loris3 commited on
Commit
a1bbe8e
·
verified ·
1 Parent(s): 3a9baec

Upload folder using huggingface_hub

Browse files
checkpoints/checkpoint-4698/config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LlamaForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 1,
8
+ "eos_token_id": 2,
9
+ "head_dim": 64,
10
+ "hidden_act": "silu",
11
+ "hidden_size": 768,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 2048,
14
+ "max_position_embeddings": 256,
15
+ "mlp_bias": false,
16
+ "model_type": "llama",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 12,
19
+ "num_key_value_heads": 12,
20
+ "pad_token_id": 1,
21
+ "pretraining_tp": 1,
22
+ "rms_norm_eps": 1e-06,
23
+ "rope_scaling": null,
24
+ "rope_theta": 10000.0,
25
+ "tie_word_embeddings": true,
26
+ "torch_dtype": "float32",
27
+ "transformers_version": "4.47.0",
28
+ "use_cache": true,
29
+ "vocab_size": 16000
30
+ }
checkpoints/checkpoint-4698/generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "eos_token_id": 2,
5
+ "pad_token_id": 1,
6
+ "transformers_version": "4.47.0"
7
+ }
checkpoints/checkpoint-4698/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:324d866345aed95ee504c567ff2d6e53f5bfb3929db0352d76e367e249748697
3
+ size 388979624
checkpoints/checkpoint-4698/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:19ef85b3502ef680132bf80384523e4e8da41cc0bcb34b05a5b31f78e431b3ee
3
+ size 778027770
checkpoints/checkpoint-4698/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c70a41264c08a1e8401b7173fe9901cfca41eb5cf987bd975ed722bcda9db818
3
+ size 14244
checkpoints/checkpoint-4698/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd9212863f777dcdbfb6c335a83079842969e83d8b384cd22c5313931e509967
3
+ size 1064
checkpoints/checkpoint-4698/trainer_state.json ADDED
@@ -0,0 +1,774 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 8.998804113848362,
5
+ "eval_steps": 500,
6
+ "global_step": 4698,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.09567089213106912,
13
+ "grad_norm": 11.793500900268555,
14
+ "learning_rate": 0.00011666666666666667,
15
+ "loss": 134.5754,
16
+ "step": 50
17
+ },
18
+ {
19
+ "epoch": 0.19134178426213824,
20
+ "grad_norm": 8.607985496520996,
21
+ "learning_rate": 0.00023333333333333333,
22
+ "loss": 103.0838,
23
+ "step": 100
24
+ },
25
+ {
26
+ "epoch": 0.2870126763932074,
27
+ "grad_norm": 10.04274845123291,
28
+ "learning_rate": 0.00035,
29
+ "loss": 90.5819,
30
+ "step": 150
31
+ },
32
+ {
33
+ "epoch": 0.3826835685242765,
34
+ "grad_norm": 7.803333759307861,
35
+ "learning_rate": 0.00046666666666666666,
36
+ "loss": 83.9902,
37
+ "step": 200
38
+ },
39
+ {
40
+ "epoch": 0.4783544606553456,
41
+ "grad_norm": 6.973517894744873,
42
+ "learning_rate": 0.0005833333333333334,
43
+ "loss": 80.1721,
44
+ "step": 250
45
+ },
46
+ {
47
+ "epoch": 0.5740253527864148,
48
+ "grad_norm": 6.4767961502075195,
49
+ "learning_rate": 0.0007,
50
+ "loss": 77.4189,
51
+ "step": 300
52
+ },
53
+ {
54
+ "epoch": 0.6696962449174838,
55
+ "grad_norm": 6.103850841522217,
56
+ "learning_rate": 0.000699821634561209,
57
+ "loss": 75.1767,
58
