xzuyn commited on
Commit
bb043ad
1 Parent(s): 8477e66

Upload folder using huggingface_hub

Browse files
checkpoint-128/README.md ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
checkpoint-128/adapter_config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
3
+ "bias": "none",
4
+ "fan_in_fan_out": false,
5
+ "inference_mode": true,
6
+ "init_lora_weights": true,
7
+ "layers_pattern": null,
8
+ "layers_to_transform": null,
9
+ "lora_alpha": 32,
10
+ "lora_dropout": 0.05,
11
+ "modules_to_save": null,
12
+ "peft_type": "LORA",
13
+ "r": 32,
14
+ "revision": null,
15
+ "target_modules": [
16
+ "gate_proj",
17
+ "down_proj",
18
+ "up_proj",
19
+ "q_proj",
20
+ "k_proj",
21
+ "v_proj",
22
+ "o_proj"
23
+ ],
24
+ "task_type": "CAUSAL_LM"
25
+ }
checkpoint-128/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80a63a315ff77e0396ac807010978634f4080a579e4d24a982374f5fa9cd8202
3
+ size 319977674
checkpoint-128/adapter_model/README.md ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
checkpoint-128/adapter_model/adapter_config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
3
+ "bias": "none",
4
+ "fan_in_fan_out": false,
5
+ "inference_mode": true,
6
+ "init_lora_weights": true,
7
+ "layers_pattern": null,
8
+ "layers_to_transform": null,
9
+ "lora_alpha": 32,
10
+ "lora_dropout": 0.05,
11
+ "modules_to_save": null,
12
+ "peft_type": "LORA",
13
+ "r": 32,
14
+ "revision": null,
15
+ "target_modules": [
16
+ "gate_proj",
17
+ "down_proj",
18
+ "up_proj",
19
+ "q_proj",
20
+ "k_proj",
21
+ "v_proj",
22
+ "o_proj"
23
+ ],
24
+ "task_type": "CAUSAL_LM"
25
+ }
checkpoint-128/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80a63a315ff77e0396ac807010978634f4080a579e4d24a982374f5fa9cd8202
3
+ size 319977674
checkpoint-128/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff8b608bf64e8f628bb90f81b1fea2c81ab3ff2364deaf9ec90a1f3b272139d3
3
+ size 160746900
checkpoint-128/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b546edf0c4db53a664c3625f2de4b10f336ee74a3c9bc65fdfbc67f3823e2118
3
+ size 14244
checkpoint-128/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be4181246c026df0ebb9f3b5762d7c7ba2acd3ca14083db52d93f5b2a829c863
3
+ size 1064
checkpoint-128/trainer_state.json ADDED
@@ -0,0 +1,784 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.9825242718446603,
5
+ "global_step": 128,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.02,
12
+ "learning_rate": 2.631578947368421e-06,
13
+ "loss": 1.6384,
14
+ "step": 1
15
+ },
16
+ {
17
+ "epoch": 0.05,
18
+ "learning_rate": 5.263157894736842e-06,
19
+ "loss": 1.677,
20
+ "step": 2
21
+ },
22
+ {
