zaas12 commited on
Commit
c2584bd
·
verified ·
1 Parent(s): ae4de02

Upload folder using huggingface_hub

Browse files
checkpoint-3033/config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "EleutherAI/pythia-1.4b",
3
+ "architectures": [
4
+ "GPTNeoXForCausalLM"
5
+ ],
6
+ "attention_bias": true,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 0,
9
+ "classifier_dropout": 0.1,
10
+ "eos_token_id": 0,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout": 0.0,
13
+ "hidden_size": 2048,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 8192,
16
+ "layer_norm_eps": 1e-05,
17
+ "max_position_embeddings": 2048,
18
+ "model_type": "gpt_neox",
19
+ "num_attention_heads": 16,
20
+ "num_hidden_layers": 24,
21
+ "partial_rotary_factor": 0.25,
22
+ "rope_scaling": null,
23
+ "rope_theta": 10000,
24
+ "rotary_emb_base": 10000,
25
+ "rotary_pct": 0.25,
26
+ "tie_word_embeddings": false,
27
+ "torch_dtype": "bfloat16",
28
+ "transformers_version": "4.45.0",
29
+ "use_cache": false,
30
+ "use_parallel_residual": true,
31
+ "vocab_size": 50304
32
+ }
checkpoint-3033/generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "eos_token_id": 0,
5
+ "transformers_version": "4.45.0",
6
+ "use_cache": false
7
+ }
checkpoint-3033/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:781d66818bee006f2e14098bcc4474c37aecd70e5d5bb5f9c5e50a612dfbeca8
3
+ size 2829330208
checkpoint-3033/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b3877808b0bd257d2334e10a97148d1d150b50966074dbb19e231e243bda7678
3
+ size 11317362842
checkpoint-3033/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6baa6f1319458ab835d45b4ee7f546881cf929298434e89057cdfecabc80521c
3
+ size 14244
checkpoint-3033/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c223b69b0867e268d8bb8a6a1b406faafaef12c6e6add1644b0c076c5a734146
3
+ size 1064
checkpoint-3033/trainer_state.json ADDED
@@ -0,0 +1,461 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.5828993320465088,
3
+ "best_model_checkpoint": "./checkpoints/ultrafeedback_binarized/pythia-1.4b-ultrafeedback_binarized-lambda0.25-ORPO-20-9-44/checkpoint-3033",
4
+ "epoch": 1.0,
5
+ "eval_steps": 500,
6
+ "global_step": 3033,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.016485328058028353,
13
+ "grad_norm": 3.171875,
14
+ "learning_rate": 2.0000000000000002e-07,
15
+ "loss": 1.6616,
16
+ "step": 50
17
+ },
18
+ {
19
+ "epoch": 0.03297065611605671,
20
+ "grad_norm": 2.4375,
21
+ "learning_rate": 4.0000000000000003e-07,
22
+ "loss": 1.1362,
23
+ "step": 100
24
+ },
25
+ {
26
+ "epoch": 0.04945598417408507,
27
+ "grad_norm": 3.046875,
28
+ "learning_rate": 6.000000000000001e-07,
29
+ "loss": 1.0031,
30
+ "step": 150
31
+ },
32
+ {
33
+ "epoch": 0.06594131223211341,
34
+ "grad_norm": 4.03125,
35
+ "learning_rate": 8.000000000000001e-07,
36
+ "loss": 1.1643,
37
+ "step": 200
38
+ },
39
+ {
40
+ "epoch": 0.08242664029014177,
41
+ "grad_norm": 5.90625,
42
+ "learning_rate": 1.0000000000000002e-06,
43
+ "loss": 1.3066,
44
+ "step": 250
45
+ },
46
+ {
47
+ "epoch": 0.09891196834817013,
48
+ "grad_norm": 3.203125,
49
+ "learning_rate": 1.2000000000000002e-06,
50
+ "loss": 1.1156,
51
+ "step": 300
52
+ },
53
+ {
54
+ "epoch": 0.11539729640619849,
55
+ "grad_norm": 5.25,
56
+ "learning_rate": 1.4000000000000001e-06,
57
+ "loss": 1.1995,
58
+ "step": 350
59
+ },
60
+ {
61
+ "epoch": 0.13188262446422683,
62
+ "grad_norm": 4.5,
63
+ "learning_rate": 1.6000000000000001e-06,
64
+ "loss": 1.3933,
65
+ "step": 400
66
+ },
67
+ {
68
+ "epoch": 0.14836795252225518,
69
+ "grad_norm": 2.90625,
70
+ "learning_rate": 1.8000000000000001e-06,
71
+ "loss": 1.4222,
72
+ "step": 450
73
+ },
74
+ {
75
+ "epoch": 0.16485328058028353,
76
+ "grad_norm": 2.890625,
77
+ "learning_rate": 2.0000000000000003e-06,
78
+ "loss": 1.2595,
79
+ "step": 500
80
+ },
81
+ {
82
+ "epoch": 0.18133860863831192,
83
+ "grad_norm": 6.28125,
84
+ "learning_rate": 2.2e-06,
85
+ "loss": 1.1306,
86
+ "step": 550
87
+ },
88
+ {
89
+ "epoch": 0.19782393669634027,
90
+ "grad_norm": 2.0,
91
+ "learning_rate": 2.4000000000000003e-06,
92
+ "loss": 1.1537,
93
+ "step": 600
94
+ },
95
+ {
96
+ "epoch": 0.21430926475436862,
97
+ "grad_norm": 1.734375,
98
+ "learning_rate": 2.6e-06,
99
+ "loss": 0.8461,
100
+ "step": 650
101
+ },
102
+ {
103
+ "epoch": 0.23079459281239698,
104
+ "grad_norm": 1.5859375,
105
+ "learning_rate": 2.8000000000000003e-06,
106
+ "loss": 0.8137,
107
+ "step": 700
108
+ },
109
+ {
110
+ "epoch": 0.24727992087042533,
111
+ "grad_norm": 0.97265625,
112
+ "learning_rate": 3e-06,
113
+ "loss": 0.6845,
114
+ "step": 750
115
+ },
116
+ {
117
+ "epoch": 0.26376524892845365,
118
+ "grad_norm": 1.25,
119
+ "learning_rate": 3.2000000000000003e-06,
120