+ "step": 350
59
+ },
60
+ {
61
+ "epoch": 0.765367137048553,
62
+ "grad_norm": 5.816777229309082,
63
+ "learning_rate": 0.0006992867200404345,
64
+ "loss": 73.4409,
65
+ "step": 400
66
+ },
67
+ {
68
+ "epoch": 0.8610380291796221,
69
+ "grad_norm": 5.716831684112549,
70
+ "learning_rate": 0.0006983958016391807,
71
+ "loss": 72.072,
72
+ "step": 450
73
+ },
74
+ {
75
+ "epoch": 0.9567089213106912,
76
+ "grad_norm": 6.124087333679199,
77
+ "learning_rate": 0.0006971497874091708,
78
+ "loss": 71.1102,
79
+ "step": 500
80
+ },
81
+ {
82
+ "epoch": 0.9988041138483617,
83
+ "eval_accuracy": 0.0,
84
+ "eval_loss": 4.415195941925049,
85
+ "eval_normalizer": 679845.0,
86
+ "eval_runtime": 152.8944,
87
+ "eval_samples_per_second": 349.634,
88
+ "eval_steps_per_second": 0.687,
89
+ "step": 522
90
+ },
91
+ {
92
+ "epoch": 1.0535756995933987,
93
+ "grad_norm": 5.71658182144165,
94
+ "learning_rate": 0.0006955499473268326,
95
+ "loss": 69.4994,
96
+ "step": 550
97
+ },
98
+ {
99
+ "epoch": 1.1492465917244679,
100
+ "grad_norm": 5.8848676681518555,
101
+ "learning_rate": 0.0006935979119988993,
102
+ "loss": 68.1432,
103
+ "step": 600
104
+ },
105
+ {
106
+ "epoch": 1.244917483855537,
107
+ "grad_norm": 5.555721282958984,
108
+ "learning_rate": 0.0006912956710004438,
109
+ "loss": 67.7763,
110
+ "step": 650
111
+ },
112
+ {
113
+ "epoch": 1.3405883759866062,
114
+ "grad_norm": 5.780415058135986,
115
+ "learning_rate": 0.0006886455708470427,
116
+ "loss": 67.4878,
117
+ "step": 700
118
+ },
119
+ {
120
+ "epoch": 1.4362592681176751,
121
+ "grad_norm": 5.483816146850586,
122
+ "learning_rate": 0.0006856503126031346,
123
+ "loss": 67.1099,
124
+ "step": 750
125
+ },
126
+ {
127
+ "epoch": 1.5319301602487443,
128
+ "grad_norm": 5.490396976470947,
129
+ "learning_rate": 0.0006823129491290102,
130
+ "loss": 66.8468,
131
+ "step": 800
132
+ },
133
+ {
134
+ "epoch": 1.6276010523798135,
135
+ "grad_norm": 5.378612995147705,
136
+ "learning_rate": 0.0006786368819692442,
137
+ "loss": 66.4973,
138
+ "step": 850
139
+ },
140
+ {
141
+ "epoch": 1.7232719445108826,
142
+ "grad_norm": 5.397339344024658,
143
+ "learning_rate": 0.0006746258578857331,
144
+ "loss": 66.2689,
145
+ "step": 900
146
+ },
147
+ {
148
+ "epoch": 1.8189428366419516,
149
+ "grad_norm": 5.1414971351623535,
150
+ "learning_rate": 0.000670283965038881,
151
+ "loss": 65.7849,
152
+ "step": 950
153
+ },
154
+ {
155
+ "epoch": 1.9146137287730207,
156
+ "grad_norm": 5.261592864990234,
157
+ "learning_rate": 0.0006656156288208179,
158
+ "loss": 65.699,
159
+ "step": 1000
160
+ },
161
+ {
162
+ "epoch": 1.9988041138483617,
163
+ "eval_accuracy": 0.0,
164
+ "eval_loss": 4.1399054527282715,
165
+ "eval_normalizer": 679845.0,
166
+ "eval_runtime": 164.3524,
167
+ "eval_samples_per_second": 325.259,
168
+ "eval_steps_per_second": 0.639,
169
+ "step": 1044
170
+ },
171
+ {
172
+ "epoch": 2.0114805070557282,
173
+ "grad_norm": 5.168030261993408,
174
+ "learning_rate": 0.000660625607344904,
175
+ "loss": 65.1215,
176
+ "step": 1050
177
+ },
178
+ {
179
+ "epoch": 2.1071513991867974,
180
