Davlan commited on
Commit
ea50a15
1 Parent(s): b8b2875

Upload trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +1717 -0
trainer_state.json ADDED
@@ -0,0 +1,1717 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "global_step": 141057,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 4.982276668297213e-05,
13
+ "loss": 2.4577,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.02,
18
+ "learning_rate": 4.964553336594426e-05,
19
+ "loss": 2.4076,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.03,
24
+ "learning_rate": 4.94683000489164e-05,
25
+ "loss": 2.3823,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.04,
30
+ "learning_rate": 4.929106673188853e-05,
31
+ "loss": 2.3764,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.05,
36
+ "learning_rate": 4.911383341486066e-05,
37
+ "loss": 2.3482,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.06,
42
+ "learning_rate": 4.893660009783279e-05,
43
+ "loss": 2.3238,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.07,
48
+ "learning_rate": 4.8759366780804925e-05,
49
+ "loss": 2.3344,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.09,
54
+ "learning_rate": 4.8582133463777055e-05,
55
+ "loss": 2.3137,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.1,
60
+ "learning_rate": 4.840490014674919e-05,
61
+ "loss": 2.3036,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.11,
66
+ "learning_rate": 4.822766682972132e-05,
67
+ "loss": 2.2799,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.12,
72
+ "learning_rate": 4.805043351269346e-05,
73
+ "loss": 2.283,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.13,
78
+ "learning_rate": 4.787320019566559e-05,
79
+ "loss": 2.271,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.14,
84
+ "learning_rate": 4.769596687863772e-05,
85
+ "loss": 2.267,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.15,
90
+ "learning_rate": 4.751873356160985e-05,
91
+ "loss": 2.2637,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.16,
96
+ "learning_rate": 4.7341500244581984e-05,
97
+ "loss": 2.2565,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.17,
102
+ "learning_rate": 4.7164266927554114e-05,
103
+ "loss": 2.2501,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.18,
108
+ "learning_rate": 4.6987033610526244e-05,
109
+ "loss": 2.2233,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.19,
114
+ "learning_rate": 4.6809800293498373e-05,
115
+ "loss": 2.2321,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.2,
120
+ "learning_rate": 4.663256697647051e-05,
121
+ "loss": 2.221,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.21,
126
+ "learning_rate": 4.645533365944264e-05,
127
+ "loss": 2.2343,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.22,
132
+ "learning_rate": 4.627810034241477e-05,
133
+ "loss": 2.2038,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.23,
138
+ "learning_rate": 4.61008670253869e-05,
139
+ "loss": 2.2013,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.24,
144
+ "learning_rate": 4.592363370835903e-05,
145
+ "loss": 2.2105,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.26,
150
+ "learning_rate": 4.5746400391331166e-05,
151
+ "loss": 2.1881,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.27,
156
+ "learning_rate": 4.5569167074303296e-05,
157
+ "loss": 2.1832,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.28,
162
+ "learning_rate": 4.5391933757275426e-05,
163
+ "loss": 2.1774,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.29,
168
+ "learning_rate": 4.521470044024756e-05,
169
+ "loss": 2.1714,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.3,
174
+ "learning_rate": 4.503746712321969e-05,
175
+ "loss": 2.1602,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.31,
180
+ "learning_rate": 4.486023380619183e-05,
181
+ "loss": 2.1583,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.32,
186
+ "learning_rate": 4.468300048916396e-05,
187
+ "loss": 2.1707,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.33,
192
+ "learning_rate": 4.450576717213609e-05,
193
+ "loss": 2.1689,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.34,
198
+ "learning_rate": 4.4328533855108225e-05,
199
+ "loss": 2.1479,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.35,
204
+ "learning_rate": 4.4151300538080355e-05,
205
+ "loss": 2.1554,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.36,
210
+ "learning_rate": 4.3974067221052485e-05,
211
+ "loss": 2.1309,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.37,
216
+ "learning_rate": 4.3796833904024615e-05,
217
+ "loss": 2.1328,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.38,
222