+ "loss": 0.6661,
121
+ "step": 800
122
+ },
123
+ {
124
+ "epoch": 0.28025057698648204,
125
+ "grad_norm": 1.109375,
126
+ "learning_rate": 3.4000000000000005e-06,
127
+ "loss": 0.6483,
128
+ "step": 850
129
+ },
130
+ {
131
+ "epoch": 0.29673590504451036,
132
+ "grad_norm": 1.3125,
133
+ "learning_rate": 3.6000000000000003e-06,
134
+ "loss": 0.6598,
135
+ "step": 900
136
+ },
137
+ {
138
+ "epoch": 0.31322123310253874,
139
+ "grad_norm": 1.3046875,
140
+ "learning_rate": 3.8000000000000005e-06,
141
+ "loss": 0.6339,
142
+ "step": 950
143
+ },
144
+ {
145
+ "epoch": 0.32970656116056707,
146
+ "grad_norm": 1.234375,
147
+ "learning_rate": 4.000000000000001e-06,
148
+ "loss": 0.6764,
149
+ "step": 1000
150
+ },
151
+ {
152
+ "epoch": 0.34619188921859545,
153
+ "grad_norm": 1.3046875,
154
+ "learning_rate": 4.2000000000000004e-06,
155
+ "loss": 0.6544,
156
+ "step": 1050
157
+ },
158
+ {
159
+ "epoch": 0.36267721727662383,
160
+ "grad_norm": 1.171875,
161
+ "learning_rate": 4.4e-06,
162
+ "loss": 0.6414,
163
+ "step": 1100
164
+ },
165
+ {
166
+ "epoch": 0.37916254533465216,
167
+ "grad_norm": 1.1640625,
168
+ "learning_rate": 4.600000000000001e-06,
169
+ "loss": 0.675,
170
+ "step": 1150
171
+ },
172
+ {
173
+ "epoch": 0.39564787339268054,
174
+ "grad_norm": 1.0625,
175
+ "learning_rate": 4.800000000000001e-06,
176
+ "loss": 0.6089,
177
+ "step": 1200
178
+ },
179
+ {
180
+ "epoch": 0.41213320145070886,
181
+ "grad_norm": 1.3125,
182
+ "learning_rate": 5e-06,
183
+ "loss": 0.6111,
184
+ "step": 1250
185
+ },
186
+ {
187
+ "epoch": 0.42861852950873724,
188
+ "grad_norm": 1.265625,
189
+ "learning_rate": 5.2e-06,
190
+ "loss": 0.6293,
191
+ "step": 1300
192
+ },
193
+ {
194
+ "epoch": 0.44510385756676557,
195
+ "grad_norm": 1.078125,
196
+ "learning_rate": 5.400000000000001e-06,
197
+ "loss": 0.6115,
198
+ "step": 1350
199
+ },
200
+ {
201
+ "epoch": 0.46158918562479395,
202
+ "grad_norm": 1.1796875,
203
+ "learning_rate": 5.600000000000001e-06,
204
+ "loss": 0.6035,
205
+ "step": 1400
206
+ },
207
+ {
208
+ "epoch": 0.4780745136828223,
209
+ "grad_norm": 1.2421875,
210
+ "learning_rate": 5.8e-06,
211
+ "loss": 0.6247,
212
+ "step": 1450
213
+ },
214
+ {
215
+ "epoch": 0.49455984174085066,
216
+ "grad_norm": 1.25,
217
+ "learning_rate": 6e-06,
218
+ "loss": 0.6121,
219
+ "step": 1500
220
+ },
221
+ {
222
+ "epoch": 0.511045169798879,
223
+ "grad_norm": 1.1796875,
224
+ "learning_rate": 6.200000000000001e-06,
225
+ "loss": 0.5945,
226
+ "step": 1550
227
+ },
228
+ {
229
+ "epoch": 0.5275304978569073,
230
+ "grad_norm": 1.1796875,
231
+ "learning_rate": 6.4000000000000006e-06,
232
+ "loss": 0.5971,
233
+ "step": 1600
234
+ },
235
+ {
236
+ "epoch": 0.5440158259149357,
237
+ "grad_norm": 1.1953125,
238
+ "learning_rate": 6.600000000000001e-06,
239
+ "loss": 0.6137,
240
+ "step": 1650
241
+ },
242
+ {
243
+ "epoch": 0.5605011539729641,
244
+ "grad_norm": 1.15625,
245
+ "learning_rate": 6.800000000000001e-06,
246
+ "loss": 0.6205,
247
+ "step": 1700
248
+ },
249
+ {
250
+ "epoch": 0.5769864820309925,
251
+ "grad_norm": 1.03125,
252
+ "learning_rate": 7e-06,
253
+ "loss": 0.6136,
254
+ "step": 1750
255
+ },
256
+ {
257
+ "epoch": 0.5934718100890207,
258
+ "grad_norm": 1.078125,
259
+ "learning_rate": 7.2000000000000005e-06,
260
+ "loss": 0.6166,
261
+ "step": 1800
262
+ },
263
+ {
264
+ "epoch": 0.6099571381470491,
265
+ "grad_norm": 1.0703125,
266
+ "learning_rate": 7.4e-06,
267
+ "loss": 0.6011,
268
+ "step": 1850
269
+ },
270
+ {
271
+ "epoch": 0.6264424662050775,
272
+ "grad_norm": 1.140625,
273
+ "learning_rate": 7.600000000000001e-06,
274
+ "loss": 0.6205,
275
+ "step": 1900
276
+ },
277
+ {
278
+ "epoch": 0.6429277942631059,
279
+ "grad_norm": 1.234375,
280
+ "learning_rate": 7.800000000000002e-06,
281
+ "loss": 0.6323,
282
+ "step": 1950
283
+ },
284
+ {
285
+ "epoch": 0.6594131223211341,
286
+ "grad_norm": 1.1328125,
287
+ "learning_rate": 8.000000000000001e-06,
288
+ "loss": 0.6075,
289
+ "step": 2000
290
+ },
291
+ {
292
+ "epoch": 0.6758984503791625,
293
+ "grad_norm": 1.3671875,
294
+ "learning_rate": 8.2e-06,
295
+ "loss": 0.6165,
296
+ "step": 2050
297
+ },
298
+ {
299
+ "epoch": 0.6923837784371909,
300
+ "grad_norm": 1.203125,
301
+ "learning_rate": 8.400000000000001e-06,
302
+ "loss": 0.6248,
303
+ "step": 2100
304
+ },
305
+ {
306
+ "epoch": 0.7088691064952193,
307
+ "grad_norm": 1.2578125,
308
+ "learning_rate": 8.6e-06,
309
+ "loss": 0.6098,
310
+ "step": 2150
311
+ },
312
+ {
313
+ "epoch": 0.7253544345532477,
314
+ "grad_norm": 1.1796875,
315
+ "learning_rate": 8.8e-06,
316
+ "loss": 0.6096,
317
+ "step": 2200
318
+ },
319
+ {
320
+ "epoch": 0.7418397626112759,
321
+ "grad_norm": 1.296875,
322
+ "learning_rate": 9e-06,