+ "grad_norm": 5.651010990142822,
181
+ "learning_rate": 0.0006553189865961112,
182
+ "loss": 62.0117,
183
+ "step": 1100
184
+ },
185
+ {
186
+ "epoch": 2.2028222913178666,
187
+ "grad_norm": 5.6423115730285645,
188
+ "learning_rate": 0.0006497011752472301,
189
+ "loss": 62.0058,
190
+ "step": 1150
191
+ },
192
+ {
193
+ "epoch": 2.2984931834489357,
194
+ "grad_norm": 5.54484224319458,
195
+ "learning_rate": 0.0006437778991461825,
196
+ "loss": 62.0492,
197
+ "step": 1200
198
+ },
199
+ {
200
+ "epoch": 2.394164075580005,
201
+ "grad_norm": 5.494040489196777,
202
+ "learning_rate": 0.0006375551954800587,
203
+ "loss": 62.1302,
204
+ "step": 1250
205
+ },
206
+ {
207
+ "epoch": 2.489834967711074,
208
+ "grad_norm": 5.8661298751831055,
209
+ "learning_rate": 0.0006310394066218296,
210
+ "loss": 62.2036,
211
+ "step": 1300
212
+ },
213
+ {
214
+ "epoch": 2.5855058598421428,
215
+ "grad_norm": 5.392300605773926,
216
+ "learning_rate": 0.0006242371736660025,
217
+ "loss": 62.1324,
218
+ "step": 1350
219
+ },
220
+ {
221
+ "epoch": 2.6811767519732124,
222
+ "grad_norm": 5.379298210144043,
223
+ "learning_rate": 0.000617155429659811,
224
+ "loss": 62.155,
225
+ "step": 1400
226
+ },
227
+ {
228
+ "epoch": 2.776847644104281,
229
+ "grad_norm": 5.4863152503967285,
230
+ "learning_rate": 0.0006098013925368385,
231
+ "loss": 62.1645,
232
+ "step": 1450
233
+ },
234
+ {
235
+ "epoch": 2.8725185362353503,
236
+ "grad_norm": 5.372952938079834,
237
+ "learning_rate": 0.0006021825577602754,
238
+ "loss": 61.9254,
239
+ "step": 1500
240
+ },
241
+ {
242
+ "epoch": 2.9681894283664194,
243
+ "grad_norm": 5.275949954986572,
244
+ "learning_rate": 0.0005943066906833104,
245
+ "loss": 62.0202,
246
+ "step": 1550
247
+ },
248
+ {
249
+ "epoch": 2.9988041138483617,
250
+ "eval_accuracy": 0.0,
251
+ "eval_loss": 4.02716588973999,
252
+ "eval_normalizer": 679845.0,
253
+ "eval_runtime": 100.1843,
254
+ "eval_samples_per_second": 533.587,
255
+ "eval_steps_per_second": 1.048,
256
+ "step": 1566
257
+ },
258
+ {
259
+ "epoch": 3.065056206649127,
260
+ "grad_norm": 5.55876350402832,
261
+ "learning_rate": 0.0005861818186344407,
262
+ "loss": 58.7311,
263
+ "step": 1600
264
+ },
265
+ {
266
+ "epoch": 3.160727098780196,
267
+ "grad_norm": 5.992660045623779,
268
+ "learning_rate": 0.00057781622273577,
269
+ "loss": 57.1846,
270
+ "step": 1650
271
+ },
272
+ {
273
+ "epoch": 3.2563979909112652,
274
+ "grad_norm": 5.9196953773498535,
275
+ "learning_rate": 0.0005692184294626307,
276
+ "loss": 57.5236,
277
+ "step": 1700
278
+ },
279
+ {
280
+ "epoch": 3.3520688830423344,
281
+ "grad_norm": 6.049771785736084,
282
+ "learning_rate": 0.0005603972019531362,
283
+ "loss": 57.7079,
284
+ "step": 1750
285
+ },
286
+ {
287
+ "epoch": 3.4477397751734036,
288
+ "grad_norm": 5.912595748901367,
289
+ "learning_rate": 0.0005513615310765172,
290
+ "loss": 57.7866,
291
+ "step": 1800
292
+ },
293
+ {
294
+ "epoch": 3.5434106673044727,
295
+ "grad_norm": 6.013986587524414,
296
+ "learning_rate": 0.0005421206262693491,
297
+ "loss": 58.0619,
298
+ "step": 1850
299
+ },
300
+ {
301
