Davlan commited on
Commit
db33061
1 Parent(s): 58b3969

Upload trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +1771 -0
trainer_state.json ADDED
@@ -0,0 +1,1771 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "global_step": 145677,
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": 4.982838745992847e-05,
13
+ "loss": 3.161,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.02,
18
+ "learning_rate": 4.965677491985695e-05,
19
+ "loss": 2.686,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.03,
24
+ "learning_rate": 4.948516237978542e-05,
25
+ "loss": 2.5,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.04,
30
+ "learning_rate": 4.9313549839713894e-05,
31
+ "loss": 2.3421,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.05,
36
+ "learning_rate": 4.914193729964236e-05,
37
+ "loss": 2.3123,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.06,
42
+ "learning_rate": 4.897032475957083e-05,
43
+ "loss": 2.2321,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.07,
48
+ "learning_rate": 4.87987122194993e-05,
49
+ "loss": 2.1761,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.08,
54
+ "learning_rate": 4.862709967942778e-05,
55
+ "loss": 2.1261,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.09,
60
+ "learning_rate": 4.845548713935625e-05,
61
+ "loss": 2.109,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.1,
66
+ "learning_rate": 4.8283874599284724e-05,
67
+ "loss": 2.0872,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.11,
72
+ "learning_rate": 4.811226205921319e-05,
73
+ "loss": 2.036,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.12,
78
+ "learning_rate": 4.794064951914167e-05,
79
+ "loss": 2.0286,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.13,
84
+ "learning_rate": 4.776903697907014e-05,
85
+ "loss": 1.9971,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.14,
90
+ "learning_rate": 4.759742443899861e-05,
91
+ "loss": 1.9656,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.15,
96
+ "learning_rate": 4.742581189892708e-05,
97
+ "loss": 1.9784,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.16,
102
+ "learning_rate": 4.725419935885555e-05,
103
+ "loss": 1.924,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.18,
108
+ "learning_rate": 4.708258681878402e-05,
109
+ "loss": 1.9137,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.19,
114
+ "learning_rate": 4.691097427871249e-05,
115
+ "loss": 1.9194,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.2,
120
+ "learning_rate": 4.673936173864097e-05,
121
+ "loss": 1.8744,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.21,
126
+ "learning_rate": 4.656774919856944e-05,
127
+ "loss": 1.8964,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.22,
132
+ "learning_rate": 4.6396136658497914e-05,
133
+ "loss": 1.8515,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.23,
138
+ "learning_rate": 4.6224524118426383e-05,
139
+ "loss": 1.842,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.24,
144
+ "learning_rate": 4.605291157835486e-05,
145
+ "loss": 1.8577,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.25,
150
+ "learning_rate": 4.588129903828333e-05,
151
+ "loss": 1.8402,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.26,
156
+ "learning_rate": 4.57096864982118e-05,
157
+ "loss": 1.8155,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.27,
162
+ "learning_rate": 4.553807395814027e-05,
163
+ "loss": 1.8154,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.28,