323
+ "loss": 0.5938,
324
+ "step": 2250
325
+ },
326
+ {
327
+ "epoch": 0.7583250906693043,
328
+ "grad_norm": 1.234375,
329
+ "learning_rate": 9.200000000000002e-06,
330
+ "loss": 0.5825,
331
+ "step": 2300
332
+ },
333
+ {
334
+ "epoch": 0.7748104187273327,
335
+ "grad_norm": 1.265625,
336
+ "learning_rate": 9.4e-06,
337
+ "loss": 0.5849,
338
+ "step": 2350
339
+ },
340
+ {
341
+ "epoch": 0.7912957467853611,
342
+ "grad_norm": 1.203125,
343
+ "learning_rate": 9.600000000000001e-06,
344
+ "loss": 0.5872,
345
+ "step": 2400
346
+ },
347
+ {
348
+ "epoch": 0.8077810748433893,
349
+ "grad_norm": 1.0703125,
350
+ "learning_rate": 9.800000000000001e-06,
351
+ "loss": 0.5864,
352
+ "step": 2450
353
+ },
354
+ {
355
+ "epoch": 0.8242664029014177,
356
+ "grad_norm": 1.1015625,
357
+ "learning_rate": 1e-05,
358
+ "loss": 0.5809,
359
+ "step": 2500
360
+ },
361
+ {
362
+ "epoch": 0.8407517309594461,
363
+ "grad_norm": 1.0390625,
364
+ "learning_rate": 1.02e-05,
365
+ "loss": 0.5855,
366
+ "step": 2550
367
+ },
368
+ {
369
+ "epoch": 0.8572370590174745,
370
+ "grad_norm": 1.1875,
371
+ "learning_rate": 1.04e-05,
372
+ "loss": 0.607,
373
+ "step": 2600
374
+ },
375
+ {
376
+ "epoch": 0.8737223870755028,
377
+ "grad_norm": 1.015625,
378
+ "learning_rate": 1.0600000000000002e-05,
379
+ "loss": 0.5834,
380
+ "step": 2650
381
+ },
382
+ {
383
+ "epoch": 0.8902077151335311,
384
+ "grad_norm": 1.2578125,
385
+ "learning_rate": 1.0800000000000002e-05,
386
+ "loss": 0.5902,
387
+ "step": 2700
388
+ },
389
+ {
390
+ "epoch": 0.9066930431915595,
391
+ "grad_norm": 1.1796875,
392
+ "learning_rate": 1.1000000000000001e-05,
393
+ "loss": 0.6158,
394
+ "step": 2750
395
+ },
396
+ {
397
+ "epoch": 0.9231783712495879,
398
+ "grad_norm": 1.234375,
399
+ "learning_rate": 1.1200000000000001e-05,
400
+ "loss": 0.5687,
401
+ "step": 2800
402
+ },
403
+ {
404
+ "epoch": 0.9396636993076162,
405
+ "grad_norm": 1.0625,
406
+ "learning_rate": 1.14e-05,
407
+ "loss": 0.5906,
408
+ "step": 2850
409
+ },
410
+ {
411
+ "epoch": 0.9561490273656446,
412
+ "grad_norm": 1.0078125,
413
+ "learning_rate": 1.16e-05,
414
+ "loss": 0.6102,
415
+ "step": 2900
416
+ },
417
+ {
418
+ "epoch": 0.9726343554236729,
419
+ "grad_norm": 1.1328125,
420
+ "learning_rate": 1.18e-05,
421
+ "loss": 0.5748,
422
+ "step": 2950
423
+ },
424
+ {
425
+ "epoch": 0.9891196834817013,
426
+ "grad_norm": 1.078125,
427
+ "learning_rate": 1.2e-05,
428
+ "loss": 0.5793,
429
+ "step": 3000
430
+ },
431
+ {
432
+ "epoch": 1.0,
433
+ "eval_loss": 0.5828993320465088,
434
+ "eval_runtime": 76.6531,
435
+ "eval_samples_per_second": 21.252,
436
+ "eval_steps_per_second": 1.331,
437
+ "step": 3033
438
+ }
439
+ ],
440
+ "logging_steps": 50,
441
+ "max_steps": 6066,
442
+ "num_input_tokens_seen": 0,
443
+ "num_train_epochs": 2,
444
+ "save_steps": 500,
445
+ "stateful_callbacks": {
446
+ "TrainerControl": {
447
+ "args": {
448
+ "should_epoch_stop": false,
449
+ "should_evaluate": false,
450
+ "should_log": false,
451
+ "should_save": true,
452
+ "should_training_stop": false
453
+ },
454
+ "attributes": {}
455
+ }
456
+ },
457
+ "total_flos": 3.909883836204319e+17,
458
+ "train_batch_size": 16,
459
+ "trial_name": null,
460
+ "trial_params": null
461
+ }
checkpoint-3033/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e0675a81995d37478683f01d22ce39f1a71fbb9775020cfc1de7932d0dabdb9
3
+ size 5432
checkpoint-6066/config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "EleutherAI/pythia-1.4b",
3
+ "architectures": [
4
+ "GPTNeoXForCausalLM"
5
+ ],
6
+ "attention_bias": true,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 0,
9
+ "classifier_dropout": 0.1,
10
+ "eos_token_id": 0,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout": 0.0,
13
+ "hidden_size": 2048,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 8192,
16
+ "layer_norm_eps": 1e-05,
17
+ "max_position_embeddings": 2048,
18
+ "model_type": "gpt_neox",
19
+ "num_attention_heads": 16,
20
+ "num_hidden_layers": 24,
21
+ "partial_rotary_factor": 0.25,
22
+ "rope_scaling": null,
23
+ "rope_theta": 10000,
24
+ "rotary_emb_base": 10000,
25
+ "rotary_pct": 0.25,
26
+ "tie_word_embeddings": false,
27
+ "torch_dtype": "bfloat16",
28
+ "transformers_version": "4.45.0",
29
+ "use_cache": false,
30
+ "use_parallel_residual": true,
31
+ "vocab_size": 50304
32
+ }
checkpoint-6066/generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "eos_token_id": 0,
5
+ "transformers_version": "4.45.0",
6
+ "use_cache": false
7
+ }
checkpoint-6066/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac8a5ad2319c9ddc5c45584371ff1c8f83b721b9a9794beb97376e606a185fe2
3
+ size 2829330208
