218,
893
+ "step": 1240
894
+ },
895
+ {
896
+ "epoch": 2.962085308056872,
897
+ "grad_norm": 0.5580173773855611,
898
+ "learning_rate": 5e-06,
899
+ "loss": 0.2211,
900
+ "step": 1250
901
+ },
902
+ {
903
+ "epoch": 2.985781990521327,
904
+ "grad_norm": 0.5398857489449794,
905
+ "learning_rate": 5e-06,
906
+ "loss": 0.2241,
907
+ "step": 1260
908
+ },
909
+ {
910
+ "epoch": 3.0,
911
+ "eval_loss": 0.30088362097740173,
912
+ "eval_runtime": 40.2088,
913
+ "eval_samples_per_second": 282.624,
914
+ "eval_steps_per_second": 1.119,
915
+ "step": 1266
916
+ },
917
+ {
918
+ "epoch": 3.0,
919
+ "step": 1266,
920
+ "total_flos": 2120597152727040.0,
921
+ "train_loss": 0.2836323574444317,
922
+ "train_runtime": 7986.5453,
923
+ "train_samples_per_second": 81.101,
924
+ "train_steps_per_second": 0.159
925
+ }
926
+ ],
927
+ "logging_steps": 10,
928
+ "max_steps": 1266,
929
+ "num_input_tokens_seen": 0,
930
+ "num_train_epochs": 3,
931
+ "save_steps": 500,
932
+ "stateful_callbacks": {
933
+ "TrainerControl": {
934
+ "args": {
935
+ "should_epoch_stop": false,
936
+ "should_evaluate": false,
937
+ "should_log": false,
938
+ "should_save": true,
939
+ "should_training_stop": true
940
+ },
941
+ "attributes": {}
942
+ }
943
+ },
944
+ "total_flos": 2120597152727040.0,
945
+ "train_batch_size": 16,
946
+ "trial_name": null,
947
+ "trial_params": null
948
+ }
training_eval_loss.png ADDED
training_loss.png ADDED