+ "epoch": 3.639081559435542,
302
+ "grad_norm": 5.911418914794922,
303
+ "learning_rate": 0.0005326839061490078,
304
+ "loss": 58.0844,
305
+ "step": 1900
306
+ },
307
+ {
308
+ "epoch": 3.734752451566611,
309
+ "grad_norm": 5.901169776916504,
310
+ "learning_rate": 0.0005230609889139216,
311
+ "loss": 58.1622,
312
+ "step": 1950
313
+ },
314
+ {
315
+ "epoch": 3.83042334369768,
316
+ "grad_norm": 6.09370756149292,
317
+ "learning_rate": 0.0005132616825404055,
318
+ "loss": 58.1267,
319
+ "step": 2000
320
+ },
321
+ {
322
+ "epoch": 3.926094235828749,
323
+ "grad_norm": 6.066126346588135,
324
+ "learning_rate": 0.0005032959747860662,
325
+ "loss": 58.3174,
326
+ "step": 2050
327
+ },
328
+ {
329
+ "epoch": 3.9988041138483617,
330
+ "eval_accuracy": 0.0,
331
+ "eval_loss": 3.9929206371307373,
332
+ "eval_normalizer": 679845.0,
333
+ "eval_runtime": 96.8965,
334
+ "eval_samples_per_second": 551.692,
335
+ "eval_steps_per_second": 1.084,
336
+ "step": 2088
337
+ },
338
+ {
339
+ "epoch": 4.0229610141114565,
340
+ "grad_norm": 6.213118553161621,
341
+ "learning_rate": 0.000493174023009969,
342
+ "loss": 56.8785,
343
+ "step": 2100
344
+ },
345
+ {
346
+ "epoch": 4.118631906242526,
347
+ "grad_norm": 6.573572158813477,
348
+ "learning_rate": 0.00048290614381994235,
349
+ "loss": 52.002,
350
+ "step": 2150
351
+ },
352
+ {
353
+ "epoch": 4.214302798373595,
354
+ "grad_norm": 6.8537468910217285,
355
+ "learning_rate": 0.00047250280255757023,
356
+ "loss": 52.3166,
357
+ "step": 2200
358
+ },
359
+ {
360
+ "epoch": 4.309973690504664,
361
+ "grad_norm": 6.924459934234619,
362
+ "learning_rate": 0.0004619746026315906,
363
+ "loss": 52.6308,
364
+ "step": 2250
365
+ },
366
+ {
367
+ "epoch": 4.405644582635733,
368
+ "grad_norm": 7.022147178649902,
369
+ "learning_rate": 0.00045133227471057203,
370
+ "loss": 52.8349,
371
+ "step": 2300
372
+ },
373
+ {
374
+ "epoch": 4.501315474766802,
375
+ "grad_norm": 7.115175247192383,
376
+ "learning_rate": 0.00044058666578588224,
377
+ "loss": 53.0948,
378
+ "step": 2350
379
+ },
380
+ {
381
+ "epoch": 4.596986366897871,
382
+ "grad_norm": 7.212996482849121,
383
+ "learning_rate": 0.0004297487281160982,
384
+ "loss": 53.1418,
385
+ "step": 2400
386
+ },
387
+ {
388
+ "epoch": 4.69265725902894,
389
+ "grad_norm": 7.168196201324463,
390
+ "learning_rate": 0.00041882950806412285,
391
+ "loss": 53.3312,
392
+ "step": 2450
393
+ },
394
+ {
395
+ "epoch": 4.78832815116001,
396
+ "grad_norm": 7.141201972961426,
397
+ "learning_rate": 0.0004078401348383897,
398
+ "loss": 53.3364,
399
+ "step": 2500
400
+ },
401
+ {
402
+ "epoch": 4.8839990432910785,
403
+ "grad_norm": 7.150181770324707,
404
+ "learning_rate": 0.00039679180914962693,
405
+ "loss": 53.5222,
406
+ "step": 2550
407
+ },
408
+ {
409
+ "epoch": 4.979669935422148,
410
+ "grad_norm": 7.120785236358643,
411
+ "learning_rate": 0.00038569579179474536,
412
+ "loss": 53.5096,
413
+ "step": 2600
414
+ },
415
+ {
416
+ "epoch": 4.998804113848362,
417
+ "eval_accuracy": 0.0,
418
+ "eval_loss": 4.0353522300720215,
419
+ "eval_normalizer": 679845.0,
420
+ "eval_runtime": 99.3537,