168
+ "learning_rate": 4.5366461418068744e-05,
169
+ "loss": 1.796,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.29,
174
+ "learning_rate": 4.5194848877997213e-05,
175
+ "loss": 1.7842,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.3,
180
+ "learning_rate": 4.502323633792569e-05,
181
+ "loss": 1.7819,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.31,
186
+ "learning_rate": 4.485162379785416e-05,
187
+ "loss": 1.7769,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.32,
192
+ "learning_rate": 4.4680011257782635e-05,
193
+ "loss": 1.7649,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.33,
198
+ "learning_rate": 4.4508398717711105e-05,
199
+ "loss": 1.7653,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.34,
204
+ "learning_rate": 4.4336786177639574e-05,
205
+ "loss": 1.7418,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.35,
210
+ "learning_rate": 4.416517363756805e-05,
211
+ "loss": 1.7284,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.36,
216
+ "learning_rate": 4.399356109749652e-05,
217
+ "loss": 1.7218,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.37,
222
+ "learning_rate": 4.382194855742499e-05,
223
+ "loss": 1.719,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.38,
228
+ "learning_rate": 4.365033601735346e-05,
229
+ "loss": 1.6981,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.39,
234
+ "learning_rate": 4.3478723477281935e-05,
235
+ "loss": 1.702,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.4,
240
+ "learning_rate": 4.3307110937210404e-05,
241
+ "loss": 1.7055,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.41,
246
+ "learning_rate": 4.313549839713888e-05,
247
+ "loss": 1.7072,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.42,
252
+ "learning_rate": 4.296388585706735e-05,
253
+ "loss": 1.6702,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.43,
258
+ "learning_rate": 4.2792273316995826e-05,
259
+ "loss": 1.6595,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.44,
264
+ "learning_rate": 4.2620660776924295e-05,
265
+ "loss": 1.6889,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.45,
270
+ "learning_rate": 4.2449048236852765e-05,
271
+ "loss": 1.6576,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.46,
276
+ "learning_rate": 4.2277435696781234e-05,
277
+ "loss": 1.6732,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.47,
282
+ "learning_rate": 4.210582315670971e-05,
283
+ "loss": 1.6441,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.48,
288
+ "learning_rate": 4.193421061663818e-05,
289
+ "loss": 1.6511,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.49,
294
+ "learning_rate": 4.176259807656665e-05,
295
+ "loss": 1.6529,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.5,
300
+ "learning_rate": 4.1590985536495125e-05,
301
+ "loss": 1.6562,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.51,
306
+ "learning_rate": 4.1419372996423594e-05,
307
+ "loss": 1.6359,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.53,
312
+ "learning_rate": 4.124776045635207e-05,
313
+ "loss": 1.6404,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.54,
318
+ "learning_rate": 4.107614791628054e-05,
319
+ "loss": 1.6256,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.55,
324
+ "learning_rate": 4.0904535376209016e-05,
325
+ "loss": 1.6337,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.56,
330
+ "learning_rate": 4.0732922836137486e-05,
331
