abhinavp commited on
Commit
cd9fc5a
1 Parent(s): b1cb74d

Training in progress, epoch 0, checkpoint

Browse files
checkpoint-76320/config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_remove_final_layer_norm": false,
3
+ "activation_function": "relu",
4
+ "architectures": [
5
+ "OPTForCausalLM"
6
+ ],
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 2,
9
+ "do_layer_norm_before": true,
10
+ "dropout": 0.1,
11
+ "enable_bias": true,
12
+ "eos_token_id": 2,
13
+ "ffn_dim": 768,
14
+ "hidden_size": 768,
15
+ "init_std": 0.02,
16
+ "layer_norm_elementwise_affine": true,
17
+ "layerdrop": 0.0,
18
+ "max_position_embeddings": 512,
19
+ "model_type": "opt",
20
+ "num_attention_heads": 8,
21
+ "num_hidden_layers": 8,
22
+ "pad_token_id": 1,
23
+ "torch_dtype": "float32",
24
+ "transformers_version": "4.33.3",
25
+ "use_cache": true,
26
+ "vocab_size": 50002,
27
+ "word_embed_proj_dim": 768
28
+ }
checkpoint-76320/generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 2,
4
+ "eos_token_id": 2,
5
+ "pad_token_id": 1,
6
+ "transformers_version": "4.33.3"
7
+ }
checkpoint-76320/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8551f0f370a0a4f4912bfa6b2083d3c1f8f0cfc786b2553191d248af33eba7c9
3
+ size 537476677
checkpoint-76320/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac0d34824fa8d33268533881cc79b9a63c561129fb6fdeb241c575303bb7dc5d
3
+ size 268727709
checkpoint-76320/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a43fb7d3bed5e93a10489f237aa3766bf2ddeab43dd2046e04a07b1edd518cd2
3
+ size 14575
checkpoint-76320/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cf11f1455a7c1dfeea4b4706604b9a9686a940ee59e059424f6c23c271a8890e
3
+ size 627
checkpoint-76320/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "eos_token": "<eos>",
3
+ "pad_token": "<pad>",
4
+ "unk_token": "<unk>"
5
+ }
checkpoint-76320/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-76320/tokenizer_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "clean_up_tokenization_spaces": true,
3
+ "eos_token": "<eos>",
4
+ "max_length": null,
5
+ "model_max_length": 512,
6
+ "pad_to_multiple_of": null,
7
+ "pad_token": "<pad>",
8
+ "pad_token_type_id": 0,
9
+ "padding_side": "right",
10
+ "tokenizer_class": "PreTrainedTokenizerFast",
11
+ "unk_token": "<unk>"
12
+ }
checkpoint-76320/trainer_state.json ADDED
@@ -0,0 +1,927 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 4.196377277374268,
3
+ "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/npi-only/transformer/3/checkpoints/checkpoint-76320",
4
+ "epoch": 0.025000606015738065,
5
+ "eval_steps": 10,
6
+ "global_step": 76320,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 4.999998362119627e-05,
14
+ "loss": 10.9828,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 4.999161405248948e-05,