421
+ "eval_samples_per_second": 538.047,
422
+ "eval_steps_per_second": 1.057,
423
+ "step": 2610
424
+ },
425
+ {
426
+ "epoch": 5.076536713704855,
427
+ "grad_norm": 7.744894504547119,
428
+ "learning_rate": 0.00037456339217948394,
429
+ "loss": 47.6374,
430
+ "step": 2650
431
+ },
432
+ {
433
+ "epoch": 5.172207605835925,
434
+ "grad_norm": 8.048383712768555,
435
+ "learning_rate": 0.0003634059567915124,
436
+ "loss": 46.0838,
437
+ "step": 2700
438
+ },
439
+ {
440
+ "epoch": 5.2678784979669935,
441
+ "grad_norm": 8.176159858703613,
442
+ "learning_rate": 0.00035223485763573775,
443
+ "loss": 46.4321,
444
+ "step": 2750
445
+ },
446
+ {
447
+ "epoch": 5.363549390098063,
448
+ "grad_norm": 8.581367492675781,
449
+ "learning_rate": 0.00034106148064360405,
450
+ "loss": 46.7736,
451
+ "step": 2800
452
+ },
453
+ {
454
+ "epoch": 5.459220282229132,
455
+ "grad_norm": 8.688493728637695,
456
+ "learning_rate": 0.0003298972140681969,
457
+ "loss": 47.1028,
458
+ "step": 2850
459
+ },
460
+ {
461
+ "epoch": 5.5548911743602005,
462
+ "grad_norm": 8.841804504394531,
463
+ "learning_rate": 0.00031875343687698203,
464
+ "loss": 47.1435,
465
+ "step": 2900
466
+ },
467
+ {
468
+ "epoch": 5.65056206649127,
469
+ "grad_norm": 8.769726753234863,
470
+ "learning_rate": 0.000307641507154008,
471
+ "loss": 47.3549,
472
+ "step": 2950
473
+ },
474
+ {
475
+ "epoch": 5.746232958622339,
476
+ "grad_norm": 8.779825210571289,
477
+ "learning_rate": 0.0002965727505233939,
478
+ "loss": 47.5233,
479
+ "step": 3000
480
+ },
481
+ {
482
+ "epoch": 5.8419038507534085,
483
+ "grad_norm": 8.91053295135498,
484
+ "learning_rate": 0.0002855584486059016,
485
+ "loss": 47.5871,
486
+ "step": 3050
487
+ },
488
+ {
489
+ "epoch": 5.937574742884477,
490
+ "grad_norm": 8.902570724487305,
491
+ "learning_rate": 0.00027460982752035653,
492
+ "loss": 47.7148,
493
+ "step": 3100
494
+ },
495
+ {
496
+ "epoch": 5.998804113848362,
497
+ "eval_accuracy": 0.0,
498
+ "eval_loss": 4.167338848114014,
499
+ "eval_normalizer": 679845.0,
500
+ "eval_runtime": 95.6702,
501
+ "eval_samples_per_second": 558.763,
502
+ "eval_steps_per_second": 1.098,
503
+ "step": 3132
504
+ },
505
+ {
506
+ "epoch": 6.034441521167185,
507
+ "grad_norm": 8.682727813720703,
508
+ "learning_rate": 0.00026373804644163934,
509
+ "loss": 44.9027,
510
+ "step": 3150
511
+ },
512
+ {
513
+ "epoch": 6.130112413298254,
514
+ "grad_norm": 9.522422790527344,
515
+ "learning_rate": 0.00025295418622690716,
516
+ "loss": 39.3671,
517
+ "step": 3200
518
+ },
519
+ {
520
+ "epoch": 6.2257833054293235,
521
+ "grad_norm": 9.705806732177734,
522
+ "learning_rate": 0.00024226923812164069,
523
+ "loss": 39.7118,
524
+ "step": 3250
525
+ },
526
+ {
527
+ "epoch": 6.321454197560392,
528
+ "grad_norm": 9.989520072937012,
529
+ "learning_rate": 0.0002316940925570241,
530
+ "loss": 40.0659,
531
+ "step": 3300
532
+ },
533
+ {
534
+ "epoch": 6.417125089691462,
535
+ "grad_norm": 10.2572021484375,
536
+ "learning_rate": 0.0002212395280500796,
537
+ "loss": 40.2838,
538
+ "step": 3350
539
+ },
540
+ {
541