+ "learning_rate": 4.361960058699675e-05,
223
+ "loss": 2.1469,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.39,
228
+ "learning_rate": 4.344236726996888e-05,
229
+ "loss": 2.1316,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.4,
234
+ "learning_rate": 4.326513395294101e-05,
235
+ "loss": 2.1387,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.41,
240
+ "learning_rate": 4.308790063591314e-05,
241
+ "loss": 2.1143,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.43,
246
+ "learning_rate": 4.291066731888528e-05,
247
+ "loss": 2.1301,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.44,
252
+ "learning_rate": 4.273343400185741e-05,
253
+ "loss": 2.1275,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.45,
258
+ "learning_rate": 4.255620068482954e-05,
259
+ "loss": 2.1172,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.46,
264
+ "learning_rate": 4.237896736780167e-05,
265
+ "loss": 2.118,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.47,
270
+ "learning_rate": 4.2201734050773804e-05,
271
+ "loss": 2.1287,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.48,
276
+ "learning_rate": 4.202450073374593e-05,
277
+ "loss": 2.1148,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.49,
282
+ "learning_rate": 4.184726741671807e-05,
283
+ "loss": 2.1019,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.5,
288
+ "learning_rate": 4.16700340996902e-05,
289
+ "loss": 2.1088,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.51,
294
+ "learning_rate": 4.149280078266233e-05,
295
+ "loss": 2.1092,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.52,
300
+ "learning_rate": 4.1315567465634466e-05,
301
+ "loss": 2.0987,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.53,
306
+ "learning_rate": 4.1138334148606596e-05,
307
+ "loss": 2.0778,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.54,
312
+ "learning_rate": 4.0961100831578726e-05,
313
+ "loss": 2.0963,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.55,
318
+ "learning_rate": 4.0783867514550856e-05,
319
+ "loss": 2.0868,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.56,
324
+ "learning_rate": 4.060663419752299e-05,
325
+ "loss": 2.0881,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.57,
330
+ "learning_rate": 4.042940088049512e-05,
331
+ "loss": 2.0887,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.58,
336
+ "learning_rate": 4.025216756346725e-05,
337
+ "loss": 2.0921,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.6,
342
+ "learning_rate": 4.007493424643938e-05,
343
+ "loss": 2.0782,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.61,
348
+ "learning_rate": 3.989770092941152e-05,
349
+ "loss": 2.0626,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.62,
354
+ "learning_rate": 3.972046761238365e-05,
355
+ "loss": 2.077,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.63,
360
+ "learning_rate": 3.954323429535578e-05,
361
+ "loss": 2.0728,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.64,
366
+ "learning_rate": 3.936600097832791e-05,
367
+ "loss": 2.0528,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.65,
372
+ "learning_rate": 3.918876766130004e-05,
373
+ "loss": 2.0661,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.66,
378
+ "learning_rate": 3.9011534344272175e-05,
379
+ "loss": 2.0639,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 0.67,
384
+ "learning_rate": 3.8834301027244304e-05,
385
+ "loss": 2.063,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 0.68,
390
+ "learning_rate": 3.865706771021644e-05,
391
+ "loss": 2.0445,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 0.69,
396
+ "learning_rate": 3.847983439318857e-05,
397
+ "loss": 2.0514,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 0.7,
402
+ "learning_rate": 3.830260107616071e-05,
403
+ "loss": 2.0538,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 0.71,
408
+ "learning_rate": 3.812536775913284e-05,
409
+ "loss": 2.041,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 0.72,
414
+ "learning_rate": 3.794813444210497e-05,
415
+ "loss": 2.0441,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 0.73,
420
+ "learning_rate": 3.77709011250771e-05,
421
+ "loss": 2.044,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 0.74,
426
+ "learning_rate": 3.7593667808049234e-05,
427
+ "loss": 2.0449,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 0.76,
432
+ "learning_rate": 3.7416434491021363e-05,
433
+ "loss": 2.0381,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 0.77,
438
+ "learning_rate": 3.723920117399349e-05,
439
+ "loss": 2.0302,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 0.78,