20
+ "loss": 6.815,
21
+ "step": 512
22
+ },
23
+ {
24
+ "epoch": 0.0,
25
+ "learning_rate": 4.998322810497896e-05,
26
+ "loss": 6.1821,
27
+ "step": 1024
28
+ },
29
+ {
30
+ "epoch": 0.0,
31
+ "learning_rate": 4.997484215746844e-05,
32
+ "loss": 5.9729,
33
+ "step": 1536
34
+ },
35
+ {
36
+ "epoch": 0.0,
37
+ "learning_rate": 4.996645620995792e-05,
38
+ "loss": 5.8165,
39
+ "step": 2048
40
+ },
41
+ {
42
+ "epoch": 0.0,
43
+ "learning_rate": 4.99580702624474e-05,
44
+ "loss": 5.7082,
45
+ "step": 2560
46
+ },
47
+ {
48
+ "epoch": 0.0,
49
+ "learning_rate": 4.994968431493688e-05,
50
+ "loss": 5.6061,
51
+ "step": 3072
52
+ },
53
+ {
54
+ "epoch": 0.0,
55
+ "learning_rate": 4.994129836742636e-05,
56
+ "loss": 5.5386,
57
+ "step": 3584
58
+ },
59
+ {
60
+ "epoch": 0.0,
61
+ "learning_rate": 4.993291241991584e-05,
62
+ "loss": 5.4514,
63
+ "step": 4096
64
+ },
65
+ {
66
+ "epoch": 0.0,
67
+ "learning_rate": 4.992452647240532e-05,
68
+ "loss": 5.4102,
69
+ "step": 4608
70
+ },
71
+ {
72
+ "epoch": 0.0,
73
+ "learning_rate": 4.99161405248948e-05,
74
+ "loss": 5.3599,
75
+ "step": 5120
76
+ },
77
+ {
78
+ "epoch": 0.0,
79
+ "learning_rate": 4.990775457738428e-05,
80
+ "loss": 5.3198,
81
+ "step": 5632
82
+ },
83
+ {
84
+ "epoch": 0.0,
85
+ "learning_rate": 4.989938500867749e-05,
86
+ "loss": 5.2546,
87
+ "step": 6144
88
+ },
89
+ {
90
+ "epoch": 0.0,
91
+ "learning_rate": 4.989099906116697e-05,
92
+ "loss": 5.2159,
93
+ "step": 6656
94
+ },
95
+ {
96
+ "epoch": 0.0,
97
+ "learning_rate": 4.988261311365645e-05,
98
+ "loss": 5.183,
99
+ "step": 7168
100
+ },
101
+ {
102
+ "epoch": 0.0,
103
+ "learning_rate": 4.987422716614593e-05,
104
+ "loss": 5.1335,
105
+ "step": 7680
106
+ },
107
+ {
108
+ "epoch": 0.0,
109
+ "learning_rate": 4.986584121863541e-05,
110
+ "loss": 5.1172,
111
+ "step": 8192
112
+ },
113
+ {
114
+ "epoch": 0.0,
115
+ "learning_rate": 4.985745527112489e-05,
116
+ "loss": 5.0848,
117
+ "step": 8704
118
+ },
119
+ {
120
+ "epoch": 0.0,
121
+ "learning_rate": 4.984906932361437e-05,
122
+ "loss": 5.0452,
123
+ "step": 9216
124
+ },
125
+ {
126
+ "epoch": 0.0,
127
+ "learning_rate": 4.984068337610385e-05,
128
+ "loss": 5.023,
129
+ "step": 9728
130
+ },
131
+ {
132
+ "epoch": 0.0,
133
+ "learning_rate": 4.983229742859333e-05,
134
+ "loss": 5.0069,
135
+ "step": 10240
136
+ },
137
+ {
138
+ "epoch": 0.0,
139
+ "learning_rate": 4.9823927859886547e-05,
140
+ "loss": 4.9763,
141
+ "step": 10752
142
+ },
143
+ {
144
+ "epoch": 0.0,
145
+ "learning_rate": 4.9815541912376026e-05,
146
+ "loss": 4.9588,
147
+ "step": 11264
148
+ },
149
+ {
150
+ "epoch": 0.0,
151
+ "learning_rate": 4.9807155964865506e-05,
152
+ "loss": 4.926,
153
+ "step": 11776
154
+ },