23
+ "epoch": 0.07,
24
+ "learning_rate": 7.894736842105263e-06,
25
+ "loss": 1.9569,
26
+ "step": 3
27
+ },
28
+ {
29
+ "epoch": 0.09,
30
+ "learning_rate": 1.0526315789473684e-05,
31
+ "loss": 1.8787,
32
+ "step": 4
33
+ },
34
+ {
35
+ "epoch": 0.12,
36
+ "learning_rate": 1.3157894736842106e-05,
37
+ "loss": 1.9318,
38
+ "step": 5
39
+ },
40
+ {
41
+ "epoch": 0.14,
42
+ "learning_rate": 1.5789473684210526e-05,
43
+ "loss": 2.1207,
44
+ "step": 6
45
+ },
46
+ {
47
+ "epoch": 0.16,
48
+ "learning_rate": 1.8421052631578947e-05,
49
+ "loss": 2.0227,
50
+ "step": 7
51
+ },
52
+ {
53
+ "epoch": 0.19,
54
+ "learning_rate": 2.105263157894737e-05,
55
+ "loss": 2.103,
56
+ "step": 8
57
+ },
58
+ {
59
+ "epoch": 0.21,
60
+ "learning_rate": 2.368421052631579e-05,
61
+ "loss": 2.2184,
62
+ "step": 9
63
+ },
64
+ {
65
+ "epoch": 0.23,
66
+ "learning_rate": 2.6315789473684212e-05,
67
+ "loss": 2.455,
68
+ "step": 10
69
+ },
70
+ {
71
+ "epoch": 0.26,
72
+ "learning_rate": 2.8947368421052634e-05,
73
+ "loss": 1.6635,
74
+ "step": 11
75
+ },
76
+ {
77
+ "epoch": 0.28,
78
+ "learning_rate": 3.157894736842105e-05,
79
+ "loss": 1.6389,
80
+ "step": 12
81
+ },
82
+ {
83
+ "epoch": 0.3,
84
+ "learning_rate": 3.421052631578947e-05,
85
+ "loss": 1.7803,
86
+ "step": 13
87
+ },
88
+ {
89
+ "epoch": 0.33,
90
+ "learning_rate": 3.6842105263157895e-05,
91
+ "loss": 1.8378,
92
+ "step": 14
93
+ },
94
+ {
95
+ "epoch": 0.35,
96
+ "learning_rate": 3.9473684210526316e-05,
97
+ "loss": 1.9676,
98
+ "step": 15
99
+ },
100
+ {
101
+ "epoch": 0.37,
102
+ "learning_rate": 4.210526315789474e-05,
103
+ "loss": 2.0158,
104
+ "step": 16
105
+ },
106
+ {
107
+ "epoch": 0.4,
108
+ "learning_rate": 4.473684210526316e-05,
109
+ "loss": 1.8251,
110
+ "step": 17
111
+ },
112
+ {
113
+ "epoch": 0.42,
114
+ "learning_rate": 4.736842105263158e-05,
115
+ "loss": 1.9671,
116
+ "step": 18
117
+ },
118
+ {
119
+ "epoch": 0.44,
120
+ "learning_rate": 5e-05,
121
+ "loss": 1.9491,
122
+ "step": 19
123
+ },
124
+ {
125
+ "epoch": 0.47,
126
+ "learning_rate": 5.2631578947368424e-05,
127
+ "loss": 1.9589,
128
+ "step": 20
129
+ },
130
+ {
131
+ "epoch": 0.49,
132
+ "learning_rate": 5.526315789473685e-05,
133
+ "loss": 1.6453,
134
+ "step": 21
135
+ },
136
+ {
137
+ "epoch": 0.51,
138
+ "learning_rate": 5.789473684210527e-05,
139
+ "loss": 1.5859,
140
+ "step": 22
141
+ },
142
+ {
143
+ "epoch": 0.54,
144
+ "learning_rate": 6.052631578947369e-05,
145
+ "loss": 1.7574,
146
+ "step": 23
147
+ },
148
+ {
149
+ "epoch": 0.56,
150
+ "learning_rate": 6.31578947368421e-05,
151
+ "loss": 1.8601,
152
+ "step": 24
153
+ },
154
+ {