444
+ "learning_rate": 3.706196785696562e-05,
445
+ "loss": 2.028,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 0.79,
450
+ "learning_rate": 3.688473453993776e-05,
451
+ "loss": 2.0335,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 0.8,
456
+ "learning_rate": 3.670750122290989e-05,
457
+ "loss": 2.0289,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 0.81,
462
+ "learning_rate": 3.653026790588202e-05,
463
+ "loss": 2.0072,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 0.82,
468
+ "learning_rate": 3.635303458885415e-05,
469
+ "loss": 2.0341,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 0.83,
474
+ "learning_rate": 3.6175801271826286e-05,
475
+ "loss": 2.0233,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 0.84,
480
+ "learning_rate": 3.5998567954798416e-05,
481
+ "loss": 2.0146,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 0.85,
486
+ "learning_rate": 3.5821334637770546e-05,
487
+ "loss": 2.0205,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 0.86,
492
+ "learning_rate": 3.5644101320742675e-05,
493
+ "loss": 2.009,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 0.87,
498
+ "learning_rate": 3.546686800371481e-05,
499
+ "loss": 2.0023,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 0.88,
504
+ "learning_rate": 3.528963468668695e-05,
505
+ "loss": 2.0102,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 0.89,
510
+ "learning_rate": 3.511240136965908e-05,
511
+ "loss": 2.0074,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 0.9,
516
+ "learning_rate": 3.493516805263121e-05,
517
+ "loss": 1.9968,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 0.91,
522
+ "learning_rate": 3.475793473560334e-05,
523
+ "loss": 1.9968,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 0.93,
528
+ "learning_rate": 3.4580701418575475e-05,
529
+ "loss": 2.0042,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 0.94,
534
+ "learning_rate": 3.4403468101547605e-05,
535
+ "loss": 1.9947,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 0.95,
540
+ "learning_rate": 3.4226234784519735e-05,
541
+ "loss": 1.9995,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 0.96,
546
+ "learning_rate": 3.4049001467491864e-05,
547
+ "loss": 1.9929,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 0.97,
552
+ "learning_rate": 3.3871768150464e-05,
553
+ "loss": 1.9935,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 0.98,
558
+ "learning_rate": 3.369453483343613e-05,
559
+ "loss": 1.9931,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 0.99,
564
+ "learning_rate": 3.351730151640826e-05,
565
+ "loss": 1.9944,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 1.0,
570
+ "learning_rate": 3.334006819938039e-05,
571
+ "loss": 1.9779,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 1.01,
576
+ "learning_rate": 3.316283488235253e-05,
577
+ "loss": 1.9819,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 1.02,
582
+ "learning_rate": 3.298560156532466e-05,
583
+ "loss": 1.9798,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 1.03,
588
+ "learning_rate": 3.280836824829679e-05,
589
+ "loss": 1.9767,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 1.04,
594
+ "learning_rate": 3.263113493126892e-05,
595
+ "loss": 1.963,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 1.05,
600
+ "learning_rate": 3.245390161424105e-05,
601
+ "loss": 1.9686,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 1.06,
606
+ "learning_rate": 3.227666829721318e-05,
607
+ "loss": 1.9529,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 1.07,
612
+ "learning_rate": 3.209943498018532e-05,
613
+ "loss": 1.96,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 1.08,
618
+ "learning_rate": 3.192220166315745e-05,
619
+ "loss": 1.9683,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 1.1,
624
+ "learning_rate": 3.174496834612958e-05,
625
+ "loss": 1.9574,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 1.11,
630
+ "learning_rate": 3.1567735029101716e-05,
631
+ "loss": 1.9573,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 1.12,
636
+ "learning_rate": 3.1390501712073846e-05,
637
+ "loss": 1.9582,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 1.13,
642
+ "learning_rate": 3.1213268395045976e-05,
643
+ "loss": 1.9559,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 1.14,
648
+ "learning_rate": 3.1036035078018106e-05,
649
+ "loss": 1.9464,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 1.15,
654
+ "learning_rate": 3.085880176099024e-05,
655
+ "loss": 1.9512,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 1.16,
660
+ "learning_rate": 3.068156844396237e-05,