+ "loss": 1.6198,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.57,
336
+ "learning_rate": 4.0561310296065955e-05,
337
+ "loss": 1.616,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.58,
342
+ "learning_rate": 4.0389697755994424e-05,
343
+ "loss": 1.6317,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.59,
348
+ "learning_rate": 4.02180852159229e-05,
349
+ "loss": 1.6145,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.6,
354
+ "learning_rate": 4.004647267585137e-05,
355
+ "loss": 1.6096,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.61,
360
+ "learning_rate": 3.9874860135779846e-05,
361
+ "loss": 1.5697,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.62,
366
+ "learning_rate": 3.9703247595708316e-05,
367
+ "loss": 1.5838,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.63,
372
+ "learning_rate": 3.953163505563679e-05,
373
+ "loss": 1.5767,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.64,
378
+ "learning_rate": 3.936002251556526e-05,
379
+ "loss": 1.5776,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 0.65,
384
+ "learning_rate": 3.918840997549373e-05,
385
+ "loss": 1.5802,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 0.66,
390
+ "learning_rate": 3.90167974354222e-05,
391
+ "loss": 1.5711,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 0.67,
396
+ "learning_rate": 3.884518489535067e-05,
397
+ "loss": 1.5662,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 0.68,
402
+ "learning_rate": 3.8673572355279146e-05,
403
+ "loss": 1.592,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 0.69,
408
+ "learning_rate": 3.8501959815207615e-05,
409
+ "loss": 1.5665,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 0.7,
414
+ "learning_rate": 3.833034727513609e-05,
415
+ "loss": 1.5698,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 0.71,
420
+ "learning_rate": 3.815873473506456e-05,
421
+ "loss": 1.5548,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 0.72,
426
+ "learning_rate": 3.798712219499304e-05,
427
+ "loss": 1.5993,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 0.73,
432
+ "learning_rate": 3.7815509654921506e-05,
433
+ "loss": 1.5348,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 0.74,
438
+ "learning_rate": 3.764389711484998e-05,
439
+ "loss": 1.5567,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 0.75,
444
+ "learning_rate": 3.747228457477845e-05,
445
+ "loss": 1.566,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 0.76,
450
+ "learning_rate": 3.730067203470692e-05,
451
+ "loss": 1.5338,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 0.77,
456
+ "learning_rate": 3.712905949463539e-05,
457
+ "loss": 1.5232,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 0.78,
462
+ "learning_rate": 3.695744695456387e-05,
463
+ "loss": 1.5297,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 0.79,
468
+ "learning_rate": 3.6785834414492336e-05,
469
+ "loss": 1.5412,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 0.8,
474
+ "learning_rate": 3.661422187442081e-05,
475
+ "loss": 1.5383,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 0.81,
480
+ "learning_rate": 3.644260933434928e-05,
481
+ "loss": 1.5126,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 0.82,
486
+ "learning_rate": 3.627099679427775e-05,
487
+ "loss": 1.4971,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 0.83,
492
+ "learning_rate": 3.609938425420623e-05,
493
+ "loss": 1.4952,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 0.84,