155
+ "epoch": 0.58,
156
+ "learning_rate": 6.578947368421054e-05,
157
+ "loss": 1.7867,
158
+ "step": 25
159
+ },
160
+ {
161
+ "epoch": 0.61,
162
+ "learning_rate": 6.842105263157895e-05,
163
+ "loss": 1.9736,
164
+ "step": 26
165
+ },
166
+ {
167
+ "epoch": 0.63,
168
+ "learning_rate": 7.105263157894737e-05,
169
+ "loss": 1.96,
170
+ "step": 27
171
+ },
172
+ {
173
+ "epoch": 0.65,
174
+ "learning_rate": 7.368421052631579e-05,
175
+ "loss": 1.7974,
176
+ "step": 28
177
+ },
178
+ {
179
+ "epoch": 0.68,
180
+ "learning_rate": 7.631578947368422e-05,
181
+ "loss": 1.8705,
182
+ "step": 29
183
+ },
184
+ {
185
+ "epoch": 0.7,
186
+ "learning_rate": 7.894736842105263e-05,
187
+ "loss": 1.8634,
188
+ "step": 30
189
+ },
190
+ {
191
+ "epoch": 0.72,
192
+ "learning_rate": 8.157894736842105e-05,
193
+ "loss": 1.6359,
194
+ "step": 31
195
+ },
196
+ {
197
+ "epoch": 0.75,
198
+ "learning_rate": 8.421052631578948e-05,
199
+ "loss": 1.6283,
200
+ "step": 32
201
+ },
202
+ {
203
+ "epoch": 0.77,
204
+ "learning_rate": 8.68421052631579e-05,
205
+ "loss": 1.8596,
206
+ "step": 33
207
+ },
208
+ {
209
+ "epoch": 0.79,
210
+ "learning_rate": 8.947368421052632e-05,
211
+ "loss": 1.7455,
212
+ "step": 34
213
+ },
214
+ {
215
+ "epoch": 0.82,
216
+ "learning_rate": 9.210526315789474e-05,
217
+ "loss": 1.7927,
218
+ "step": 35
219
+ },
220
+ {
221
+ "epoch": 0.84,
222
+ "learning_rate": 9.473684210526316e-05,
223
+ "loss": 1.8419,
224
+ "step": 36
225
+ },
226
+ {
227
+ "epoch": 0.86,
228
+ "learning_rate": 9.736842105263158e-05,
229
+ "loss": 1.8011,
230
+ "step": 37
231
+ },
232
+ {
233
+ "epoch": 0.89,
234
+ "learning_rate": 0.0001,
235
+ "loss": 1.8074,
236
+ "step": 38
237
+ },
238
+ {
239
+ "epoch": 0.91,
240
+ "learning_rate": 9.983108108108109e-05,
241
+ "loss": 2.0587,
242
+ "step": 39
243
+ },
244
+ {
245
+ "epoch": 0.93,
246
+ "learning_rate": 9.966216216216217e-05,
247
+ "loss": 1.9083,
248
+ "step": 40
249
+ },
250
+ {
251
+ "epoch": 0.96,
252
+ "learning_rate": 9.949324324324325e-05,
253
+ "loss": 1.7381,
254
+ "step": 41
255
+ },
256
+ {
257
+ "epoch": 0.98,
258
+ "learning_rate": 9.932432432432433e-05,
259
+ "loss": 1.9415,
260
+ "step": 42
261
+ },
262
+ {
263
+ "epoch": 1.0,
264
+ "learning_rate": 9.915540540540541e-05,
265
+ "loss": 1.6324,
266
+ "step": 43
267
+ },
268
+ {
269
+ "epoch": 1.03,
270
+ "learning_rate": 9.89864864864865e-05,
271
+ "loss": 1.5839,
272
+ "step": 44
273
+ },
274
+ {
275
+ "epoch": 1.05,
276
+ "learning_rate": 9.881756756756757e-05,
277
+ "loss": 1.8506,
278
+ "step": 45
279
+ },
280
+ {
281
+ "epoch": 1.07,
282
+ "learning_rate": 9.864864864864865e-05,