155
+ {
156
+ "epoch": 0.0,
157
+ "learning_rate": 4.9798770017354986e-05,
158
+ "loss": 4.9188,
159
+ "step": 12288
160
+ },
161
+ {
162
+ "epoch": 0.0,
163
+ "learning_rate": 4.9790400448648195e-05,
164
+ "loss": 4.8964,
165
+ "step": 12800
166
+ },
167
+ {
168
+ "epoch": 0.0,
169
+ "learning_rate": 4.9782014501137675e-05,
170
+ "loss": 4.8732,
171
+ "step": 13312
172
+ },
173
+ {
174
+ "epoch": 0.0,
175
+ "learning_rate": 4.9773628553627155e-05,
176
+ "loss": 4.8532,
177
+ "step": 13824
178
+ },
179
+ {
180
+ "epoch": 0.0,
181
+ "learning_rate": 4.9765242606116635e-05,
182
+ "loss": 4.828,
183
+ "step": 14336
184
+ },
185
+ {
186
+ "epoch": 0.0,
187
+ "learning_rate": 4.9756873037409844e-05,
188
+ "loss": 4.8199,
189
+ "step": 14848
190
+ },
191
+ {
192
+ "epoch": 0.01,
193
+ "learning_rate": 4.9748487089899324e-05,
194
+ "loss": 4.8034,
195
+ "step": 15360
196
+ },
197
+ {
198
+ "epoch": 0.01,
199
+ "learning_rate": 4.9740117521192533e-05,
200
+ "loss": 4.7967,
201
+ "step": 15872
202
+ },
203
+ {
204
+ "epoch": 0.01,
205
+ "learning_rate": 4.9731731573682013e-05,
206
+ "loss": 4.7807,
207
+ "step": 16384
208
+ },
209
+ {
210
+ "epoch": 0.01,
211
+ "learning_rate": 4.97233456261715e-05,
212
+ "loss": 4.7672,
213
+ "step": 16896
214
+ },
215
+ {
216
+ "epoch": 0.01,
217
+ "learning_rate": 4.971495967866098e-05,
218
+ "loss": 4.7584,
219
+ "step": 17408
220
+ },
221
+ {
222
+ "epoch": 0.01,
223
+ "learning_rate": 4.970657373115046e-05,
224
+ "loss": 4.7461,
225
+ "step": 17920
226
+ },
227
+ {
228
+ "epoch": 0.01,
229
+ "learning_rate": 4.969818778363994e-05,
230
+ "loss": 4.7364,
231
+ "step": 18432
232
+ },
233
+ {
234
+ "epoch": 0.01,
235
+ "learning_rate": 4.968981821493315e-05,
236
+ "loss": 4.7037,
237
+ "step": 18944
238
+ },
239
+ {
240
+ "epoch": 0.01,
241
+ "learning_rate": 4.968143226742263e-05,
242
+ "loss": 4.7038,
243
+ "step": 19456
244
+ },
245
+ {
246
+ "epoch": 0.01,
247
+ "learning_rate": 4.967304631991211e-05,
248
+ "loss": 4.6797,
249
+ "step": 19968
250
+ },
251
+ {
252
+ "epoch": 0.01,
253
+ "learning_rate": 4.966466037240159e-05,
254
+ "loss": 4.6841,
255
+ "step": 20480
256
+ },
257
+ {
258
+ "epoch": 0.01,
259
+ "learning_rate": 4.965627442489107e-05,
260
+ "loss": 4.6644,
261
+ "step": 20992
262
+ },
263
+ {
264
+ "epoch": 0.01,
265
+ "learning_rate": 4.964788847738054e-05,
266
+ "loss": 4.6666,
267
+ "step": 21504
268
+ },
269
+ {
270
+ "epoch": 0.01,
271
+ "learning_rate": 4.963950252987002e-05,
272
+ "loss": 4.6455,
273
+ "step": 22016
274
+ },
275
+ {
276
+ "epoch": 0.01,
277
+ "learning_rate": 4.96311165823595e-05,
278
+ "loss": 4.6428,
279
+ "step": 22528
280
+ },
281
+ {
282
+ "epoch": 0.01,
283
+ "learning_rate": 4.962274701365272e-05,
284
+ "loss": 4.6342,
285