checkpoint-6066/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80173f1f6ba97480bc9ea51e0d2cd9e8dcffcc88699a0b5d6354b4b26d8e0dad
3
+ size 11317362842
checkpoint-6066/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3aebab2d660d9d8cf7ffa59b71e5ddf124bdeb2f6c12f7f02bce490047f93714
3
+ size 14244
checkpoint-6066/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fbb511d3311a10a9d4ce10cb1f8968d168ae324c656df94c0496ee483fb81659
3
+ size 1064
checkpoint-6066/trainer_state.json ADDED
@@ -0,0 +1,896 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.5490379333496094,
3
+ "best_model_checkpoint": "./checkpoints/ultrafeedback_binarized/pythia-1.4b-ultrafeedback_binarized-lambda0.25-ORPO-20-9-44/checkpoint-6066",
4
+ "epoch": 2.0,
5
+ "eval_steps": 500,
6
+ "global_step": 6066,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.016485328058028353,
13
+ "grad_norm": 3.171875,
14
+ "learning_rate": 2.0000000000000002e-07,
15
+ "loss": 1.6616,
16
+ "step": 50
17
+ },
18
+ {
19
+ "epoch": 0.03297065611605671,
20
+ "grad_norm": 2.4375,
21
+ "learning_rate": 4.0000000000000003e-07,
22
+ "loss": 1.1362,
23
+ "step": 100
24
+ },
25
+ {
26
+ "epoch": 0.04945598417408507,
27
+ "grad_norm": 3.046875,
28
+ "learning_rate": 6.000000000000001e-07,
29
+ "loss": 1.0031,
30
+ "step": 150
31
+ },
32
+ {
33
+ "epoch": 0.06594131223211341,
34
+ "grad_norm": 4.03125,
35
+ "learning_rate": 8.000000000000001e-07,
36
+ "loss": 1.1643,
37
+ "step": 200
38
+ },
39
+ {
40
+ "epoch": 0.08242664029014177,
41
+ "grad_norm": 5.90625,
42
+ "learning_rate": 1.0000000000000002e-06,
43
+ "loss": 1.3066,
44
+ "step": 250
45
+ },
46
+ {
47
+ "epoch": 0.09891196834817013,
48
+ "grad_norm": 3.203125,
49
+ "learning_rate": 1.2000000000000002e-06,
50
+ "loss": 1.1156,
51
+ "step": 300
52
+ },
53
+ {
54
+ "epoch": 0.11539729640619849,
55
+ "grad_norm": 5.25,
56
+ "learning_rate": 1.4000000000000001e-06,
57
+ "loss": 1.1995,
58
+ "step": 350
59
+ },
60
+ {
61
+ "epoch": 0.13188262446422683,
62
+ "grad_norm": 4.5,
63
+ "learning_rate": 1.6000000000000001e-06,
64
+ "loss": 1.3933,
65
+ "step": 400
66
+ },
67
+ {
68
+ "epoch": 0.14836795252225518,
69
+ "grad_norm": 2.90625,
70
+ "learning_rate": 1.8000000000000001e-06,
71
+ "loss": 1.4222,
72
+ "step": 450
73
+ },
74
+ {
75
+ "epoch": 0.16485328058028353,
76
+ "grad_norm": 2.890625,
77
+ "learning_rate": 2.0000000000000003e-06,
78
+ "loss": 1.2595,
79
+ "step": 500
80
+ },
81
+ {
82
+ "epoch": 0.18133860863831192,
83
+ "grad_norm": 6.28125,
84
+ "learning_rate": 2.2e-06,
85
+ "loss": 1.1306,
86
+ "step": 550
87
+ },
88
+ {
89
+ "epoch": 0.19782393669634027,
90
+ "grad_norm": 2.0,
91
+ "learning_rate": 2.4000000000000003e-06,
92
+ "loss": 1.1537,
93
+ "step": 600
94
+ },
95
+ {
96
+ "epoch": 0.21430926475436862,
97
+ "grad_norm": 1.734375,
98
+ "learning_rate": 2.6e-06,
99
+ "loss": 0.8461,
100
+ "step": 650
101
+ },
102
+ {
103
+ "epoch": 0.23079459281239698,
104
+ "grad_norm": 1.5859375,
105
+ "learning_rate": 2.8000000000000003e-06,
106
+ "loss": 0.8137,
107
+ "step": 700
108
+ },
109
+ {
110
+ "epoch": 0.24727992087042533,
111
+ "grad_norm": 0.97265625,
112
+ "learning_rate": 3e-06,
113
+ "loss": 0.6845,
114
+ "step": 750
115
+ },
116
+ {
117
+ "epoch": 0.26376524892845365,
118
+ "grad_norm": 1.25,
119
+ "learning_rate": 3.2000000000000003e-06,
120
+ "loss": 0.6661,
121
+ "step": 800
122
+ },
123
+ {
124
+ "epoch": 0.28025057698648204,
125
+ "grad_norm": 1.109375,
126
+ "learning_rate": 3.4000000000000005e-06,
127
+ "loss": 0.6483,
128
+ "step": 850
129
+ },
130
+ {
131
+ "epoch": 0.29673590504451036,
132
+ "grad_norm": 1.3125,
133
+ "learning_rate": 3.6000000000000003e-06,
134
+ "loss": 0.6598,
135
+ "step": 900
136
+ },
137
+ {
138
+ "epoch": 0.31322123310253874,
139
+ "grad_norm": 1.3046875,
140
+ "learning_rate": 3.8000000000000005e-06,
141
+ "loss": 0.6339,
142
+ "step": 950
143
+ },
144
+ {
145
+ "epoch": 0.32970656116056707,
146
+ "grad_norm": 1.234375,
147
+ "learning_rate": 4.000000000000001e-06,
148
+ "loss": 0.6764,
149
+ "step": 1000
150
+ },
151
+ {
152
+ "epoch": 0.34619188921859545,
153
+ "grad_norm": 1.3046875,
154
+ "learning_rate": 4.2000000000000004e-06,
155
+ "loss": 0.6544,
156
+ "step": 1050
157
+ },
158
+ {
159
+ "epoch": 0.36267721727662383,
160
+ "grad_norm": 1.171875,
161
+ "learning_rate": 4.4e-06,
162
+ "loss": 0.6414,
163
+ "step": 1100
164
+ },
165
+ {
166
+ "epoch": 0.37916254533465216,
167
+ "grad_norm": 1.1640625,
168
+ "learning_rate": 4.600000000000001e-06,
169
+ "loss": 0.675,
170
+ "step": 1150
171
+ },
172
+ {
173
+ "epoch": 0.39564787339268054,
174
+ "grad_norm": 1.0625,
175
+ "learning_rate": 4.800000000000001e-06,
176
+ "loss": 0.6089,
177
+ "step": 1200
178
+ },
179
+ {
180
+ "epoch": 0.41213320145070886,
181
+ "grad_norm": 1.3125,