+ "epoch": 6.5127959818225305,
542
+ "grad_norm": 10.23260498046875,
543
+ "learning_rate": 0.00021091620021786818,
544
+ "loss": 40.504,
545
+ "step": 3400
546
+ },
547
+ {
548
+ "epoch": 6.608466873953599,
549
+ "grad_norm": 10.518167495727539,
550
+ "learning_rate": 0.0002007346309169531,
551
+ "loss": 40.5793,
552
+ "step": 3450
553
+ },
554
+ {
555
+ "epoch": 6.704137766084669,
556
+ "grad_norm": 10.158038139343262,
557
+ "learning_rate": 0.00019070519751919701,
558
+ "loss": 40.6567,
559
+ "step": 3500
560
+ },
561
+ {
562
+ "epoch": 6.7998086582157375,
563
+ "grad_norm": 10.4249267578125,
564
+ "learning_rate": 0.0001808381223348215,
565
+ "loss": 40.8632,
566
+ "step": 3550
567
+ },
568
+ {
569
+ "epoch": 6.895479550346807,
570
+ "grad_norm": 10.527756690979004,
571
+ "learning_rate": 0.0001711434621935113,
572
+ "loss": 40.8792,
573
+ "step": 3600
574
+ },
575
+ {
576
+ "epoch": 6.991150442477876,
577
+ "grad_norm": 10.697628021240234,
578
+ "learning_rate": 0.00016163109819418135,
579
+ "loss": 40.9673,
580
+ "step": 3650
581
+ },
582
+ {
583
+ "epoch": 6.998804113848362,
584
+ "eval_accuracy": 0.0,
585
+ "eval_loss": 4.374588489532471,
586
+ "eval_normalizer": 679845.0,
587
+ "eval_runtime": 111.926,
588
+ "eval_samples_per_second": 477.61,
589
+ "eval_steps_per_second": 0.938,
590
+ "step": 3654
591
+ },
592
+ {
593
+ "epoch": 7.088017220760584,
594
+ "grad_norm": 10.376972198486328,
595
+ "learning_rate": 0.0001523107256338523,
596
+ "loss": 33.8772,
597
+ "step": 3700
598
+ },
599
+ {
600
+ "epoch": 7.1836881128916525,
601
+ "grad_norm": 11.042153358459473,
602
+ "learning_rate": 0.00014319184412590392,
603
+ "loss": 33.3445,
604
+ "step": 3750
605
+ },
606
+ {
607
+ "epoch": 7.279359005022722,
608
+ "grad_norm": 11.128947257995605,
609
+ "learning_rate": 0.00013428374791777268,
610
+ "loss": 33.6749,
611
+ "step": 3800
612
+ },
613
+ {
614
+ "epoch": 7.375029897153791,
615
+ "grad_norm": 11.447359085083008,
616
+ "learning_rate": 0.0001255955164179673,
617
+ "loss": 33.8207,
618
+ "step": 3850
619
+ },
620
+ {
621
+ "epoch": 7.4707007892848605,
622
+ "grad_norm": 11.455679893493652,
623
+ "learning_rate": 0.00011713600494205216,
624
+ "loss": 33.9581,
625
+ "step": 3900
626
+ },
627
+ {
628
+ "epoch": 7.566371681415929,
629
+ "grad_norm": 11.351855278015137,
630
+ "learning_rate": 0.0001089138356870349,
631
+ "loss": 34.0369,
632
+ "step": 3950
633
+ },
634
+ {
635
+ "epoch": 7.662042573546998,
636
+ "grad_norm": 11.856739044189453,
637
+ "learning_rate": 0.00010093738894335547,
638
+ "loss": 34.1848,
639
+ "step": 4000
640
+ },
641
+ {
642
+ "epoch": 7.7577134656780675,
643
+ "grad_norm": 11.871599197387695,
644
+ "learning_rate": 9.321479455343351e-05,
645
+ "loss": 34.2809,
646
+ "step": 4050
647
+ },
648
+ {
649
+ "epoch": 7.853384357809136,
650
+ "grad_norm": 11.887659072875977,
651
+ "learning_rate": 8.575392362548022e-05,
652
+ "loss": 34.2916,
653
+ "step": 4100
654
+ },
655
+ {
656
+ "epoch": 7.949055249940206,
657
+ "grad_norm": 11.994010925292969,
658
+ "learning_rate": 7.856238051102045e-05,
659
+ "loss": 34.2647,