661
+ "loss": 1.9676,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 1.17,
666
+ "learning_rate": 3.0504335126934502e-05,
667
+ "loss": 1.946,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 1.18,
672
+ "learning_rate": 3.0327101809906632e-05,
673
+ "loss": 1.9586,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 1.19,
678
+ "learning_rate": 3.014986849287877e-05,
679
+ "loss": 1.9553,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 1.2,
684
+ "learning_rate": 2.9972635175850898e-05,
685
+ "loss": 1.9424,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 1.21,
690
+ "learning_rate": 2.979540185882303e-05,
691
+ "loss": 1.9431,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 1.22,
696
+ "learning_rate": 2.961816854179516e-05,
697
+ "loss": 1.9562,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 1.23,
702
+ "learning_rate": 2.9440935224767298e-05,
703
+ "loss": 1.9376,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 1.24,
708
+ "learning_rate": 2.9263701907739428e-05,
709
+ "loss": 1.9313,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 1.25,
714
+ "learning_rate": 2.9086468590711558e-05,
715
+ "loss": 1.9492,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 1.27,
720
+ "learning_rate": 2.8909235273683687e-05,
721
+ "loss": 1.9396,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 1.28,
726
+ "learning_rate": 2.8732001956655824e-05,
727
+ "loss": 1.9343,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 1.29,
732
+ "learning_rate": 2.8554768639627954e-05,
733
+ "loss": 1.9446,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 1.3,
738
+ "learning_rate": 2.8377535322600084e-05,
739
+ "loss": 1.9325,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 1.31,
744
+ "learning_rate": 2.8200302005572217e-05,
745
+ "loss": 1.9321,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 1.32,
750
+ "learning_rate": 2.8023068688544347e-05,
751
+ "loss": 1.9295,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 1.33,
756
+ "learning_rate": 2.7845835371516483e-05,
757
+ "loss": 1.928,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 1.34,
762
+ "learning_rate": 2.7668602054488613e-05,
763
+ "loss": 1.939,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 1.35,
768
+ "learning_rate": 2.7491368737460743e-05,
769
+ "loss": 1.9172,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 1.36,
774
+ "learning_rate": 2.7314135420432873e-05,
775
+ "loss": 1.9185,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 1.37,
780
+ "learning_rate": 2.713690210340501e-05,
781
+ "loss": 1.9299,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 1.38,
786
+ "learning_rate": 2.695966878637714e-05,
787
+ "loss": 1.9301,
788
+ "step": 65000
789
+ },
790
+ {
791
+ "epoch": 1.39,
792
+ "learning_rate": 2.678243546934927e-05,
793
+ "loss": 1.9189,
794
+ "step": 65500
795
+ },
796
+ {
797
+ "epoch": 1.4,
798
+ "learning_rate": 2.6605202152321402e-05,
799
+ "loss": 1.9303,
800
+ "step": 66000
801
+ },
802
+ {
803
+ "epoch": 1.41,
804
+ "learning_rate": 2.6427968835293536e-05,
805
+ "loss": 1.9151,
806
+ "step": 66500
807
+ },
808
+ {
809
+ "epoch": 1.42,
810
+ "learning_rate": 2.625073551826567e-05,
811
+ "loss": 1.9236,
812
+ "step": 67000
813
+ },
814
+ {
815
+ "epoch": 1.44,
816
+ "learning_rate": 2.60735022012378e-05,
817
+ "loss": 1.9198,
818
+ "step": 67500
819
+ },
820
+ {
821
+ "epoch": 1.45,
822
+ "learning_rate": 2.589626888420993e-05,
823
+ "loss": 1.91,
824
+ "step": 68000
825
+ },
826
+ {
827
+ "epoch": 1.46,
828
+ "learning_rate": 2.5719035567182065e-05,
829
+ "loss": 1.8999,
830
+ "step": 68500
831
+ },
832
+ {
833
+ "epoch": 1.47,
834
+ "learning_rate": 2.5541802250154195e-05,
835
+ "loss": 1.9141,
836
+ "step": 69000
837
+ },
838
+ {
839
+ "epoch": 1.48,
840
+ "learning_rate": 2.5364568933126325e-05,
841
+ "loss": 1.9012,
842
+ "step": 69500
843
+ },
844
+ {
845
+ "epoch": 1.49,
846
+ "learning_rate": 2.5187335616098455e-05,
847
+ "loss": 1.8999,
848
+ "step": 70000
849
+ },
850
+ {
851
+ "epoch": 1.5,
852
+ "learning_rate": 2.501010229907059e-05,
853
+ "loss": 1.8923,
854
+ "step": 70500
855
+ },
856
+ {
857
+ "epoch": 1.51,
858
+ "learning_rate": 2.483286898204272e-05,
859
+ "loss": 1.9068,
860
+ "step": 71000
861
+ },
862
+ {
863
+ "epoch": 1.52,
864
+ "learning_rate": 2.4655635665014854e-05,
865
+ "loss": 1.9013,
866
+ "step": 71500
867
+ },
868
+ {
869
+ "epoch": 1.53,
870
+ "learning_rate": 2.4478402347986984e-05,
871
+ "loss": 1.9018,
872
+ "step": 72000
873
+ },
874
+ {
875
+ "epoch": 1.54,
876
+ "learning_rate": 2.4301169030959117e-05,
877
+ "loss": 1.8892,
878
+ "step": 72500
879
+ },
880