182
+ "learning_rate": 5e-06,
183
+ "loss": 0.6111,
184
+ "step": 1250
185
+ },
186
+ {
187
+ "epoch": 0.42861852950873724,
188
+ "grad_norm": 1.265625,
189
+ "learning_rate": 5.2e-06,
190
+ "loss": 0.6293,
191
+ "step": 1300
192
+ },
193
+ {
194
+ "epoch": 0.44510385756676557,
195
+ "grad_norm": 1.078125,
196
+ "learning_rate": 5.400000000000001e-06,
197
+ "loss": 0.6115,
198
+ "step": 1350
199
+ },
200
+ {
201
+ "epoch": 0.46158918562479395,
202
+ "grad_norm": 1.1796875,
203
+ "learning_rate": 5.600000000000001e-06,
204
+ "loss": 0.6035,
205
+ "step": 1400
206
+ },
207
+ {
208
+ "epoch": 0.4780745136828223,
209
+ "grad_norm": 1.2421875,
210
+ "learning_rate": 5.8e-06,
211
+ "loss": 0.6247,
212
+ "step": 1450
213
+ },
214
+ {
215
+ "epoch": 0.49455984174085066,
216
+ "grad_norm": 1.25,
217
+ "learning_rate": 6e-06,
218
+ "loss": 0.6121,
219
+ "step": 1500
220
+ },
221
+ {
222
+ "epoch": 0.511045169798879,
223
+ "grad_norm": 1.1796875,
224
+ "learning_rate": 6.200000000000001e-06,
225
+ "loss": 0.5945,
226
+ "step": 1550
227
+ },
228
+ {
229
+ "epoch": 0.5275304978569073,
230
+ "grad_norm": 1.1796875,
231
+ "learning_rate": 6.4000000000000006e-06,
232
+ "loss": 0.5971,
233
+ "step": 1600
234
+ },
235
+ {
236
+ "epoch": 0.5440158259149357,
237
+ "grad_norm": 1.1953125,
238
+ "learning_rate": 6.600000000000001e-06,
239
+ "loss": 0.6137,
240
+ "step": 1650
241
+ },
242
+ {
243
+ "epoch": 0.5605011539729641,
244
+ "grad_norm": 1.15625,
245
+ "learning_rate": 6.800000000000001e-06,
246
+ "loss": 0.6205,
247
+ "step": 1700
248
+ },
249
+ {
250
+ "epoch": 0.5769864820309925,
251
+ "grad_norm": 1.03125,
252
+ "learning_rate": 7e-06,
253
+ "loss": 0.6136,
254
+ "step": 1750
255
+ },
256
+ {
257
+ "epoch": 0.5934718100890207,
258
+ "grad_norm": 1.078125,
259
+ "learning_rate": 7.2000000000000005e-06,
260
+ "loss": 0.6166,
261
+ "step": 1800
262
+ },
263
+ {
264
+ "epoch": 0.6099571381470491,
265
+ "grad_norm": 1.0703125,
266
+ "learning_rate": 7.4e-06,
267
+ "loss": 0.6011,
268
+ "step": 1850
269
+ },
270
+ {
271
+ "epoch": 0.6264424662050775,
272
+ "grad_norm": 1.140625,
273
+ "learning_rate": 7.600000000000001e-06,
274
+ "loss": 0.6205,
275
+ "step": 1900
276
+ },
277
+ {
278
+ "epoch": 0.6429277942631059,
279
+ "grad_norm": 1.234375,
280
+ "learning_rate": 7.800000000000002e-06,
281
+ "loss": 0.6323,
282
+ "step": 1950
283
+ },
284
+ {
285
+ "epoch": 0.6594131223211341,
286
+ "grad_norm": 1.1328125,
287
+ "learning_rate": 8.000000000000001e-06,
288
+ "loss": 0.6075,
289
+ "step": 2000
290
+ },
291
+ {
292
+ "epoch": 0.6758984503791625,
293
+ "grad_norm": 1.3671875,
294
+ "learning_rate": 8.2e-06,
295
+ "loss": 0.6165,
296
+ "step": 2050
297
+ },
298
+ {
299
+ "epoch": 0.6923837784371909,
300
+ "grad_norm": 1.203125,
301
+ "learning_rate": 8.400000000000001e-06,
302
+ "loss": 0.6248,
303
+ "step": 2100
304
+ },
305
+ {
306
+ "epoch": 0.7088691064952193,
307
+ "grad_norm": 1.2578125,
308
+ "learning_rate": 8.6e-06,
309
+ "loss": 0.6098,
310
+ "step": 2150
311
+ },
312
+ {
313
+ "epoch": 0.7253544345532477,
314
+ "grad_norm": 1.1796875,
315
+ "learning_rate": 8.8e-06,
316
+ "loss": 0.6096,
317
+ "step": 2200
318
+ },
319
+ {
320
+ "epoch": 0.7418397626112759,
321
+ "grad_norm": 1.296875,
322
+ "learning_rate": 9e-06,
323
+ "loss": 0.5938,
324
+ "step": 2250
325
+ },
326
+ {
327
+ "epoch": 0.7583250906693043,
328
+ "grad_norm": 1.234375,
329
+ "learning_rate": 9.200000000000002e-06,
330
+ "loss": 0.5825,
331
+ "step": 2300
332
+ },
333
+ {
334
+ "epoch": 0.7748104187273327,
335
+ "grad_norm": 1.265625,
336
+ "learning_rate": 9.4e-06,
337
+ "loss": 0.5849,
338
+ "step": 2350
339
+ },
340
+ {
341
+ "epoch": 0.7912957467853611,
342
+ "grad_norm": 1.203125,
343
+ "learning_rate": 9.600000000000001e-06,
344
+ "loss": 0.5872,
345
+ "step": 2400
346
+ },
347
+ {
348
+ "epoch": 0.8077810748433893,
349
+ "grad_norm": 1.0703125,
350
+ "learning_rate": 9.800000000000001e-06,
351
+ "loss": 0.5864,
352
+ "step": 2450
353
+ },
354
+ {
355
+ "epoch": 0.8242664029014177,
356
+ "grad_norm": 1.1015625,
357
+ "learning_rate": 1e-05,
358
+ "loss": 0.5809,
359
+ "step": 2500
360
+ },
361
+ {
362
+ "epoch": 0.8407517309594461,
363
+ "grad_norm": 1.0390625,
364
+ "learning_rate": 1.02e-05,
365
+ "loss": 0.5855,
366
+ "step": 2550
367
+ },
368
+ {
369
+ "epoch": 0.8572370590174745,
370
+ "grad_norm": 1.1875,
371
+ "learning_rate": 1.04e-05,
372
+ "loss": 0.607,
373
+ "step": 2600
374
+ },
375
+ {
376
+ "epoch": 0.8737223870755028,
377
+ "grad_norm": 1.015625,
378
+ "learning_rate": 1.0600000000000002e-05,
379
+ "loss": 0.5834,
380
+ "step": 2650
381
+ },
382
+ {
383
+ "epoch": 0.8902077151335311,
384
+ "grad_norm": 1.2578125,
385
+ "learning_rate": 1.0800000000000002e-05,