283
+ "loss": 1.8398,
284
+ "step": 46
285
+ },
286
+ {
287
+ "epoch": 1.1,
288
+ "learning_rate": 9.847972972972973e-05,
289
+ "loss": 1.8458,
290
+ "step": 47
291
+ },
292
+ {
293
+ "epoch": 1.12,
294
+ "learning_rate": 9.831081081081081e-05,
295
+ "loss": 1.8309,
296
+ "step": 48
297
+ },
298
+ {
299
+ "epoch": 1.14,
300
+ "learning_rate": 9.81418918918919e-05,
301
+ "loss": 1.8339,
302
+ "step": 49
303
+ },
304
+ {
305
+ "epoch": 1.17,
306
+ "learning_rate": 9.797297297297297e-05,
307
+ "loss": 1.815,
308
+ "step": 50
309
+ },
310
+ {
311
+ "epoch": 1.19,
312
+ "learning_rate": 9.780405405405407e-05,
313
+ "loss": 1.7453,
314
+ "step": 51
315
+ },
316
+ {
317
+ "epoch": 1.21,
318
+ "learning_rate": 9.763513513513513e-05,
319
+ "loss": 1.7974,
320
+ "step": 52
321
+ },
322
+ {
323
+ "epoch": 1.23,
324
+ "learning_rate": 9.746621621621623e-05,
325
+ "loss": 1.7996,
326
+ "step": 53
327
+ },
328
+ {
329
+ "epoch": 1.26,
330
+ "learning_rate": 9.729729729729731e-05,
331
+ "loss": 1.5471,
332
+ "step": 54
333
+ },
334
+ {
335
+ "epoch": 1.28,
336
+ "learning_rate": 9.712837837837838e-05,
337
+ "loss": 1.6582,
338
+ "step": 55
339
+ },
340
+ {
341
+ "epoch": 1.3,
342
+ "learning_rate": 9.695945945945947e-05,
343
+ "loss": 1.7868,
344
+ "step": 56
345
+ },
346
+ {
347
+ "epoch": 1.33,
348
+ "learning_rate": 9.679054054054054e-05,
349
+ "loss": 1.7449,
350
+ "step": 57
351
+ },
352
+ {
353
+ "epoch": 1.35,
354
+ "learning_rate": 9.662162162162163e-05,
355
+ "loss": 1.9309,
356
+ "step": 58
357
+ },
358
+ {
359
+ "epoch": 1.37,
360
+ "learning_rate": 9.64527027027027e-05,
361
+ "loss": 1.6582,
362
+ "step": 59
363
+ },
364
+ {
365
+ "epoch": 1.4,
366
+ "learning_rate": 9.628378378378379e-05,
367
+ "loss": 1.749,
368
+ "step": 60
369
+ },
370
+ {
371
+ "epoch": 1.42,
372
+ "learning_rate": 9.611486486486487e-05,
373
+ "loss": 1.7815,
374
+ "step": 61
375
+ },
376
+ {
377
+ "epoch": 1.44,
378
+ "learning_rate": 9.594594594594595e-05,
379
+ "loss": 1.7758,
380
+ "step": 62
381
+ },
382
+ {
383
+ "epoch": 1.47,
384
+ "learning_rate": 9.577702702702703e-05,
385
+ "loss": 1.6416,
386
+ "step": 63
387
+ },
388
+ {
389
+ "epoch": 1.49,
390
+ "learning_rate": 9.56081081081081e-05,
391
+ "loss": 1.6242,
392
+ "step": 64
393
+ },
394
+ {
395
+ "epoch": 1.51,
396
+ "learning_rate": 9.54391891891892e-05,
397
+ "loss": 1.5596,
398
+ "step": 65
399
+ },
400
+ {
401
+ "epoch": 1.54,
402
+ "learning_rate": 9.527027027027028e-05,
403
+ "loss": 1.7654,
404
+ "step": 66
405
+ },
406
+ {
407
+ "epoch": 1.56,
408
+ "learning_rate": 9.510135135135136e-05,
409
+ "loss": 1.7598,