+ "step": 23040
286
+ },
287
+ {
288
+ "epoch": 0.01,
289
+ "learning_rate": 4.96143610661422e-05,
290
+ "loss": 4.6276,
291
+ "step": 23552
292
+ },
293
+ {
294
+ "epoch": 0.01,
295
+ "learning_rate": 4.960597511863168e-05,
296
+ "loss": 4.6153,
297
+ "step": 24064
298
+ },
299
+ {
300
+ "epoch": 0.01,
301
+ "learning_rate": 4.959758917112116e-05,
302
+ "loss": 4.5907,
303
+ "step": 24576
304
+ },
305
+ {
306
+ "epoch": 0.01,
307
+ "learning_rate": 4.9589219602414374e-05,
308
+ "loss": 4.5971,
309
+ "step": 25088
310
+ },
311
+ {
312
+ "epoch": 0.01,
313
+ "learning_rate": 4.958083365490385e-05,
314
+ "loss": 4.5978,
315
+ "step": 25600
316
+ },
317
+ {
318
+ "epoch": 0.01,
319
+ "learning_rate": 4.957244770739333e-05,
320
+ "loss": 4.5916,
321
+ "step": 26112
322
+ },
323
+ {
324
+ "epoch": 0.01,
325
+ "learning_rate": 4.956406175988281e-05,
326
+ "loss": 4.5719,
327
+ "step": 26624
328
+ },
329
+ {
330
+ "epoch": 0.01,
331
+ "learning_rate": 4.9555692191176016e-05,
332
+ "loss": 4.5592,
333
+ "step": 27136
334
+ },
335
+ {
336
+ "epoch": 0.01,
337
+ "learning_rate": 4.954732262246923e-05,
338
+ "loss": 4.5614,
339
+ "step": 27648
340
+ },
341
+ {
342
+ "epoch": 0.01,
343
+ "learning_rate": 4.953893667495871e-05,
344
+ "loss": 4.5481,
345
+ "step": 28160
346
+ },
347
+ {
348
+ "epoch": 0.01,
349
+ "learning_rate": 4.953055072744819e-05,
350
+ "loss": 4.563,
351
+ "step": 28672
352
+ },
353
+ {
354
+ "epoch": 0.01,
355
+ "learning_rate": 4.952216477993767e-05,
356
+ "loss": 4.5249,
357
+ "step": 29184
358
+ },
359
+ {
360
+ "epoch": 0.01,
361
+ "learning_rate": 4.951377883242715e-05,
362
+ "loss": 4.5405,
363
+ "step": 29696
364
+ },
365
+ {
366
+ "epoch": 0.01,
367
+ "learning_rate": 4.950539288491663e-05,
368
+ "loss": 4.5303,
369
+ "step": 30208
370
+ },
371
+ {
372
+ "epoch": 0.01,
373
+ "learning_rate": 4.949700693740611e-05,
374
+ "loss": 4.5116,
375
+ "step": 30720
376
+ },
377
+ {
378
+ "epoch": 0.01,
379
+ "learning_rate": 4.948862098989559e-05,
380
+ "loss": 4.5213,
381
+ "step": 31232
382
+ },
383
+ {
384
+ "epoch": 0.01,
385
+ "learning_rate": 4.94802514211888e-05,
386
+ "loss": 4.5053,
387
+ "step": 31744
388
+ },
389
+ {
390
+ "epoch": 0.01,
391
+ "learning_rate": 4.947186547367828e-05,
392
+ "loss": 4.4964,
393
+ "step": 32256
394
+ },
395
+ {
396
+ "epoch": 0.01,
397
+ "learning_rate": 4.946347952616776e-05,
398
+ "loss": 4.4904,
399
+ "step": 32768
400
+ },
401
+ {
402
+ "epoch": 0.01,
403
+ "learning_rate": 4.945509357865724e-05,
404
+ "loss": 4.5027,
405
+ "step": 33280
406
+ },
407
+ {
408
+ "epoch": 0.01,
409
+ "learning_rate": 4.944672400995045e-05,
410
+ "loss": 4.4811,
411
+ "step": 33792
412
+ },
413
+ {
414
+ "epoch": 0.01,
415
+ "learning_rate": 4.943833806243993e-05,