386
+ "loss": 0.5902,
387
+ "step": 2700
388
+ },
389
+ {
390
+ "epoch": 0.9066930431915595,
391
+ "grad_norm": 1.1796875,
392
+ "learning_rate": 1.1000000000000001e-05,
393
+ "loss": 0.6158,
394
+ "step": 2750
395
+ },
396
+ {
397
+ "epoch": 0.9231783712495879,
398
+ "grad_norm": 1.234375,
399
+ "learning_rate": 1.1200000000000001e-05,
400
+ "loss": 0.5687,
401
+ "step": 2800
402
+ },
403
+ {
404
+ "epoch": 0.9396636993076162,
405
+ "grad_norm": 1.0625,
406
+ "learning_rate": 1.14e-05,
407
+ "loss": 0.5906,
408
+ "step": 2850
409
+ },
410
+ {
411
+ "epoch": 0.9561490273656446,
412
+ "grad_norm": 1.0078125,
413
+ "learning_rate": 1.16e-05,
414
+ "loss": 0.6102,
415
+ "step": 2900
416
+ },
417
+ {
418
+ "epoch": 0.9726343554236729,
419
+ "grad_norm": 1.1328125,
420
+ "learning_rate": 1.18e-05,
421
+ "loss": 0.5748,
422
+ "step": 2950
423
+ },
424
+ {
425
+ "epoch": 0.9891196834817013,
426
+ "grad_norm": 1.078125,
427
+ "learning_rate": 1.2e-05,
428
+ "loss": 0.5793,
429
+ "step": 3000
430
+ },
431
+ {
432
+ "epoch": 1.0,
433
+ "eval_loss": 0.5828993320465088,
434
+ "eval_runtime": 76.6531,
435
+ "eval_samples_per_second": 21.252,
436
+ "eval_steps_per_second": 1.331,
437
+ "step": 3033
438
+ },
439
+ {
440
+ "epoch": 1.0056050115397297,
441
+ "grad_norm": 1.3515625,
442
+ "learning_rate": 1.22e-05,
443
+ "loss": 0.6107,
444
+ "step": 3050
445
+ },
446
+ {
447
+ "epoch": 1.022090339597758,
448
+ "grad_norm": 1.2265625,
449
+ "learning_rate": 1.2400000000000002e-05,
450
+ "loss": 0.595,
451
+ "step": 3100
452
+ },
453
+ {
454
+ "epoch": 1.0385756676557865,
455
+ "grad_norm": 1.1328125,
456
+ "learning_rate": 1.2600000000000001e-05,
457
+ "loss": 0.5787,
458
+ "step": 3150
459
+ },
460
+ {
461
+ "epoch": 1.0550609957138146,
462
+ "grad_norm": 1.0078125,
463
+ "learning_rate": 1.2800000000000001e-05,
464
+ "loss": 0.5655,
465
+ "step": 3200
466
+ },
467
+ {
468
+ "epoch": 1.071546323771843,
469
+ "grad_norm": 1.15625,
470
+ "learning_rate": 1.3000000000000001e-05,
471
+ "loss": 0.5676,
472
+ "step": 3250
473
+ },
474
+ {
475
+ "epoch": 1.0880316518298714,
476
+ "grad_norm": 1.171875,
477
+ "learning_rate": 1.3200000000000002e-05,
478
+ "loss": 0.592,
479
+ "step": 3300
480
+ },
481
+ {
482
+ "epoch": 1.1045169798878998,
483
+ "grad_norm": 0.96875,
484
+ "learning_rate": 1.3400000000000002e-05,
485
+ "loss": 0.5779,
486
+ "step": 3350
487
+ },
488
+ {
489
+ "epoch": 1.1210023079459281,
490
+ "grad_norm": 1.21875,
491
+ "learning_rate": 1.3600000000000002e-05,
492
+ "loss": 0.5848,
493
+ "step": 3400
494
+ },
495
+ {
496
+ "epoch": 1.1374876360039565,
497
+ "grad_norm": 1.2109375,
498
+ "learning_rate": 1.38e-05,
499
+ "loss": 0.5635,
500
+ "step": 3450
501
+ },
502
+ {
503
+ "epoch": 1.153972964061985,
504
+ "grad_norm": 1.2109375,
505
+ "learning_rate": 1.4e-05,
506
+ "loss": 0.587,
507
+ "step": 3500
508
+ },
509
+ {
510
+ "epoch": 1.170458292120013,
511
+ "grad_norm": 1.0546875,
512
+ "learning_rate": 1.4200000000000001e-05,
513
+ "loss": 0.5734,
514
+ "step": 3550
515
+ },
516
+ {
517
+ "epoch": 1.1869436201780414,
518
+ "grad_norm": 1.1953125,
519
+ "learning_rate": 1.4400000000000001e-05,
520
+ "loss": 0.5658,
521
+ "step": 3600
522
+ },
523
+ {
524
+ "epoch": 1.2034289482360698,
525
+ "grad_norm": 0.9921875,
526
+ "learning_rate": 1.46e-05,
527
+ "loss": 0.5762,
528
+ "step": 3650
529
+ },
530
+ {
531
+ "epoch": 1.2199142762940982,
532
+ "grad_norm": 1.0234375,
533
+ "learning_rate": 1.48e-05,
534
+ "loss": 0.5835,
535
+ "step": 3700
536
+ },
537
+ {
538
+ "epoch": 1.2363996043521266,
539
+ "grad_norm": 1.0703125,
540
+ "learning_rate": 1.5000000000000002e-05,
541
+ "loss": 0.5689,
542
+ "step": 3750
543
+ },
544
+ {
545
+ "epoch": 1.252884932410155,
546
+ "grad_norm": 1.21875,
547
+ "learning_rate": 1.5200000000000002e-05,
548
+ "loss": 0.5629,
549
+ "step": 3800
550
+ },
551
+ {
552
+ "epoch": 1.2693702604681834,
553
+ "grad_norm": 1.0546875,
554
+ "learning_rate": 1.54e-05,
555
+ "loss": 0.5587,
556
+ "step": 3850
557
+ },
558
+ {
559
+ "epoch": 1.2858555885262117,
560
+ "grad_norm": 1.046875,
561
+ "learning_rate": 1.5600000000000003e-05,
562
+ "loss": 0.5995,
563
+ "step": 3900
564
+ },
565
+ {
566
+ "epoch": 1.3023409165842401,
567
+ "grad_norm": 1.1640625,
568
+ "learning_rate": 1.58e-05,
569
+ "loss": 0.569,
570
+ "step": 3950
571
+ },
572
+ {
573
+ "epoch": 1.3188262446422683,
574
+ "grad_norm": 1.171875,
575
+ "learning_rate": 1.6000000000000003e-05,
576
+ "loss": 0.5836,
577
+ "step": 4000
578
+ },
579
+ {
580
+ "epoch": 1.3353115727002967,
581
+ "grad_norm": 1.125,
582
+ "learning_rate": 1.62e-05,
583
+ "loss": 0.5539,
584
+ "step": 4050
585
+ },
586
+ {
587
+ "epoch": 1.351796900758325,