+ {
881
+ "epoch": 1.55,
882
+ "learning_rate": 2.4123935713931247e-05,
883
+ "loss": 1.9016,
884
+ "step": 73000
885
+ },
886
+ {
887
+ "epoch": 1.56,
888
+ "learning_rate": 2.394670239690338e-05,
889
+ "loss": 1.9058,
890
+ "step": 73500
891
+ },
892
+ {
893
+ "epoch": 1.57,
894
+ "learning_rate": 2.376946907987551e-05,
895
+ "loss": 1.91,
896
+ "step": 74000
897
+ },
898
+ {
899
+ "epoch": 1.58,
900
+ "learning_rate": 2.3592235762847644e-05,
901
+ "loss": 1.889,
902
+ "step": 74500
903
+ },
904
+ {
905
+ "epoch": 1.6,
906
+ "learning_rate": 2.3415002445819777e-05,
907
+ "loss": 1.8988,
908
+ "step": 75000
909
+ },
910
+ {
911
+ "epoch": 1.61,
912
+ "learning_rate": 2.323776912879191e-05,
913
+ "loss": 1.8957,
914
+ "step": 75500
915
+ },
916
+ {
917
+ "epoch": 1.62,
918
+ "learning_rate": 2.306053581176404e-05,
919
+ "loss": 1.8938,
920
+ "step": 76000
921
+ },
922
+ {
923
+ "epoch": 1.63,
924
+ "learning_rate": 2.2883302494736173e-05,
925
+ "loss": 1.8979,
926
+ "step": 76500
927
+ },
928
+ {
929
+ "epoch": 1.64,
930
+ "learning_rate": 2.2706069177708303e-05,
931
+ "loss": 1.8898,
932
+ "step": 77000
933
+ },
934
+ {
935
+ "epoch": 1.65,
936
+ "learning_rate": 2.2528835860680436e-05,
937
+ "loss": 1.891,
938
+ "step": 77500
939
+ },
940
+ {
941
+ "epoch": 1.66,
942
+ "learning_rate": 2.2351602543652566e-05,
943
+ "loss": 1.8858,
944
+ "step": 78000
945
+ },
946
+ {
947
+ "epoch": 1.67,
948
+ "learning_rate": 2.21743692266247e-05,
949
+ "loss": 1.8749,
950
+ "step": 78500
951
+ },
952
+ {
953
+ "epoch": 1.68,
954
+ "learning_rate": 2.199713590959683e-05,
955
+ "loss": 1.8833,
956
+ "step": 79000
957
+ },
958
+ {
959
+ "epoch": 1.69,
960
+ "learning_rate": 2.1819902592568962e-05,
961
+ "loss": 1.8788,
962
+ "step": 79500
963
+ },
964
+ {
965
+ "epoch": 1.7,
966
+ "learning_rate": 2.1642669275541096e-05,
967
+ "loss": 1.8809,
968
+ "step": 80000
969
+ },
970
+ {
971
+ "epoch": 1.71,
972
+ "learning_rate": 2.146543595851323e-05,
973
+ "loss": 1.8796,
974
+ "step": 80500
975
+ },
976
+ {
977
+ "epoch": 1.72,
978
+ "learning_rate": 2.128820264148536e-05,
979
+ "loss": 1.8863,
980
+ "step": 81000
981
+ },
982
+ {
983
+ "epoch": 1.73,
984
+ "learning_rate": 2.111096932445749e-05,
985
+ "loss": 1.8814,
986
+ "step": 81500
987
+ },
988
+ {
989
+ "epoch": 1.74,
990
+ "learning_rate": 2.0933736007429622e-05,
991
+ "loss": 1.8814,
992
+ "step": 82000
993
+ },
994
+ {
995
+ "epoch": 1.75,
996
+ "learning_rate": 2.075650269040175e-05,
997
+ "loss": 1.8654,
998
+ "step": 82500
999
+ },
1000
+ {
1001
+ "epoch": 1.77,
1002
+ "learning_rate": 2.0579269373373885e-05,
1003
+ "loss": 1.8806,
1004
+ "step": 83000
1005
+ },
1006
+ {
1007
+ "epoch": 1.78,
1008
+ "learning_rate": 2.0402036056346015e-05,
1009
+ "loss": 1.8659,
1010
+ "step": 83500
1011
+ },
1012
+ {
1013
+ "epoch": 1.79,
1014
+ "learning_rate": 2.0224802739318148e-05,
1015
+ "loss": 1.8837,
1016
+ "step": 84000
1017
+ },
1018
+ {
1019
+ "epoch": 1.8,
1020
+ "learning_rate": 2.004756942229028e-05,
1021
+ "loss": 1.8608,
1022
+ "step": 84500
1023
+ },
1024
+ {
1025
+ "epoch": 1.81,
1026
+ "learning_rate": 1.9870336105262414e-05,
1027
+ "loss": 1.8754,
1028
+ "step": 85000
1029
+ },
1030
+ {
1031
+ "epoch": 1.82,
1032
+ "learning_rate": 1.9693102788234544e-05,
1033
+ "loss": 1.8703,
1034
+ "step": 85500
1035
+ },
1036
+ {
1037
+ "epoch": 1.83,
1038
+ "learning_rate": 1.9515869471206677e-05,
1039
+ "loss": 1.8642,
1040
+ "step": 86000
1041
+ },
1042
+ {
1043
+ "epoch": 1.84,
1044
+ "learning_rate": 1.9338636154178807e-05,
1045
+ "loss": 1.8701,
1046
+ "step": 86500
1047
+ },
1048
+ {
1049
+ "epoch": 1.85,
1050
+ "learning_rate": 1.916140283715094e-05,
1051
+ "loss": 1.863,
1052
+ "step": 87000
1053
+ },
1054
+ {
1055
+ "epoch": 1.86,
1056
+ "learning_rate": 1.898416952012307e-05,
1057
+ "loss": 1.8672,
1058
+ "step": 87500
1059
+ },
1060
+ {
1061
+ "epoch": 1.87,
1062
+ "learning_rate": 1.8806936203095204e-05,
1063
+ "loss": 1.8638,
1064
+ "step": 88000
1065
+ },
1066
+ {
1067
+ "epoch": 1.88,
1068
+ "learning_rate": 1.8629702886067333e-05,
1069
+ "loss": 1.8709,
1070
+ "step": 88500
1071
+ },
1072
+ {
1073
+ "epoch": 1.89,
1074
+ "learning_rate": 1.845246956903947e-05,
1075
+ "loss": 1.8496,
1076
+ "step": 89000
1077
+ },
1078
+ {
1079
+ "epoch": 1.9,
1080
+ "learning_rate": 1.82752362520116e-05,
1081
+ "loss": 1.8601,
1082
+ "step": 89500
1083
+ },
1084
+ {
1085
+ "epoch": 1.91,
1086
+ "learning_rate": 1.8098002934983733e-05,
1087
+ "loss": 1.8628,
1088
+ "step": 90000
1089
+ },
1090
+ {
1091
+ "epoch": 1.92,
1092
+ "learning_rate": 1.7920769617955863e-05,
1093
+ "loss": 1.8551,
1094
+ "step": 90500