498
+ "learning_rate": 3.59277717141347e-05,
499
+ "loss": 1.5381,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 0.85,
504
+ "learning_rate": 3.575615917406317e-05,
505
+ "loss": 1.5049,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 0.86,
510
+ "learning_rate": 3.558454663399164e-05,
511
+ "loss": 1.499,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 0.88,
516
+ "learning_rate": 3.541293409392011e-05,
517
+ "loss": 1.5128,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 0.89,
522
+ "learning_rate": 3.524132155384858e-05,
523
+ "loss": 1.4833,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 0.9,
528
+ "learning_rate": 3.506970901377706e-05,
529
+ "loss": 1.5054,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 0.91,
534
+ "learning_rate": 3.489809647370553e-05,
535
+ "loss": 1.4875,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 0.92,
540
+ "learning_rate": 3.4726483933634e-05,
541
+ "loss": 1.4746,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 0.93,
546
+ "learning_rate": 3.455487139356247e-05,
547
+ "loss": 1.5014,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 0.94,
552
+ "learning_rate": 3.438325885349095e-05,
553
+ "loss": 1.4916,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 0.95,
558
+ "learning_rate": 3.421164631341942e-05,
559
+ "loss": 1.5108,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 0.96,
564
+ "learning_rate": 3.404003377334789e-05,
565
+ "loss": 1.4931,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 0.97,
570
+ "learning_rate": 3.386842123327636e-05,
571
+ "loss": 1.4794,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 0.98,
576
+ "learning_rate": 3.3696808693204826e-05,
577
+ "loss": 1.482,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 0.99,
582
+ "learning_rate": 3.35251961531333e-05,
583
+ "loss": 1.454,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 1.0,
588
+ "learning_rate": 3.335358361306177e-05,
589
+ "loss": 1.4703,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 1.01,
594
+ "learning_rate": 3.318197107299025e-05,
595
+ "loss": 1.3302,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 1.02,
600
+ "learning_rate": 3.301035853291872e-05,
601
+ "loss": 1.3441,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 1.03,
606
+ "learning_rate": 3.283874599284719e-05,
607
+ "loss": 1.311,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 1.04,
612
+ "learning_rate": 3.266713345277566e-05,
613
+ "loss": 1.3116,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 1.05,
618
+ "learning_rate": 3.249552091270414e-05,
619
+ "loss": 1.3205,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 1.06,
624
+ "learning_rate": 3.232390837263261e-05,
625
+ "loss": 1.3269,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 1.07,
630
+ "learning_rate": 3.215229583256108e-05,
631
+ "loss": 1.3125,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 1.08,
636
+ "learning_rate": 3.198068329248955e-05,
637
+ "loss": 1.3155,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 1.09,
642
+ "learning_rate": 3.180907075241802e-05,
643
+ "loss": 1.2983,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 1.1,
648
+ "learning_rate": 3.163745821234649e-05,
649
+ "loss": 1.3261,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 1.11,
654
+ "learning_rate": 3.146584567227497e-05,
655
+ "loss": 1.3233,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 1.12,
660
+ "learning_rate": 3.129423313220344e-05,
661