416
+ "loss": 4.4772,
417
+ "step": 34304
418
+ },
419
+ {
420
+ "epoch": 0.01,
421
+ "learning_rate": 4.942995211492941e-05,
422
+ "loss": 4.4696,
423
+ "step": 34816
424
+ },
425
+ {
426
+ "epoch": 0.01,
427
+ "learning_rate": 4.942156616741889e-05,
428
+ "loss": 4.4621,
429
+ "step": 35328
430
+ },
431
+ {
432
+ "epoch": 0.01,
433
+ "learning_rate": 4.9413196598712105e-05,
434
+ "loss": 4.4697,
435
+ "step": 35840
436
+ },
437
+ {
438
+ "epoch": 0.01,
439
+ "learning_rate": 4.940482703000532e-05,
440
+ "loss": 4.4767,
441
+ "step": 36352
442
+ },
443
+ {
444
+ "epoch": 0.01,
445
+ "learning_rate": 4.9396441082494794e-05,
446
+ "loss": 4.4497,
447
+ "step": 36864
448
+ },
449
+ {
450
+ "epoch": 0.01,
451
+ "learning_rate": 4.9388055134984274e-05,
452
+ "loss": 4.4607,
453
+ "step": 37376
454
+ },
455
+ {
456
+ "epoch": 0.01,
457
+ "learning_rate": 4.9379669187473754e-05,
458
+ "loss": 4.4656,
459
+ "step": 37888
460
+ },
461
+ {
462
+ "epoch": 0.01,
463
+ "learning_rate": 4.9371283239963234e-05,
464
+ "loss": 4.4472,
465
+ "step": 38400
466
+ },
467
+ {
468
+ "epoch": 0.01,
469
+ "learning_rate": 4.936291367125644e-05,
470
+ "loss": 4.4359,
471
+ "step": 38912
472
+ },
473
+ {
474
+ "epoch": 0.01,
475
+ "learning_rate": 4.935452772374592e-05,
476
+ "loss": 4.4374,
477
+ "step": 39424
478
+ },
479
+ {
480
+ "epoch": 0.01,
481
+ "learning_rate": 4.93461417762354e-05,
482
+ "loss": 4.4348,
483
+ "step": 39936
484
+ },
485
+ {
486
+ "epoch": 0.01,
487
+ "learning_rate": 4.933775582872488e-05,
488
+ "loss": 4.421,
489
+ "step": 40448
490
+ },
491
+ {
492
+ "epoch": 0.01,
493
+ "learning_rate": 4.932936988121436e-05,
494
+ "loss": 4.4193,
495
+ "step": 40960
496
+ },
497
+ {
498
+ "epoch": 0.01,
499
+ "learning_rate": 4.932098393370384e-05,
500
+ "loss": 4.4273,
501
+ "step": 41472
502
+ },
503
+ {
504
+ "epoch": 0.01,
505
+ "learning_rate": 4.931261436499706e-05,
506
+ "loss": 4.4282,
507
+ "step": 41984
508
+ },
509
+ {
510
+ "epoch": 0.01,
511
+ "learning_rate": 4.930422841748654e-05,
512
+ "loss": 4.4108,
513
+ "step": 42496
514
+ },
515
+ {
516
+ "epoch": 0.01,
517
+ "learning_rate": 4.929584246997602e-05,
518
+ "loss": 4.3884,
519
+ "step": 43008
520
+ },
521
+ {
522
+ "epoch": 0.01,
523
+ "learning_rate": 4.92874565224655e-05,
524
+ "loss": 4.4021,
525
+ "step": 43520
526
+ },
527
+ {
528
+ "epoch": 0.01,
529
+ "learning_rate": 4.927907057495498e-05,
530
+ "loss": 4.406,
531
+ "step": 44032
532
+ },
533
+ {
534
+ "epoch": 0.01,
535
+ "learning_rate": 4.927068462744446e-05,
536
+ "loss": 4.3941,
537
+ "step": 44544
538
+ },
539
+ {
540
+ "epoch": 0.01,
541
+ "learning_rate": 4.926229867993394e-05,
542
+ "loss": 4.3996,
543
+ "step": 45056
544
+ },
545
+ {
546
+ "epoch": 0.01,
547