410
+ "step": 67
411
+ },
412
+ {
413
+ "epoch": 1.58,
414
+ "learning_rate": 9.493243243243244e-05,
415
+ "loss": 1.7758,
416
+ "step": 68
417
+ },
418
+ {
419
+ "epoch": 1.61,
420
+ "learning_rate": 9.476351351351352e-05,
421
+ "loss": 1.9316,
422
+ "step": 69
423
+ },
424
+ {
425
+ "epoch": 1.63,
426
+ "learning_rate": 9.45945945945946e-05,
427
+ "loss": 1.766,
428
+ "step": 70
429
+ },
430
+ {
431
+ "epoch": 1.65,
432
+ "learning_rate": 9.442567567567568e-05,
433
+ "loss": 1.8492,
434
+ "step": 71
435
+ },
436
+ {
437
+ "epoch": 1.68,
438
+ "learning_rate": 9.425675675675676e-05,
439
+ "loss": 1.8198,
440
+ "step": 72
441
+ },
442
+ {
443
+ "epoch": 1.7,
444
+ "learning_rate": 9.408783783783784e-05,
445
+ "loss": 1.445,
446
+ "step": 73
447
+ },
448
+ {
449
+ "epoch": 1.72,
450
+ "learning_rate": 9.391891891891892e-05,
451
+ "loss": 1.5804,
452
+ "step": 74
453
+ },
454
+ {
455
+ "epoch": 1.75,
456
+ "learning_rate": 9.375e-05,
457
+ "loss": 1.5489,
458
+ "step": 75
459
+ },
460
+ {
461
+ "epoch": 1.77,
462
+ "learning_rate": 9.35810810810811e-05,
463
+ "loss": 1.7003,
464
+ "step": 76
465
+ },
466
+ {
467
+ "epoch": 1.79,
468
+ "learning_rate": 9.341216216216216e-05,
469
+ "loss": 1.6788,
470
+ "step": 77
471
+ },
472
+ {
473
+ "epoch": 1.82,
474
+ "learning_rate": 9.324324324324324e-05,
475
+ "loss": 1.8201,
476
+ "step": 78
477
+ },
478
+ {
479
+ "epoch": 1.84,
480
+ "learning_rate": 9.307432432432432e-05,
481
+ "loss": 1.891,
482
+ "step": 79
483
+ },
484
+ {
485
+ "epoch": 1.86,
486
+ "learning_rate": 9.29054054054054e-05,
487
+ "loss": 1.7963,
488
+ "step": 80
489
+ },
490
+ {
491
+ "epoch": 1.89,
492
+ "learning_rate": 9.27364864864865e-05,
493
+ "loss": 1.8285,
494
+ "step": 81
495
+ },
496
+ {
497
+ "epoch": 1.91,
498
+ "learning_rate": 9.256756756756757e-05,
499
+ "loss": 1.9649,
500
+ "step": 82
501
+ },
502
+ {
503
+ "epoch": 1.93,
504
+ "learning_rate": 9.239864864864866e-05,
505
+ "loss": 1.471,
506
+ "step": 83
507
+ },
508
+ {
509
+ "epoch": 1.96,
510
+ "learning_rate": 9.222972972972973e-05,
511
+ "loss": 1.6209,
512
+ "step": 84
513
+ },
514
+ {
515
+ "epoch": 1.98,
516
+ "learning_rate": 9.206081081081082e-05,
517
+ "loss": 1.8244,
518
+ "step": 85
519
+ },
520
+ {
521
+ "epoch": 2.0,
522
+ "learning_rate": 9.18918918918919e-05,
523
+ "loss": 1.8515,
524
+ "step": 86
525
+ },
526
+ {
527
+ "epoch": 2.03,
528
+ "learning_rate": 9.172297297297297e-05,
529
+ "loss": 1.6504,
530
+ "step": 87
531
+ },
532
+ {
533
+ "epoch": 2.05,
534
+ "learning_rate": 9.155405405405406e-05,
535
+ "loss": 1.5222,
536
+ "step": 88
537
+ },
538
+ {
539