+ "loss": 1.3272,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 1.13,
666
+ "learning_rate": 3.1122620592131915e-05,
667
+ "loss": 1.3353,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 1.14,
672
+ "learning_rate": 3.0951008052060384e-05,
673
+ "loss": 1.3086,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 1.15,
678
+ "learning_rate": 3.077939551198885e-05,
679
+ "loss": 1.3211,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 1.16,
684
+ "learning_rate": 3.060778297191732e-05,
685
+ "loss": 1.2961,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 1.17,
690
+ "learning_rate": 3.0436170431845795e-05,
691
+ "loss": 1.3105,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 1.18,
696
+ "learning_rate": 3.026455789177427e-05,
697
+ "loss": 1.3043,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 1.19,
702
+ "learning_rate": 3.0092945351702738e-05,
703
+ "loss": 1.2852,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 1.2,
708
+ "learning_rate": 2.9921332811631214e-05,
709
+ "loss": 1.3173,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 1.22,
714
+ "learning_rate": 2.9749720271559683e-05,
715
+ "loss": 1.2992,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 1.23,
720
+ "learning_rate": 2.957810773148816e-05,
721
+ "loss": 1.3263,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 1.24,
726
+ "learning_rate": 2.940649519141663e-05,
727
+ "loss": 1.3091,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 1.25,
732
+ "learning_rate": 2.92348826513451e-05,
733
+ "loss": 1.2984,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 1.26,
738
+ "learning_rate": 2.906327011127357e-05,
739
+ "loss": 1.3037,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 1.27,
744
+ "learning_rate": 2.8891657571202047e-05,
745
+ "loss": 1.3045,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 1.28,
750
+ "learning_rate": 2.8720045031130517e-05,
751
+ "loss": 1.2961,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 1.29,
756
+ "learning_rate": 2.854843249105899e-05,
757
+ "loss": 1.306,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 1.3,
762
+ "learning_rate": 2.837681995098746e-05,
763
+ "loss": 1.3032,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 1.31,
768
+ "learning_rate": 2.8205207410915928e-05,
769
+ "loss": 1.2992,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 1.32,
774
+ "learning_rate": 2.8033594870844404e-05,
775
+ "loss": 1.3061,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 1.33,
780
+ "learning_rate": 2.7861982330772874e-05,
781
+ "loss": 1.3085,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 1.34,
786
+ "learning_rate": 2.7690369790701347e-05,
787
+ "loss": 1.312,
788
+ "step": 65000
789
+ },
790
+ {
791
+ "epoch": 1.35,
792
+ "learning_rate": 2.7518757250629816e-05,
793
+ "loss": 1.2912,
794
+ "step": 65500
795
+ },
796
+ {
797
+ "epoch": 1.36,
798
+ "learning_rate": 2.7347144710558292e-05,
799
+ "loss": 1.3026,
800
+ "step": 66000
801
+ },
802
+ {
803
+ "epoch": 1.37,
804
+ "learning_rate": 2.717553217048676e-05,
805
+ "loss": 1.2848,
806
+ "step": 66500
807
+ },
808
+ {
809
+ "epoch": 1.38,
810
+ "learning_rate": 2.7003919630415238e-05,
811
+ "loss": 1.3018,
812
+ "step": 67000
813
+ },
814
+ {
815
+ "epoch": 1.39,
816
+ "learning_rate": 2.6832307090343707e-05,
817
+ "loss": 1.2972,
818
+ "step": 67500
819
+ },
820
+ {
821
+ "epoch": 1.4,
822
+ "learning_rate": 2.666069455027218e-05,
823
+ "loss": 1.2736,
824
+ "step": 68000
825