+ "learning_rate": 4.925392911122715e-05,
548
+ "loss": 4.3806,
549
+ "step": 45568
550
+ },
551
+ {
552
+ "epoch": 0.02,
553
+ "learning_rate": 4.924554316371663e-05,
554
+ "loss": 4.3806,
555
+ "step": 46080
556
+ },
557
+ {
558
+ "epoch": 0.02,
559
+ "learning_rate": 4.923715721620611e-05,
560
+ "loss": 4.3754,
561
+ "step": 46592
562
+ },
563
+ {
564
+ "epoch": 0.02,
565
+ "learning_rate": 4.922878764749932e-05,
566
+ "loss": 4.3734,
567
+ "step": 47104
568
+ },
569
+ {
570
+ "epoch": 0.02,
571
+ "learning_rate": 4.92204016999888e-05,
572
+ "loss": 4.3795,
573
+ "step": 47616
574
+ },
575
+ {
576
+ "epoch": 0.02,
577
+ "learning_rate": 4.921201575247828e-05,
578
+ "loss": 4.3668,
579
+ "step": 48128
580
+ },
581
+ {
582
+ "epoch": 0.02,
583
+ "learning_rate": 4.920362980496776e-05,
584
+ "loss": 4.3828,
585
+ "step": 48640
586
+ },
587
+ {
588
+ "epoch": 0.02,
589
+ "learning_rate": 4.919524385745724e-05,
590
+ "loss": 4.3664,
591
+ "step": 49152
592
+ },
593
+ {
594
+ "epoch": 0.02,
595
+ "learning_rate": 4.918685790994672e-05,
596
+ "loss": 4.3492,
597
+ "step": 49664
598
+ },
599
+ {
600
+ "epoch": 0.02,
601
+ "learning_rate": 4.91784719624362e-05,
602
+ "loss": 4.3528,
603
+ "step": 50176
604
+ },
605
+ {
606
+ "epoch": 0.02,
607
+ "learning_rate": 4.9170086014925676e-05,
608
+ "loss": 4.3598,
609
+ "step": 50688
610
+ },
611
+ {
612
+ "epoch": 0.02,
613
+ "learning_rate": 4.916171644621889e-05,
614
+ "loss": 4.3504,
615
+ "step": 51200
616
+ },
617
+ {
618
+ "epoch": 0.02,
619
+ "learning_rate": 4.915333049870837e-05,
620
+ "loss": 4.3481,
621
+ "step": 51712
622
+ },
623
+ {
624
+ "epoch": 0.02,
625
+ "learning_rate": 4.914494455119785e-05,
626
+ "loss": 4.3471,
627
+ "step": 52224
628
+ },
629
+ {
630
+ "epoch": 0.02,
631
+ "learning_rate": 4.9136558603687325e-05,
632
+ "loss": 4.3356,
633
+ "step": 52736
634
+ },
635
+ {
636
+ "epoch": 0.02,
637
+ "learning_rate": 4.9128172656176805e-05,
638
+ "loss": 4.3382,
639
+ "step": 53248
640
+ },
641
+ {
642
+ "epoch": 0.02,
643
+ "learning_rate": 4.9119786708666285e-05,
644
+ "loss": 4.3182,
645
+ "step": 53760
646
+ },
647
+ {
648
+ "epoch": 0.02,
649
+ "learning_rate": 4.9111400761155765e-05,
650
+ "loss": 4.324,
651
+ "step": 54272
652
+ },
653
+ {
654
+ "epoch": 0.02,
655
+ "learning_rate": 4.910303119244898e-05,
656
+ "loss": 4.3292,
657
+ "step": 54784
658
+ },
659
+ {
660
+ "epoch": 0.02,
661
+ "learning_rate": 4.909464524493846e-05,
662
+ "loss": 4.3239,
663
+ "step": 55296
664
+ },
665
+ {
666
+ "epoch": 0.02,
667
+ "learning_rate": 4.908625929742794e-05,
668
+ "loss": 4.3165,
669
+ "step": 55808
670
+ },
671
+ {
672
+ "epoch": 0.02,
673
+ "learning_rate": 4.907787334991742e-05,
674
+ "loss": 4.3255,
675
+ "step": 56320
676
+ },
677
+ {
678
+ "epoch": 0.02,