588
+ "grad_norm": 1.1171875,
589
+ "learning_rate": 1.64e-05,
590
+ "loss": 0.5569,
591
+ "step": 4100
592
+ },
593
+ {
594
+ "epoch": 1.3682822288163534,
595
+ "grad_norm": 1.1171875,
596
+ "learning_rate": 1.66e-05,
597
+ "loss": 0.5645,
598
+ "step": 4150
599
+ },
600
+ {
601
+ "epoch": 1.3847675568743818,
602
+ "grad_norm": 1.1953125,
603
+ "learning_rate": 1.6800000000000002e-05,
604
+ "loss": 0.5806,
605
+ "step": 4200
606
+ },
607
+ {
608
+ "epoch": 1.4012528849324102,
609
+ "grad_norm": 1.0625,
610
+ "learning_rate": 1.7e-05,
611
+ "loss": 0.5859,
612
+ "step": 4250
613
+ },
614
+ {
615
+ "epoch": 1.4177382129904386,
616
+ "grad_norm": 1.2109375,
617
+ "learning_rate": 1.72e-05,
618
+ "loss": 0.5813,
619
+ "step": 4300
620
+ },
621
+ {
622
+ "epoch": 1.434223541048467,
623
+ "grad_norm": 1.0859375,
624
+ "learning_rate": 1.7400000000000003e-05,
625
+ "loss": 0.5492,
626
+ "step": 4350
627
+ },
628
+ {
629
+ "epoch": 1.4507088691064953,
630
+ "grad_norm": 1.140625,
631
+ "learning_rate": 1.76e-05,
632
+ "loss": 0.5555,
633
+ "step": 4400
634
+ },
635
+ {
636
+ "epoch": 1.4671941971645235,
637
+ "grad_norm": 1.1015625,
638
+ "learning_rate": 1.7800000000000002e-05,
639
+ "loss": 0.5586,
640
+ "step": 4450
641
+ },
642
+ {
643
+ "epoch": 1.4836795252225519,
644
+ "grad_norm": 1.09375,
645
+ "learning_rate": 1.8e-05,
646
+ "loss": 0.5593,
647
+ "step": 4500
648
+ },
649
+ {
650
+ "epoch": 1.5001648532805802,
651
+ "grad_norm": 0.99609375,
652
+ "learning_rate": 1.8200000000000002e-05,
653
+ "loss": 0.5557,
654
+ "step": 4550
655
+ },
656
+ {
657
+ "epoch": 1.5166501813386086,
658
+ "grad_norm": 1.1015625,
659
+ "learning_rate": 1.8400000000000003e-05,
660
+ "loss": 0.5831,
661
+ "step": 4600
662
+ },
663
+ {
664
+ "epoch": 1.533135509396637,
665
+ "grad_norm": 0.92578125,
666
+ "learning_rate": 1.86e-05,
667
+ "loss": 0.5535,
668
+ "step": 4650
669
+ },
670
+ {
671
+ "epoch": 1.5496208374546654,
672
+ "grad_norm": 1.125,
673
+ "learning_rate": 1.88e-05,
674
+ "loss": 0.55,
675
+ "step": 4700
676
+ },
677
+ {
678
+ "epoch": 1.5661061655126938,
679
+ "grad_norm": 1.2265625,
680
+ "learning_rate": 1.9e-05,
681
+ "loss": 0.5447,
682
+ "step": 4750
683
+ },
684
+ {
685
+ "epoch": 1.582591493570722,
686
+ "grad_norm": 1.265625,
687
+ "learning_rate": 1.9200000000000003e-05,
688
+ "loss": 0.5746,
689
+ "step": 4800
690
+ },
691
+ {
692
+ "epoch": 1.5990768216287505,
693
+ "grad_norm": 1.0546875,
694
+ "learning_rate": 1.94e-05,
695
+ "loss": 0.5611,
696
+ "step": 4850
697
+ },
698
+ {
699
+ "epoch": 1.6155621496867787,
700
+ "grad_norm": 1.015625,
701
+ "learning_rate": 1.9600000000000002e-05,
702
+ "loss": 0.5442,
703
+ "step": 4900
704
+ },
705
+ {
706
+ "epoch": 1.632047477744807,
707
+ "grad_norm": 1.1484375,
708
+ "learning_rate": 1.98e-05,
709
+ "loss": 0.555,
710
+ "step": 4950
711
+ },
712
+ {
713
+ "epoch": 1.6485328058028355,
714
+ "grad_norm": 1.109375,
715
+ "learning_rate": 2e-05,
716
+ "loss": 0.5564,
717
+ "step": 5000
718
+ },
719
+ {
720
+ "epoch": 1.6650181338608638,
721
+ "grad_norm": 1.2734375,
722
+ "learning_rate": 1.989162992420255e-05,
723
+ "loss": 0.5606,
724
+ "step": 5050
725
+ },
726
+ {
727
+ "epoch": 1.6815034619188922,
728
+ "grad_norm": 1.03125,
729
+ "learning_rate": 1.9568868511475862e-05,
730
+ "loss": 0.5399,
731
+ "step": 5100
732
+ },
733
+ {
734
+ "epoch": 1.6979887899769204,
735
+ "grad_norm": 1.1328125,
736
+ "learning_rate": 1.9038711297572278e-05,
737
+ "loss": 0.5604,
738
+ "step": 5150
739
+ },
740
+ {
741
+ "epoch": 1.714474118034949,
742
+ "grad_norm": 1.15625,
743
+ "learning_rate": 1.8312648917982856e-05,
744
+ "loss": 0.554,
745
+ "step": 5200
746
+ },
747
+ {
748
+ "epoch": 1.7309594460929771,
749
+ "grad_norm": 1.046875,
750
+ "learning_rate": 1.7406418059729547e-05,
751
+ "loss": 0.5244,
752
+ "step": 5250
753
+ },
754
+ {
755
+ "epoch": 1.7474447741510057,
756
+ "grad_norm": 1.0390625,
757
+ "learning_rate": 1.633966038417214e-05,
758
+ "loss": 0.5577,
759
+ "step": 5300
760
+ },
761
+ {
762
+ "epoch": 1.763930102209034,
763
+ "grad_norm": 0.953125,
764
+ "learning_rate": 1.513549681334217e-05,
765
+ "loss": 0.5539,
766
+ "step": 5350
767
+ },
768
+ {
769
+ "epoch": 1.7804154302670623,
770
+ "grad_norm": 1.0625,
771
+ "learning_rate": 1.3820026406728302e-05,
772
+ "loss": 0.5375,
773
+ "step": 5400
774
+ },
775
+ {
776
+ "epoch": 1.7969007583250907,
777
+ "grad_norm": 1.0,
778
+ "learning_rate": 1.242176068986535e-05,
779
+ "loss": 0.564,
780
+ "step": 5450
781
+ },
782
+ {
783
+ "epoch": 1.813386086383119,
784
+ "grad_norm": 1.125,
785
+ "learning_rate": 1.09710056950976e-05,
786
+ "loss": 0.5686,
787
+ "step": 5500
788
+ },
789
+ {
790
+ "epoch": 1.8298714144411474,