+ "epoch": 2.07,
540
+ "learning_rate": 9.138513513513513e-05,
541
+ "loss": 1.7524,
542
+ "step": 89
543
+ },
544
+ {
545
+ "epoch": 2.1,
546
+ "learning_rate": 9.121621621621623e-05,
547
+ "loss": 1.8002,
548
+ "step": 90
549
+ },
550
+ {
551
+ "epoch": 2.12,
552
+ "learning_rate": 9.10472972972973e-05,
553
+ "loss": 1.8035,
554
+ "step": 91
555
+ },
556
+ {
557
+ "epoch": 2.14,
558
+ "learning_rate": 9.087837837837839e-05,
559
+ "loss": 1.8115,
560
+ "step": 92
561
+ },
562
+ {
563
+ "epoch": 2.17,
564
+ "learning_rate": 9.070945945945947e-05,
565
+ "loss": 1.7192,
566
+ "step": 93
567
+ },
568
+ {
569
+ "epoch": 2.19,
570
+ "learning_rate": 9.054054054054055e-05,
571
+ "loss": 1.7066,
572
+ "step": 94
573
+ },
574
+ {
575
+ "epoch": 2.21,
576
+ "learning_rate": 9.037162162162163e-05,
577
+ "loss": 1.6312,
578
+ "step": 95
579
+ },
580
+ {
581
+ "epoch": 2.24,
582
+ "learning_rate": 9.02027027027027e-05,
583
+ "loss": 1.2849,
584
+ "step": 96
585
+ },
586
+ {
587
+ "epoch": 2.26,
588
+ "learning_rate": 9.003378378378379e-05,
589
+ "loss": 1.5149,
590
+ "step": 97
591
+ },
592
+ {
593
+ "epoch": 2.28,
594
+ "learning_rate": 8.986486486486487e-05,
595
+ "loss": 1.4894,
596
+ "step": 98
597
+ },
598
+ {
599
+ "epoch": 2.31,
600
+ "learning_rate": 8.969594594594595e-05,
601
+ "loss": 1.7987,
602
+ "step": 99
603
+ },
604
+ {
605
+ "epoch": 2.33,
606
+ "learning_rate": 8.952702702702703e-05,
607
+ "loss": 1.6454,
608
+ "step": 100
609
+ },
610
+ {
611
+ "epoch": 2.35,
612
+ "learning_rate": 8.935810810810811e-05,
613
+ "loss": 1.8256,
614
+ "step": 101
615
+ },
616
+ {
617
+ "epoch": 2.38,
618
+ "learning_rate": 8.918918918918919e-05,
619
+ "loss": 1.7403,
620
+ "step": 102
621
+ },
622
+ {
623
+ "epoch": 2.4,
624
+ "learning_rate": 8.902027027027027e-05,
625
+ "loss": 1.6801,
626
+ "step": 103
627
+ },
628
+ {
629
+ "epoch": 2.42,
630
+ "learning_rate": 8.885135135135135e-05,
631
+ "loss": 1.7089,
632
+ "step": 104
633
+ },
634
+ {
635
+ "epoch": 2.45,
636
+ "learning_rate": 8.868243243243243e-05,
637
+ "loss": 1.7009,
638
+ "step": 105
639
+ },
640
+ {
641
+ "epoch": 2.47,
642
+ "learning_rate": 8.851351351351352e-05,
643
+ "loss": 1.3085,
644
+ "step": 106
645
+ },
646
+ {
647
+ "epoch": 2.49,
648
+ "learning_rate": 8.83445945945946e-05,
649
+ "loss": 1.5615,
650
+ "step": 107
651
+ },
652
+ {
653
+ "epoch": 2.52,
654
+ "learning_rate": 8.817567567567569e-05,
655
+ "loss": 1.6415,
656
+ "step": 108
657
+ },
658
+ {
659
+ "epoch": 2.54,
660
+ "learning_rate": 8.800675675675676e-05,
661
+ "loss": 1.5999,
662
+ "step": 109
663
+ },
664
+ {
665
+ "epoch": 2.56,
666