+ },
826
+ {
827
+ "epoch": 1.41,
828
+ "learning_rate": 2.648908201020065e-05,
829
+ "loss": 1.2721,
830
+ "step": 68500
831
+ },
832
+ {
833
+ "epoch": 1.42,
834
+ "learning_rate": 2.6317469470129126e-05,
835
+ "loss": 1.2921,
836
+ "step": 69000
837
+ },
838
+ {
839
+ "epoch": 1.43,
840
+ "learning_rate": 2.6145856930057595e-05,
841
+ "loss": 1.287,
842
+ "step": 69500
843
+ },
844
+ {
845
+ "epoch": 1.44,
846
+ "learning_rate": 2.5974244389986068e-05,
847
+ "loss": 1.286,
848
+ "step": 70000
849
+ },
850
+ {
851
+ "epoch": 1.45,
852
+ "learning_rate": 2.5802631849914537e-05,
853
+ "loss": 1.2839,
854
+ "step": 70500
855
+ },
856
+ {
857
+ "epoch": 1.46,
858
+ "learning_rate": 2.5631019309843007e-05,
859
+ "loss": 1.262,
860
+ "step": 71000
861
+ },
862
+ {
863
+ "epoch": 1.47,
864
+ "learning_rate": 2.5459406769771483e-05,
865
+ "loss": 1.2712,
866
+ "step": 71500
867
+ },
868
+ {
869
+ "epoch": 1.48,
870
+ "learning_rate": 2.5287794229699952e-05,
871
+ "loss": 1.2876,
872
+ "step": 72000
873
+ },
874
+ {
875
+ "epoch": 1.49,
876
+ "learning_rate": 2.5116181689628425e-05,
877
+ "loss": 1.2771,
878
+ "step": 72500
879
+ },
880
+ {
881
+ "epoch": 1.5,
882
+ "learning_rate": 2.4944569149556898e-05,
883
+ "loss": 1.2536,
884
+ "step": 73000
885
+ },
886
+ {
887
+ "epoch": 1.51,
888
+ "learning_rate": 2.477295660948537e-05,
889
+ "loss": 1.2693,
890
+ "step": 73500
891
+ },
892
+ {
893
+ "epoch": 1.52,
894
+ "learning_rate": 2.4601344069413843e-05,
895
+ "loss": 1.2467,
896
+ "step": 74000
897
+ },
898
+ {
899
+ "epoch": 1.53,
900
+ "learning_rate": 2.4429731529342316e-05,
901
+ "loss": 1.2886,
902
+ "step": 74500
903
+ },
904
+ {
905
+ "epoch": 1.54,
906
+ "learning_rate": 2.4258118989270785e-05,
907
+ "loss": 1.2981,
908
+ "step": 75000
909
+ },
910
+ {
911
+ "epoch": 1.55,
912
+ "learning_rate": 2.4086506449199255e-05,
913
+ "loss": 1.2729,
914
+ "step": 75500
915
+ },
916
+ {
917
+ "epoch": 1.57,
918
+ "learning_rate": 2.3914893909127728e-05,
919
+ "loss": 1.2682,
920
+ "step": 76000
921
+ },
922
+ {
923
+ "epoch": 1.58,
924
+ "learning_rate": 2.37432813690562e-05,
925
+ "loss": 1.2577,
926
+ "step": 76500
927
+ },
928
+ {
929
+ "epoch": 1.59,
930
+ "learning_rate": 2.3571668828984673e-05,
931
+ "loss": 1.2735,
932
+ "step": 77000
933
+ },
934
+ {
935
+ "epoch": 1.6,
936
+ "learning_rate": 2.3400056288913143e-05,
937
+ "loss": 1.2663,
938
+ "step": 77500
939
+ },
940
+ {
941
+ "epoch": 1.61,
942
+ "learning_rate": 2.3228443748841615e-05,
943
+ "loss": 1.272,
944
+ "step": 78000
945
+ },
946
+ {
947
+ "epoch": 1.62,
948
+ "learning_rate": 2.3056831208770088e-05,
949
+ "loss": 1.2679,
950
+ "step": 78500
951
+ },
952
+ {
953
+ "epoch": 1.63,
954
+ "learning_rate": 2.288521866869856e-05,
955
+ "loss": 1.2369,
956
+ "step": 79000
957
+ },
958
+ {
959
+ "epoch": 1.64,
960
+ "learning_rate": 2.2713606128627034e-05,
961
+ "loss": 1.2633,
962
+ "step": 79500
963
+ },
964
+ {
965
+ "epoch": 1.65,
966
+ "learning_rate": 2.2541993588555503e-05,
967
+ "loss": 1.2504,
968
+ "step": 80000
969
+ },
970
+ {
971
+ "epoch": 1.66,
972
+ "learning_rate": 2.2370381048483976e-05,
973
+ "loss": 1.2702,
974
+ "step": 80500
975
+ },
976
+ {
977
+ "epoch": 1.67,
978
+ "learning_rate": 2.219876850841245e-05,
979
+ "loss": 1.2467,
980
+ "step": 81000
981
+ },
982
+ {
983
+ "epoch": 1.68,
984
+ "learning_rate": 2.202715596834092e-05,
985
+ "loss": 1.2392,
986
+ "step": 81500
987
+ },
988
+ {
989
+ "epoch": 1.69,
990