660
+ "step": 4150
661
+ },
662
+ {
663
+ "epoch": 7.998804113848362,
664
+ "eval_accuracy": 0.0,
665
+ "eval_loss": 4.619776248931885,
666
+ "eval_normalizer": 679845.0,
667
+ "eval_runtime": 137.8494,
668
+ "eval_samples_per_second": 387.793,
669
+ "eval_steps_per_second": 0.762,
670
+ "step": 4176
671
+ },
672
+ {
673
+ "epoch": 8.045922028222913,
674
+ "grad_norm": 10.849797248840332,
675
+ "learning_rate": 7.164749505430072e-05,
676
+ "loss": 31.6597,
677
+ "step": 4200
678
+ },
679
+ {
680
+ "epoch": 8.141592920353983,
681
+ "grad_norm": 11.105512619018555,
682
+ "learning_rate": 6.501631512148524e-05,
683
+ "loss": 28.7426,
684
+ "step": 4250
685
+ },
686
+ {
687
+ "epoch": 8.237263812485052,
688
+ "grad_norm": 11.203261375427246,
689
+ "learning_rate": 5.867559941725047e-05,
690
+ "loss": 28.8556,
691
+ "step": 4300
692
+ },
693
+ {
694
+ "epoch": 8.33293470461612,
695
+ "grad_norm": 11.628230094909668,
696
+ "learning_rate": 5.263181059610415e-05,
697
+ "loss": 29.077,
698
+ "step": 4350
699
+ },
700
+ {
701
+ "epoch": 8.42860559674719,
702
+ "grad_norm": 11.53283405303955,
703
+ "learning_rate": 4.689110867544645e-05,
704
+ "loss": 29.1128,
705
+ "step": 4400
706
+ },
707
+ {
708
+ "epoch": 8.52427648887826,
709
+ "grad_norm": 11.642229080200195,
710
+ "learning_rate": 4.145934475708915e-05,
711
+ "loss": 29.2207,
712
+ "step": 4450
713
+ },
714
+ {
715
+ "epoch": 8.619947381009329,
716
+ "grad_norm": 12.052610397338867,
717
+ "learning_rate": 3.634205506363058e-05,
718
+ "loss": 29.1995,
719
+ "step": 4500
720
+ },
721
+ {
722
+ "epoch": 8.715618273140397,
723
+ "grad_norm": 11.68619155883789,
724
+ "learning_rate": 3.154445529576543e-05,
725
+ "loss": 29.2274,
726
+ "step": 4550
727
+ },
728
+ {
729
+ "epoch": 8.811289165271466,
730
+ "grad_norm": 12.159764289855957,
731
+ "learning_rate": 2.7071435316280922e-05,
732
+ "loss": 29.1777,
733
+ "step": 4600
734
+ },
735
+ {
736
+ "epoch": 8.906960057402536,
737
+ "grad_norm": 11.982682228088379,
738
+ "learning_rate": 2.292755416615615e-05,
739
+ "loss": 29.2239,
740
+ "step": 4650
741
+ },
742
+ {
743
+ "epoch": 8.998804113848362,
744
+ "eval_accuracy": 0.0,
745
+ "eval_loss": 4.816934585571289,
746
+ "eval_normalizer": 679845.0,
747
+ "eval_runtime": 99.4971,
748
+ "eval_samples_per_second": 537.272,
749
+ "eval_steps_per_second": 1.055,
750
+ "step": 4698
751
+ }
752
+ ],
753
+ "logging_steps": 50,
754
+ "max_steps": 5220,
755
+ "num_input_tokens_seen": 0,
756
+ "num_train_epochs": 10,
757
+ "save_steps": 500,
758
+ "stateful_callbacks": {
759
+ "TrainerControl": {
760
+ "args": {
761
+ "should_epoch_stop": false,
762
+ "should_evaluate": false,
763
+ "should_log": false,
764
+ "should_save": true,
765
+ "should_training_stop": false
766
+ },
767
+ "attributes": {}
768
+ }
769
+ },
770
+ "total_flos": 8.086004467289948e+17,
771
+ "train_batch_size": 128,
772
+ "trial_name": null,
773
+ "trial_params": null
774
+ }
checkpoints/checkpoint-4698/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a2a4f97f7e7735163153c8175c9361045adcb7dd68c056222c2951d24c61bb2
3
+ size 5368