1095
+ },
1096
+ {
1097
+ "epoch": 1.94,
1098
+ "learning_rate": 1.7743536300927993e-05,
1099
+ "loss": 1.8493,
1100
+ "step": 91000
1101
+ },
1102
+ {
1103
+ "epoch": 1.95,
1104
+ "learning_rate": 1.7566302983900126e-05,
1105
+ "loss": 1.866,
1106
+ "step": 91500
1107
+ },
1108
+ {
1109
+ "epoch": 1.96,
1110
+ "learning_rate": 1.7389069666872256e-05,
1111
+ "loss": 1.8527,
1112
+ "step": 92000
1113
+ },
1114
+ {
1115
+ "epoch": 1.97,
1116
+ "learning_rate": 1.721183634984439e-05,
1117
+ "loss": 1.8541,
1118
+ "step": 92500
1119
+ },
1120
+ {
1121
+ "epoch": 1.98,
1122
+ "learning_rate": 1.7034603032816522e-05,
1123
+ "loss": 1.8483,
1124
+ "step": 93000
1125
+ },
1126
+ {
1127
+ "epoch": 1.99,
1128
+ "learning_rate": 1.6857369715788656e-05,
1129
+ "loss": 1.8555,
1130
+ "step": 93500
1131
+ },
1132
+ {
1133
+ "epoch": 2.0,
1134
+ "learning_rate": 1.6680136398760785e-05,
1135
+ "loss": 1.8524,
1136
+ "step": 94000
1137
+ },
1138
+ {
1139
+ "epoch": 2.01,
1140
+ "learning_rate": 1.650290308173292e-05,
1141
+ "loss": 1.8502,
1142
+ "step": 94500
1143
+ },
1144
+ {
1145
+ "epoch": 2.02,
1146
+ "learning_rate": 1.632566976470505e-05,
1147
+ "loss": 1.8332,
1148
+ "step": 95000
1149
+ },
1150
+ {
1151
+ "epoch": 2.03,
1152
+ "learning_rate": 1.614843644767718e-05,
1153
+ "loss": 1.8336,
1154
+ "step": 95500
1155
+ },
1156
+ {
1157
+ "epoch": 2.04,
1158
+ "learning_rate": 1.597120313064931e-05,
1159
+ "loss": 1.8362,
1160
+ "step": 96000
1161
+ },
1162
+ {
1163
+ "epoch": 2.05,
1164
+ "learning_rate": 1.5793969813621445e-05,
1165
+ "loss": 1.845,
1166
+ "step": 96500
1167
+ },
1168
+ {
1169
+ "epoch": 2.06,
1170
+ "learning_rate": 1.5616736496593575e-05,
1171
+ "loss": 1.8313,
1172
+ "step": 97000
1173
+ },
1174
+ {
1175
+ "epoch": 2.07,
1176
+ "learning_rate": 1.5439503179565708e-05,
1177
+ "loss": 1.8447,
1178
+ "step": 97500
1179
+ },
1180
+ {
1181
+ "epoch": 2.08,
1182
+ "learning_rate": 1.526226986253784e-05,
1183
+ "loss": 1.8309,
1184
+ "step": 98000
1185
+ },
1186
+ {
1187
+ "epoch": 2.09,
1188
+ "learning_rate": 1.5085036545509973e-05,
1189
+ "loss": 1.8321,
1190
+ "step": 98500
1191
+ },
1192
+ {
1193
+ "epoch": 2.11,
1194
+ "learning_rate": 1.4907803228482104e-05,
1195
+ "loss": 1.8326,
1196
+ "step": 99000
1197
+ },
1198
+ {
1199
+ "epoch": 2.12,
1200
+ "learning_rate": 1.4730569911454237e-05,
1201
+ "loss": 1.8386,
1202
+ "step": 99500
1203
+ },
1204
+ {
1205
+ "epoch": 2.13,
1206
+ "learning_rate": 1.4553336594426367e-05,
1207
+ "loss": 1.8316,
1208
+ "step": 100000
1209
+ },
1210
+ {
1211
+ "epoch": 2.14,
1212
+ "learning_rate": 1.43761032773985e-05,
1213
+ "loss": 1.8337,
1214
+ "step": 100500
1215
+ },
1216
+ {
1217
+ "epoch": 2.15,
1218
+ "learning_rate": 1.419886996037063e-05,
1219
+ "loss": 1.8395,
1220
+ "step": 101000
1221
+ },
1222
+ {
1223
+ "epoch": 2.16,
1224
+ "learning_rate": 1.4021636643342762e-05,
1225
+ "loss": 1.8345,
1226
+ "step": 101500
1227
+ },
1228
+ {
1229
+ "epoch": 2.17,
1230
+ "learning_rate": 1.3844403326314895e-05,
1231
+ "loss": 1.8324,
1232
+ "step": 102000
1233
+ },
1234
+ {
1235
+ "epoch": 2.18,
1236
+ "learning_rate": 1.3667170009287025e-05,
1237
+ "loss": 1.8222,
1238
+ "step": 102500
1239
+ },
1240
+ {
1241
+ "epoch": 2.19,
1242
+ "learning_rate": 1.3489936692259158e-05,
1243
+ "loss": 1.8077,
1244
+ "step": 103000
1245
+ },
1246
+ {
1247
+ "epoch": 2.2,
1248
+ "learning_rate": 1.331270337523129e-05,
1249
+ "loss": 1.8202,
1250
+ "step": 103500
1251
+ },
1252
+ {
1253
+ "epoch": 2.21,
1254
+ "learning_rate": 1.3135470058203423e-05,
1255
+ "loss": 1.8052,
1256
+ "step": 104000
1257
+ },
1258
+ {
1259
+ "epoch": 2.22,
1260
+ "learning_rate": 1.2958236741175553e-05,
1261
+ "loss": 1.8069,
1262
+ "step": 104500
1263
+ },
1264
+ {
1265
+ "epoch": 2.23,
1266
+ "learning_rate": 1.2781003424147686e-05,
1267
+ "loss": 1.8158,
1268
+ "step": 105000
1269
+ },
1270
+ {
1271
+ "epoch": 2.24,
1272
+ "learning_rate": 1.2603770107119817e-05,
1273
+ "loss": 1.8183,
1274
+ "step": 105500
1275
+ },
1276
+ {
1277
+ "epoch": 2.25,
1278
+ "learning_rate": 1.242653679009195e-05,
1279
+ "loss": 1.8116,
1280
+ "step": 106000
1281
+ },
1282
+ {
1283
+ "epoch": 2.27,
1284
+ "learning_rate": 1.224930347306408e-05,
1285
+ "loss": 1.8095,
1286
+ "step": 106500
1287
+ },
1288
+ {
1289
+ "epoch": 2.28,
1290
+ "learning_rate": 1.2072070156036212e-05,
1291
+ "loss": 1.8207,
1292
+ "step": 107000
1293
+ },
1294
+ {
1295
+ "epoch": 2.29,
1296
+ "learning_rate": 1.1894836839008344e-05,
1297
+ "loss": 1.8184,
1298
+ "step": 107500
1299
+ },
1300
+ {
1301
+ "epoch": 2.3,
1302
+ "learning_rate": 1.1717603521980477e-05,
1303
+ "loss": 1.8278,