679
+ "learning_rate": 4.90694874024069e-05,
680
+ "loss": 4.3005,
681
+ "step": 56832
682
+ },
683
+ {
684
+ "epoch": 0.02,
685
+ "learning_rate": 4.906110145489638e-05,
686
+ "loss": 4.3115,
687
+ "step": 57344
688
+ },
689
+ {
690
+ "epoch": 0.02,
691
+ "learning_rate": 4.905273188618959e-05,
692
+ "loss": 4.3071,
693
+ "step": 57856
694
+ },
695
+ {
696
+ "epoch": 0.02,
697
+ "learning_rate": 4.904434593867907e-05,
698
+ "loss": 4.3115,
699
+ "step": 58368
700
+ },
701
+ {
702
+ "epoch": 0.02,
703
+ "learning_rate": 4.903595999116855e-05,
704
+ "loss": 4.3189,
705
+ "step": 58880
706
+ },
707
+ {
708
+ "epoch": 0.02,
709
+ "learning_rate": 4.902757404365803e-05,
710
+ "loss": 4.3064,
711
+ "step": 59392
712
+ },
713
+ {
714
+ "epoch": 0.02,
715
+ "learning_rate": 4.901920447495124e-05,
716
+ "loss": 4.2984,
717
+ "step": 59904
718
+ },
719
+ {
720
+ "epoch": 0.02,
721
+ "learning_rate": 4.901081852744072e-05,
722
+ "loss": 4.2937,
723
+ "step": 60416
724
+ },
725
+ {
726
+ "epoch": 0.02,
727
+ "learning_rate": 4.90024325799302e-05,
728
+ "loss": 4.2998,
729
+ "step": 60928
730
+ },
731
+ {
732
+ "epoch": 0.02,
733
+ "learning_rate": 4.8994046632419686e-05,
734
+ "loss": 4.2947,
735
+ "step": 61440
736
+ },
737
+ {
738
+ "epoch": 0.02,
739
+ "learning_rate": 4.8985660684909166e-05,
740
+ "loss": 4.2915,
741
+ "step": 61952
742
+ },
743
+ {
744
+ "epoch": 0.02,
745
+ "learning_rate": 4.8977274737398646e-05,
746
+ "loss": 4.2846,
747
+ "step": 62464
748
+ },
749
+ {
750
+ "epoch": 0.02,
751
+ "learning_rate": 4.8968888789888125e-05,
752
+ "loss": 4.2819,
753
+ "step": 62976
754
+ },
755
+ {
756
+ "epoch": 0.02,
757
+ "learning_rate": 4.8960519221181335e-05,
758
+ "loss": 4.2823,
759
+ "step": 63488
760
+ },
761
+ {
762
+ "epoch": 0.02,
763
+ "learning_rate": 4.8952133273670815e-05,
764
+ "loss": 4.2714,
765
+ "step": 64000
766
+ },
767
+ {
768
+ "epoch": 0.02,
769
+ "learning_rate": 4.8943747326160294e-05,
770
+ "loss": 4.277,
771
+ "step": 64512
772
+ },
773
+ {
774
+ "epoch": 0.02,
775
+ "learning_rate": 4.8935361378649774e-05,
776
+ "loss": 4.2747,
777
+ "step": 65024
778
+ },
779
+ {
780
+ "epoch": 0.02,
781
+ "learning_rate": 4.8926991809942984e-05,
782
+ "loss": 4.271,
783
+ "step": 65536
784
+ },
785
+ {
786
+ "epoch": 0.02,
787
+ "learning_rate": 4.8918605862432463e-05,
788
+ "loss": 4.2606,
789
+ "step": 66048
790
+ },
791
+ {
792
+ "epoch": 0.02,
793
+ "learning_rate": 4.8910219914921943e-05,
794
+ "loss": 4.279,
795
+ "step": 66560
796
+ },
797
+ {
798
+ "epoch": 0.02,
799
+ "learning_rate": 4.8901833967411423e-05,
800
+ "loss": 4.2791,
801
+ "step": 67072
802
+ },
803
+ {
804
+ "epoch": 0.02,
805
+ "learning_rate": 4.88934480199009e-05,
806
+ "loss": 4.2664,
807
+ "step": 67584