+ "learning_rate": 8.783783783783784e-05,
667
+ "loss": 1.6671,
668
+ "step": 110
669
+ },
670
+ {
671
+ "epoch": 2.59,
672
+ "learning_rate": 8.766891891891892e-05,
673
+ "loss": 1.8566,
674
+ "step": 111
675
+ },
676
+ {
677
+ "epoch": 2.61,
678
+ "learning_rate": 8.75e-05,
679
+ "loss": 1.9677,
680
+ "step": 112
681
+ },
682
+ {
683
+ "epoch": 2.63,
684
+ "learning_rate": 8.73310810810811e-05,
685
+ "loss": 1.6473,
686
+ "step": 113
687
+ },
688
+ {
689
+ "epoch": 2.66,
690
+ "learning_rate": 8.716216216216216e-05,
691
+ "loss": 1.6044,
692
+ "step": 114
693
+ },
694
+ {
695
+ "epoch": 2.68,
696
+ "learning_rate": 8.699324324324325e-05,
697
+ "loss": 1.7437,
698
+ "step": 115
699
+ },
700
+ {
701
+ "epoch": 2.7,
702
+ "learning_rate": 8.682432432432432e-05,
703
+ "loss": 1.2056,
704
+ "step": 116
705
+ },
706
+ {
707
+ "epoch": 2.73,
708
+ "learning_rate": 8.665540540540542e-05,
709
+ "loss": 1.5716,
710
+ "step": 117
711
+ },
712
+ {
713
+ "epoch": 2.75,
714
+ "learning_rate": 8.64864864864865e-05,
715
+ "loss": 1.6996,
716
+ "step": 118
717
+ },
718
+ {
719
+ "epoch": 2.77,
720
+ "learning_rate": 8.631756756756756e-05,
721
+ "loss": 1.7896,
722
+ "step": 119
723
+ },
724
+ {
725
+ "epoch": 2.8,
726
+ "learning_rate": 8.614864864864866e-05,
727
+ "loss": 1.7133,
728
+ "step": 120
729
+ },
730
+ {
731
+ "epoch": 2.82,
732
+ "learning_rate": 8.597972972972972e-05,
733
+ "loss": 1.6978,
734
+ "step": 121
735
+ },
736
+ {
737
+ "epoch": 2.84,
738
+ "learning_rate": 8.581081081081082e-05,
739
+ "loss": 1.7591,
740
+ "step": 122
741
+ },
742
+ {
743
+ "epoch": 2.87,
744
+ "learning_rate": 8.56418918918919e-05,
745
+ "loss": 1.6284,
746
+ "step": 123
747
+ },
748
+ {
749
+ "epoch": 2.89,
750
+ "learning_rate": 8.547297297297298e-05,
751
+ "loss": 1.8141,
752
+ "step": 124
753
+ },
754
+ {
755
+ "epoch": 2.91,
756
+ "learning_rate": 8.530405405405406e-05,
757
+ "loss": 1.6254,
758
+ "step": 125
759
+ },
760
+ {
761
+ "epoch": 2.94,
762
+ "learning_rate": 8.513513513513514e-05,
763
+ "loss": 1.5776,
764
+ "step": 126
765
+ },
766
+ {
767
+ "epoch": 2.96,
768
+ "learning_rate": 8.496621621621622e-05,
769
+ "loss": 1.7454,
770
+ "step": 127
771
+ },
772
+ {
773
+ "epoch": 2.98,
774
+ "learning_rate": 8.47972972972973e-05,
775
+ "loss": 1.8205,
776
+ "step": 128
777
+ }
778
+ ],
779
+ "max_steps": 630,
780
+ "num_train_epochs": 15,
781
+ "total_flos": 4.660769981693952e+16,
782
+ "trial_name": null,
783
+ "trial_params": null
784
+ }
checkpoint-128/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c25aabb1993b3aedb881713056a431f01aa3b24e332809302d8cfa95efb4d04
3
+ size 4408