791
+ "grad_norm": 1.15625,
792
+ "learning_rate": 9.499205108174355e-06,
793
+ "loss": 0.5564,
794
+ "step": 5550
795
+ },
796
+ {
797
+ "epoch": 1.8463567424991756,
798
+ "grad_norm": 0.9296875,
799
+ "learning_rate": 8.038258757328332e-06,
800
+ "loss": 0.552,
801
+ "step": 5600
802
+ },
803
+ {
804
+ "epoch": 1.8628420705572042,
805
+ "grad_norm": 1.0078125,
806
+ "learning_rate": 6.619831215914974e-06,
807
+ "loss": 0.57,
808
+ "step": 5650
809
+ },
810
+ {
811
+ "epoch": 1.8793273986152323,
812
+ "grad_norm": 1.2109375,
813
+ "learning_rate": 5.274665503969513e-06,
814
+ "loss": 0.5607,
815
+ "step": 5700
816
+ },
817
+ {
818
+ "epoch": 1.895812726673261,
819
+ "grad_norm": 1.1015625,
820
+ "learning_rate": 4.031916763524674e-06,
821
+ "loss": 0.5578,
822
+ "step": 5750
823
+ },
824
+ {
825
+ "epoch": 1.912298054731289,
826
+ "grad_norm": 1.0390625,
827
+ "learning_rate": 2.9185203496203065e-06,
828
+ "loss": 0.5373,
829
+ "step": 5800
830
+ },
831
+ {
832
+ "epoch": 1.9287833827893175,
833
+ "grad_norm": 0.9921875,
834
+ "learning_rate": 1.95860803300989e-06,
835
+ "loss": 0.5283,
836
+ "step": 5850
837
+ },
838
+ {
839
+ "epoch": 1.9452687108473459,
840
+ "grad_norm": 1.015625,
841
+ "learning_rate": 1.1729849677954175e-06,
842
+ "loss": 0.538,
843
+ "step": 5900
844
+ },
845
+ {
846
+ "epoch": 1.9617540389053743,
847
+ "grad_norm": 1.0546875,
848
+ "learning_rate": 5.786787602020005e-07,
849
+ "loss": 0.5371,
850
+ "step": 5950
851
+ },
852
+ {
853
+ "epoch": 1.9782393669634026,
854
+ "grad_norm": 0.98046875,
855
+ "learning_rate": 1.8857041198239502e-07,
856
+ "loss": 0.5356,
857
+ "step": 6000
858
+ },
859
+ {
860
+ "epoch": 1.9947246950214308,
861
+ "grad_norm": 1.1171875,
862
+ "learning_rate": 1.111513738975445e-08,
863
+ "loss": 0.5462,
864
+ "step": 6050
865
+ },
866
+ {
867
+ "epoch": 2.0,
868
+ "eval_loss": 0.5490379333496094,
869
+ "eval_runtime": 76.8706,
870
+ "eval_samples_per_second": 21.191,
871
+ "eval_steps_per_second": 1.327,
872
+ "step": 6066
873
+ }
874
+ ],
875
+ "logging_steps": 50,
876
+ "max_steps": 6066,
877
+ "num_input_tokens_seen": 0,
878
+ "num_train_epochs": 2,
879
+ "save_steps": 500,
880
+ "stateful_callbacks": {
881
+ "TrainerControl": {
882
+ "args": {
883
+ "should_epoch_stop": false,
884
+ "should_evaluate": false,
885
+ "should_log": false,
886
+ "should_save": true,
887
+ "should_training_stop": true
888
+ },
889
+ "attributes": {}
890
+ }
891
+ },
892
+ "total_flos": 7.819767672408637e+17,
893
+ "train_batch_size": 16,
894
+ "trial_name": null,
895
+ "trial_params": null
896
+ }
checkpoint-6066/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e0675a81995d37478683f01d22ce39f1a71fbb9775020cfc1de7932d0dabdb9
3
+ size 5432
config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "EleutherAI/pythia-1.4b",
3
+ "architectures": [
4
+ "GPTNeoXForCausalLM"
5
+ ],
6
+ "attention_bias": true,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 0,
9
+ "classifier_dropout": 0.1,
10
+ "eos_token_id": 0,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout": 0.0,
13
+ "hidden_size": 2048,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 8192,
16
+ "layer_norm_eps": 1e-05,
17
+ "max_position_embeddings": 2048,
18
+ "model_type": "gpt_neox",
19
+ "num_attention_heads": 16,
20
+ "num_hidden_layers": 24,
21
+ "partial_rotary_factor": 0.25,
22
+ "rope_scaling": null,
23
+ "rope_theta": 10000,
24
+ "rotary_emb_base": 10000,
25
+ "rotary_pct": 0.25,
26
+ "tie_word_embeddings": false,
27
+ "torch_dtype": "bfloat16",
28
+ "transformers_version": "4.45.0",
29
+ "use_cache": false,
30
+ "use_parallel_residual": true,
31
+ "vocab_size": 50304
32
+ }
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "eos_token_id": 0,
5
+ "transformers_version": "4.45.0",
6
+ "use_cache": false
7
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac8a5ad2319c9ddc5c45584371ff1c8f83b721b9a9794beb97376e606a185fe2
3
+ size 2829330208
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:056766b124a4ac70afcc722084d7c2b6cab80581e0d066709d72ae5c1a8640d8
3
+ size 2930076797
special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "unk_token": "<|endoftext|>"
5
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "bos_token": "<|endoftext|>",
4
+ "eos_token": "<|endoftext|>",
5
+ "name_or_path": "EleutherAI/gpt-neox-20b",
6
+ "special_tokens_map_file": "/admin/home-hailey/.cache/huggingface/hub/models--EleutherAI--gpt-neox-20b/snapshots/4e49eadb5d14bd22f314ec3f45b69a87b88c7691/special_tokens_map.json",
7
+ "tokenizer_class": "GPTNeoXTokenizer",
8
+ "unk_token": "<|endoftext|>"
9
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e0675a81995d37478683f01d22ce39f1a71fbb9775020cfc1de7932d0dabdb9
3
+ size 5432