1304
+ "step": 108000
1305
+ },
1306
+ {
1307
+ "epoch": 2.31,
1308
+ "learning_rate": 1.1540370204952608e-05,
1309
+ "loss": 1.8164,
1310
+ "step": 108500
1311
+ },
1312
+ {
1313
+ "epoch": 2.32,
1314
+ "learning_rate": 1.136313688792474e-05,
1315
+ "loss": 1.8244,
1316
+ "step": 109000
1317
+ },
1318
+ {
1319
+ "epoch": 2.33,
1320
+ "learning_rate": 1.1185903570896871e-05,
1321
+ "loss": 1.8199,
1322
+ "step": 109500
1323
+ },
1324
+ {
1325
+ "epoch": 2.34,
1326
+ "learning_rate": 1.1008670253869003e-05,
1327
+ "loss": 1.8252,
1328
+ "step": 110000
1329
+ },
1330
+ {
1331
+ "epoch": 2.35,
1332
+ "learning_rate": 1.0831436936841136e-05,
1333
+ "loss": 1.808,
1334
+ "step": 110500
1335
+ },
1336
+ {
1337
+ "epoch": 2.36,
1338
+ "learning_rate": 1.0654203619813268e-05,
1339
+ "loss": 1.8097,
1340
+ "step": 111000
1341
+ },
1342
+ {
1343
+ "epoch": 2.37,
1344
+ "learning_rate": 1.04769703027854e-05,
1345
+ "loss": 1.8049,
1346
+ "step": 111500
1347
+ },
1348
+ {
1349
+ "epoch": 2.38,
1350
+ "learning_rate": 1.029973698575753e-05,
1351
+ "loss": 1.8081,
1352
+ "step": 112000
1353
+ },
1354
+ {
1355
+ "epoch": 2.39,
1356
+ "learning_rate": 1.0122503668729662e-05,
1357
+ "loss": 1.8134,
1358
+ "step": 112500
1359
+ },
1360
+ {
1361
+ "epoch": 2.4,
1362
+ "learning_rate": 9.945270351701796e-06,
1363
+ "loss": 1.7976,
1364
+ "step": 113000
1365
+ },
1366
+ {
1367
+ "epoch": 2.41,
1368
+ "learning_rate": 9.768037034673927e-06,
1369
+ "loss": 1.8036,
1370
+ "step": 113500
1371
+ },
1372
+ {
1373
+ "epoch": 2.42,
1374
+ "learning_rate": 9.590803717646059e-06,
1375
+ "loss": 1.7979,
1376
+ "step": 114000
1377
+ },
1378
+ {
1379
+ "epoch": 2.44,
1380
+ "learning_rate": 9.41357040061819e-06,
1381
+ "loss": 1.8165,
1382
+ "step": 114500
1383
+ },
1384
+ {
1385
+ "epoch": 2.45,
1386
+ "learning_rate": 9.236337083590323e-06,
1387
+ "loss": 1.8069,
1388
+ "step": 115000
1389
+ },
1390
+ {
1391
+ "epoch": 2.46,
1392
+ "learning_rate": 9.059103766562455e-06,
1393
+ "loss": 1.7922,
1394
+ "step": 115500
1395
+ },
1396
+ {
1397
+ "epoch": 2.47,
1398
+ "learning_rate": 8.881870449534587e-06,
1399
+ "loss": 1.7981,
1400
+ "step": 116000
1401
+ },
1402
+ {
1403
+ "epoch": 2.48,
1404
+ "learning_rate": 8.704637132506716e-06,
1405
+ "loss": 1.7998,
1406
+ "step": 116500
1407
+ },
1408
+ {
1409
+ "epoch": 2.49,
1410
+ "learning_rate": 8.52740381547885e-06,
1411
+ "loss": 1.7938,
1412
+ "step": 117000
1413
+ },
1414
+ {
1415
+ "epoch": 2.5,
1416
+ "learning_rate": 8.350170498450981e-06,
1417
+ "loss": 1.8041,
1418
+ "step": 117500
1419
+ },
1420
+ {
1421
+ "epoch": 2.51,
1422
+ "learning_rate": 8.172937181423113e-06,
1423
+ "loss": 1.8037,
1424
+ "step": 118000
1425
+ },
1426
+ {
1427
+ "epoch": 2.52,
1428
+ "learning_rate": 7.995703864395244e-06,
1429
+ "loss": 1.7945,
1430
+ "step": 118500
1431
+ },
1432
+ {
1433
+ "epoch": 2.53,
1434
+ "learning_rate": 7.818470547367376e-06,
1435
+ "loss": 1.8138,
1436
+ "step": 119000
1437
+ },
1438
+ {
1439
+ "epoch": 2.54,
1440
+ "learning_rate": 7.641237230339509e-06,
1441
+ "loss": 1.792,
1442
+ "step": 119500
1443
+ },
1444
+ {
1445
+ "epoch": 2.55,
1446
+ "learning_rate": 7.4640039133116405e-06,
1447
+ "loss": 1.7963,
1448
+ "step": 120000
1449
+ },
1450
+ {
1451
+ "epoch": 2.56,
1452
+ "learning_rate": 7.286770596283772e-06,
1453
+ "loss": 1.7906,
1454
+ "step": 120500
1455
+ },
1456
+ {
1457
+ "epoch": 2.57,
1458
+ "learning_rate": 7.1095372792559036e-06,
1459
+ "loss": 1.7933,
1460
+ "step": 121000
1461
+ },
1462
+ {
1463
+ "epoch": 2.58,
1464
+ "learning_rate": 6.932303962228036e-06,
1465
+ "loss": 1.7877,
1466
+ "step": 121500
1467
+ },
1468
+ {
1469
+ "epoch": 2.59,
1470
+ "learning_rate": 6.7550706452001675e-06,
1471
+ "loss": 1.7999,
1472
+ "step": 122000
1473
+ },
1474
+ {
1475
+ "epoch": 2.61,
1476
+ "learning_rate": 6.5778373281723e-06,
1477
+ "loss": 1.7839,
1478
+ "step": 122500
1479
+ },
1480
+ {
1481
+ "epoch": 2.62,
1482
+ "learning_rate": 6.400604011144431e-06,
1483
+ "loss": 1.7907,
1484
+ "step": 123000
1485
+ },
1486
+ {
1487
+ "epoch": 2.63,
1488
+ "learning_rate": 6.223370694116564e-06,
1489
+ "loss": 1.7869,
1490
+ "step": 123500
1491
+ },
1492
+ {
1493
+ "epoch": 2.64,
1494
+ "learning_rate": 6.0461373770886945e-06,
1495
+ "loss": 1.7736,
1496
+ "step": 124000
1497
+ },
1498
+ {
1499
+ "epoch": 2.65,
1500
+ "learning_rate": 5.868904060060827e-06,
1501
+ "loss": 1.7825,
1502
+ "step": 124500
1503
+ },
1504
+ {
1505
+ "epoch": 2.66,
1506
+ "learning_rate": 5.691670743032958e-06,
1507
+ "loss": 1.7831,
1508
+ "step": 125000
1509
+ },
1510
+ {
1511
+ "epoch": 2.67,
1512
+ "learning_rate": 5.51443742600509e-06,