+ "learning_rate": 2.1855543428269394e-05,
991
+ "loss": 1.257,
992
+ "step": 82000
993
+ },
994
+ {
995
+ "epoch": 1.7,
996
+ "learning_rate": 2.1683930888197864e-05,
997
+ "loss": 1.2431,
998
+ "step": 82500
999
+ },
1000
+ {
1001
+ "epoch": 1.71,
1002
+ "learning_rate": 2.1512318348126333e-05,
1003
+ "loss": 1.2486,
1004
+ "step": 83000
1005
+ },
1006
+ {
1007
+ "epoch": 1.72,
1008
+ "learning_rate": 2.1340705808054806e-05,
1009
+ "loss": 1.212,
1010
+ "step": 83500
1011
+ },
1012
+ {
1013
+ "epoch": 1.73,
1014
+ "learning_rate": 2.116909326798328e-05,
1015
+ "loss": 1.2743,
1016
+ "step": 84000
1017
+ },
1018
+ {
1019
+ "epoch": 1.74,
1020
+ "learning_rate": 2.099748072791175e-05,
1021
+ "loss": 1.2249,
1022
+ "step": 84500
1023
+ },
1024
+ {
1025
+ "epoch": 1.75,
1026
+ "learning_rate": 2.082586818784022e-05,
1027
+ "loss": 1.2234,
1028
+ "step": 85000
1029
+ },
1030
+ {
1031
+ "epoch": 1.76,
1032
+ "learning_rate": 2.0654255647768694e-05,
1033
+ "loss": 1.2778,
1034
+ "step": 85500
1035
+ },
1036
+ {
1037
+ "epoch": 1.77,
1038
+ "learning_rate": 2.0482643107697167e-05,
1039
+ "loss": 1.2382,
1040
+ "step": 86000
1041
+ },
1042
+ {
1043
+ "epoch": 1.78,
1044
+ "learning_rate": 2.031103056762564e-05,
1045
+ "loss": 1.2764,
1046
+ "step": 86500
1047
+ },
1048
+ {
1049
+ "epoch": 1.79,
1050
+ "learning_rate": 2.0139418027554112e-05,
1051
+ "loss": 1.2352,
1052
+ "step": 87000
1053
+ },
1054
+ {
1055
+ "epoch": 1.8,
1056
+ "learning_rate": 1.996780548748258e-05,
1057
+ "loss": 1.2464,
1058
+ "step": 87500
1059
+ },
1060
+ {
1061
+ "epoch": 1.81,
1062
+ "learning_rate": 1.9796192947411054e-05,
1063
+ "loss": 1.2257,
1064
+ "step": 88000
1065
+ },
1066
+ {
1067
+ "epoch": 1.82,
1068
+ "learning_rate": 1.9624580407339527e-05,
1069
+ "loss": 1.2331,
1070
+ "step": 88500
1071
+ },
1072
+ {
1073
+ "epoch": 1.83,
1074
+ "learning_rate": 1.9452967867268e-05,
1075
+ "loss": 1.2404,
1076
+ "step": 89000
1077
+ },
1078
+ {
1079
+ "epoch": 1.84,
1080
+ "learning_rate": 1.928135532719647e-05,
1081
+ "loss": 1.2406,
1082
+ "step": 89500
1083
+ },
1084
+ {
1085
+ "epoch": 1.85,
1086
+ "learning_rate": 1.9109742787124942e-05,
1087
+ "loss": 1.2287,
1088
+ "step": 90000
1089
+ },
1090
+ {
1091
+ "epoch": 1.86,
1092
+ "learning_rate": 1.893813024705341e-05,
1093
+ "loss": 1.2386,
1094
+ "step": 90500
1095
+ },
1096
+ {
1097
+ "epoch": 1.87,
1098
+ "learning_rate": 1.8766517706981884e-05,
1099
+ "loss": 1.2105,
1100
+ "step": 91000
1101
+ },
1102
+ {
1103
+ "epoch": 1.88,
1104
+ "learning_rate": 1.8594905166910357e-05,
1105
+ "loss": 1.2354,
1106
+ "step": 91500
1107
+ },
1108
+ {
1109
+ "epoch": 1.89,
1110
+ "learning_rate": 1.842329262683883e-05,
1111
+ "loss": 1.2233,
1112
+ "step": 92000
1113
+ },
1114
+ {
1115
+ "epoch": 1.9,
1116
+ "learning_rate": 1.82516800867673e-05,
1117
+ "loss": 1.2254,
1118
+ "step": 92500
1119
+ },
1120
+ {
1121
+ "epoch": 1.92,
1122
+ "learning_rate": 1.8080067546695772e-05,
1123
+ "loss": 1.218,
1124
+ "step": 93000
1125
+ },
1126
+ {
1127
+ "epoch": 1.93,
1128
+ "learning_rate": 1.7908455006624245e-05,
1129
+ "loss": 1.2331,
1130
+ "step": 93500
1131
+ },
1132
+ {
1133
+ "epoch": 1.94,
1134
+ "learning_rate": 1.7736842466552718e-05,
1135
+ "loss": 1.2435,
1136
+ "step": 94000
1137
+ },
1138
+ {
1139
+ "epoch": 1.95,
1140
+ "learning_rate": 1.7565229926481187e-05,
1141
+ "loss": 1.2365,
1142
+ "step": 94500
1143
+ },
1144
+ {
1145
+ "epoch": 1.96,
1146
+ "learning_rate": 1.739361738640966e-05,
1147
+ "loss": 1.2457,