808
+ },
809
+ {
810
+ "epoch": 0.02,
811
+ "learning_rate": 4.888507845119412e-05,
812
+ "loss": 4.2571,
813
+ "step": 68096
814
+ },
815
+ {
816
+ "epoch": 0.02,
817
+ "learning_rate": 4.88766925036836e-05,
818
+ "loss": 4.2637,
819
+ "step": 68608
820
+ },
821
+ {
822
+ "epoch": 0.02,
823
+ "learning_rate": 4.886830655617308e-05,
824
+ "loss": 4.2635,
825
+ "step": 69120
826
+ },
827
+ {
828
+ "epoch": 0.02,
829
+ "learning_rate": 4.885992060866256e-05,
830
+ "loss": 4.2636,
831
+ "step": 69632
832
+ },
833
+ {
834
+ "epoch": 0.02,
835
+ "learning_rate": 4.885153466115204e-05,
836
+ "loss": 4.2589,
837
+ "step": 70144
838
+ },
839
+ {
840
+ "epoch": 0.02,
841
+ "learning_rate": 4.884314871364151e-05,
842
+ "loss": 4.2573,
843
+ "step": 70656
844
+ },
845
+ {
846
+ "epoch": 0.02,
847
+ "learning_rate": 4.883477914493473e-05,
848
+ "loss": 4.2537,
849
+ "step": 71168
850
+ },
851
+ {
852
+ "epoch": 0.02,
853
+ "learning_rate": 4.882639319742421e-05,
854
+ "loss": 4.2512,
855
+ "step": 71680
856
+ },
857
+ {
858
+ "epoch": 0.02,
859
+ "learning_rate": 4.881800724991369e-05,
860
+ "loss": 4.2461,
861
+ "step": 72192
862
+ },
863
+ {
864
+ "epoch": 0.02,
865
+ "learning_rate": 4.880962130240316e-05,
866
+ "loss": 4.2477,
867
+ "step": 72704
868
+ },
869
+ {
870
+ "epoch": 0.02,
871
+ "learning_rate": 4.880123535489264e-05,
872
+ "loss": 4.239,
873
+ "step": 73216
874
+ },
875
+ {
876
+ "epoch": 0.02,
877
+ "learning_rate": 4.879286578618586e-05,
878
+ "loss": 4.2339,
879
+ "step": 73728
880
+ },
881
+ {
882
+ "epoch": 0.02,
883
+ "learning_rate": 4.878447983867534e-05,
884
+ "loss": 4.2479,
885
+ "step": 74240
886
+ },
887
+ {
888
+ "epoch": 0.02,
889
+ "learning_rate": 4.877609389116482e-05,
890
+ "loss": 4.2322,
891
+ "step": 74752
892
+ },
893
+ {
894
+ "epoch": 0.02,
895
+ "learning_rate": 4.87677079436543e-05,
896
+ "loss": 4.2427,
897
+ "step": 75264
898
+ },
899
+ {
900
+ "epoch": 0.02,
901
+ "learning_rate": 4.875932199614378e-05,
902
+ "loss": 4.2285,
903
+ "step": 75776
904
+ },
905
+ {
906
+ "epoch": 0.02,
907
+ "learning_rate": 4.875093604863326e-05,
908
+ "loss": 4.2223,
909
+ "step": 76288
910
+ },
911
+ {
912
+ "epoch": 0.03,
913
+ "eval_loss": 4.196377277374268,
914
+ "eval_runtime": 306.8923,
915
+ "eval_samples_per_second": 1243.404,
916
+ "eval_steps_per_second": 38.857,
917
+ "step": 76320
918
+ }
919
+ ],
920
+ "logging_steps": 512,
921
+ "max_steps": 3052726,
922
+ "num_train_epochs": 9223372036854775807,
923
+ "save_steps": 10,
924
+ "total_flos": 5.322027683621376e+16,
925
+ "trial_name": null,
926
+ "trial_params": null
927
+ }
checkpoint-76320/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bf06229907a00c361bcc45ea31a88489cc0a27e0af480a8dc3b171af1c1e740a
3
+ size 4219