1513
+ "loss": 1.7831,
1514
+ "step": 125500
1515
+ },
1516
+ {
1517
+ "epoch": 2.68,
1518
+ "learning_rate": 5.337204108977222e-06,
1519
+ "loss": 1.7848,
1520
+ "step": 126000
1521
+ },
1522
+ {
1523
+ "epoch": 2.69,
1524
+ "learning_rate": 5.159970791949354e-06,
1525
+ "loss": 1.793,
1526
+ "step": 126500
1527
+ },
1528
+ {
1529
+ "epoch": 2.7,
1530
+ "learning_rate": 4.982737474921486e-06,
1531
+ "loss": 1.7887,
1532
+ "step": 127000
1533
+ },
1534
+ {
1535
+ "epoch": 2.71,
1536
+ "learning_rate": 4.805504157893618e-06,
1537
+ "loss": 1.7824,
1538
+ "step": 127500
1539
+ },
1540
+ {
1541
+ "epoch": 2.72,
1542
+ "learning_rate": 4.62827084086575e-06,
1543
+ "loss": 1.7885,
1544
+ "step": 128000
1545
+ },
1546
+ {
1547
+ "epoch": 2.73,
1548
+ "learning_rate": 4.451037523837882e-06,
1549
+ "loss": 1.7892,
1550
+ "step": 128500
1551
+ },
1552
+ {
1553
+ "epoch": 2.74,
1554
+ "learning_rate": 4.273804206810013e-06,
1555
+ "loss": 1.7832,
1556
+ "step": 129000
1557
+ },
1558
+ {
1559
+ "epoch": 2.75,
1560
+ "learning_rate": 4.096570889782145e-06,
1561
+ "loss": 1.7705,
1562
+ "step": 129500
1563
+ },
1564
+ {
1565
+ "epoch": 2.76,
1566
+ "learning_rate": 3.919337572754276e-06,
1567
+ "loss": 1.7775,
1568
+ "step": 130000
1569
+ },
1570
+ {
1571
+ "epoch": 2.78,
1572
+ "learning_rate": 3.7421042557264087e-06,
1573
+ "loss": 1.7737,
1574
+ "step": 130500
1575
+ },
1576
+ {
1577
+ "epoch": 2.79,
1578
+ "learning_rate": 3.5648709386985406e-06,
1579
+ "loss": 1.7858,
1580
+ "step": 131000
1581
+ },
1582
+ {
1583
+ "epoch": 2.8,
1584
+ "learning_rate": 3.387637621670672e-06,
1585
+ "loss": 1.784,
1586
+ "step": 131500
1587
+ },
1588
+ {
1589
+ "epoch": 2.81,
1590
+ "learning_rate": 3.210404304642804e-06,
1591
+ "loss": 1.7766,
1592
+ "step": 132000
1593
+ },
1594
+ {
1595
+ "epoch": 2.82,
1596
+ "learning_rate": 3.033170987614936e-06,
1597
+ "loss": 1.7806,
1598
+ "step": 132500
1599
+ },
1600
+ {
1601
+ "epoch": 2.83,
1602
+ "learning_rate": 2.8559376705870676e-06,
1603
+ "loss": 1.7574,
1604
+ "step": 133000
1605
+ },
1606
+ {
1607
+ "epoch": 2.84,
1608
+ "learning_rate": 2.6787043535591996e-06,
1609
+ "loss": 1.7697,
1610
+ "step": 133500
1611
+ },
1612
+ {
1613
+ "epoch": 2.85,
1614
+ "learning_rate": 2.5014710365313316e-06,
1615
+ "loss": 1.7834,
1616
+ "step": 134000
1617
+ },
1618
+ {
1619
+ "epoch": 2.86,
1620
+ "learning_rate": 2.3242377195034635e-06,
1621
+ "loss": 1.7644,
1622
+ "step": 134500
1623
+ },
1624
+ {
1625
+ "epoch": 2.87,
1626
+ "learning_rate": 2.147004402475595e-06,
1627
+ "loss": 1.7794,
1628
+ "step": 135000
1629
+ },
1630
+ {
1631
+ "epoch": 2.88,
1632
+ "learning_rate": 1.9697710854477266e-06,
1633
+ "loss": 1.7657,
1634
+ "step": 135500
1635
+ },
1636
+ {
1637
+ "epoch": 2.89,
1638
+ "learning_rate": 1.7925377684198588e-06,
1639
+ "loss": 1.7786,
1640
+ "step": 136000
1641
+ },
1642
+ {
1643
+ "epoch": 2.9,
1644
+ "learning_rate": 1.6153044513919905e-06,
1645
+ "loss": 1.7756,
1646
+ "step": 136500
1647
+ },
1648
+ {
1649
+ "epoch": 2.91,
1650
+ "learning_rate": 1.4380711343641223e-06,
1651
+ "loss": 1.7742,
1652
+ "step": 137000
1653
+ },
1654
+ {
1655
+ "epoch": 2.92,
1656
+ "learning_rate": 1.2608378173362542e-06,
1657
+ "loss": 1.7711,
1658
+ "step": 137500
1659
+ },
1660
+ {
1661
+ "epoch": 2.93,
1662
+ "learning_rate": 1.0836045003083862e-06,
1663
+ "loss": 1.7596,
1664
+ "step": 138000
1665
+ },
1666
+ {
1667
+ "epoch": 2.95,
1668
+ "learning_rate": 9.063711832805178e-07,
1669
+ "loss": 1.7777,
1670
+ "step": 138500
1671
+ },
1672
+ {
1673
+ "epoch": 2.96,
1674
+ "learning_rate": 7.291378662526497e-07,
1675
+ "loss": 1.7767,
1676
+ "step": 139000
1677
+ },
1678
+ {
1679
+ "epoch": 2.97,
1680
+ "learning_rate": 5.519045492247815e-07,
1681
+ "loss": 1.7738,
1682
+ "step": 139500
1683
+ },
1684
+ {
1685
+ "epoch": 2.98,
1686
+ "learning_rate": 3.746712321969133e-07,
1687
+ "loss": 1.787,
1688
+ "step": 140000
1689
+ },
1690
+ {
1691
+ "epoch": 2.99,
1692
+ "learning_rate": 1.9743791516904515e-07,
1693
+ "loss": 1.7725,
1694
+ "step": 140500
1695
+ },
1696
+ {
1697
+ "epoch": 3.0,
1698
+ "learning_rate": 2.0204598141176972e-08,
1699
+ "loss": 1.7788,
1700
+ "step": 141000
1701
+ },
1702
+ {
1703
+ "epoch": 3.0,
1704
+ "step": 141057,
1705
+ "total_flos": 3.722159267218719e+17,
1706
+ "train_loss": 1.9451130962921488,
1707
+ "train_runtime": 85182.9847,
1708
+ "train_samples_per_second": 16.559,
1709
+ "train_steps_per_second": 1.656
1710
+ }
1711
+ ],
1712
+ "max_steps": 141057,
1713
+ "num_train_epochs": 3,
1714
+ "total_flos": 3.722159267218719e+17,
1715
+ "trial_name": null,
1716
+ "trial_params": null
1717
+ }