1148
+ "step": 95000
1149
+ },
1150
+ {
1151
+ "epoch": 1.97,
1152
+ "learning_rate": 1.7222004846338133e-05,
1153
+ "loss": 1.2267,
1154
+ "step": 95500
1155
+ },
1156
+ {
1157
+ "epoch": 1.98,
1158
+ "learning_rate": 1.7050392306266605e-05,
1159
+ "loss": 1.2258,
1160
+ "step": 96000
1161
+ },
1162
+ {
1163
+ "epoch": 1.99,
1164
+ "learning_rate": 1.6878779766195078e-05,
1165
+ "loss": 1.2074,
1166
+ "step": 96500
1167
+ },
1168
+ {
1169
+ "epoch": 2.0,
1170
+ "learning_rate": 1.6707167226123548e-05,
1171
+ "loss": 1.2509,
1172
+ "step": 97000
1173
+ },
1174
+ {
1175
+ "epoch": 2.01,
1176
+ "learning_rate": 1.653555468605202e-05,
1177
+ "loss": 1.1091,
1178
+ "step": 97500
1179
+ },
1180
+ {
1181
+ "epoch": 2.02,
1182
+ "learning_rate": 1.6363942145980493e-05,
1183
+ "loss": 1.0773,
1184
+ "step": 98000
1185
+ },
1186
+ {
1187
+ "epoch": 2.03,
1188
+ "learning_rate": 1.6192329605908963e-05,
1189
+ "loss": 1.0776,
1190
+ "step": 98500
1191
+ },
1192
+ {
1193
+ "epoch": 2.04,
1194
+ "learning_rate": 1.6020717065837435e-05,
1195
+ "loss": 1.0561,
1196
+ "step": 99000
1197
+ },
1198
+ {
1199
+ "epoch": 2.05,
1200
+ "learning_rate": 1.5849104525765908e-05,
1201
+ "loss": 1.0627,
1202
+ "step": 99500
1203
+ },
1204
+ {
1205
+ "epoch": 2.06,
1206
+ "learning_rate": 1.5677491985694378e-05,
1207
+ "loss": 1.0586,
1208
+ "step": 100000
1209
+ },
1210
+ {
1211
+ "epoch": 2.07,
1212
+ "learning_rate": 1.550587944562285e-05,
1213
+ "loss": 1.0628,
1214
+ "step": 100500
1215
+ },
1216
+ {
1217
+ "epoch": 2.08,
1218
+ "learning_rate": 1.5334266905551323e-05,
1219
+ "loss": 1.0768,
1220
+ "step": 101000
1221
+ },
1222
+ {
1223
+ "epoch": 2.09,
1224
+ "learning_rate": 1.5162654365479794e-05,
1225
+ "loss": 1.0788,
1226
+ "step": 101500
1227
+ },
1228
+ {
1229
+ "epoch": 2.1,
1230
+ "learning_rate": 1.4991041825408267e-05,
1231
+ "loss": 1.0629,
1232
+ "step": 102000
1233
+ },
1234
+ {
1235
+ "epoch": 2.11,
1236
+ "learning_rate": 1.4819429285336738e-05,
1237
+ "loss": 1.0641,
1238
+ "step": 102500
1239
+ },
1240
+ {
1241
+ "epoch": 2.12,
1242
+ "learning_rate": 1.4647816745265211e-05,
1243
+ "loss": 1.0654,
1244
+ "step": 103000
1245
+ },
1246
+ {
1247
+ "epoch": 2.13,
1248
+ "learning_rate": 1.4476204205193684e-05,
1249
+ "loss": 1.0584,
1250
+ "step": 103500
1251
+ },
1252
+ {
1253
+ "epoch": 2.14,
1254
+ "learning_rate": 1.4304591665122155e-05,
1255
+ "loss": 1.0606,
1256
+ "step": 104000
1257
+ },
1258
+ {
1259
+ "epoch": 2.15,
1260
+ "learning_rate": 1.4132979125050628e-05,
1261
+ "loss": 1.0814,
1262
+ "step": 104500
1263
+ },
1264
+ {
1265
+ "epoch": 2.16,
1266
+ "learning_rate": 1.3961366584979099e-05,
1267
+ "loss": 1.0753,
1268
+ "step": 105000
1269
+ },
1270
+ {
1271
+ "epoch": 2.17,
1272
+ "learning_rate": 1.3789754044907571e-05,
1273
+ "loss": 1.0789,
1274
+ "step": 105500
1275
+ },
1276
+ {
1277
+ "epoch": 2.18,
1278
+ "learning_rate": 1.3618141504836043e-05,
1279
+ "loss": 1.0673,
1280
+ "step": 106000
1281
+ },
1282
+ {
1283
+ "epoch": 2.19,
1284
+ "learning_rate": 1.3446528964764512e-05,
1285
+ "loss": 1.0591,
1286
+ "step": 106500
1287
+ },
1288
+ {
1289
+ "epoch": 2.2,
1290
+ "learning_rate": 1.3274916424692985e-05,
1291
+ "loss": 1.0724,
1292
+ "step": 107000
1293
+ },
1294
+ {
1295
+ "epoch": 2.21,
1296
+ "learning_rate": 1.3103303884621458e-05,
1297
+ "loss": 1.0467,
1298
+ "step": 107500
1299
+ },
1300
+ {
1301
+ "epoch": 2.22,
1302
+ "learning_rate": 1.2931691344549929e-05,
1303
+ "loss": 1.048,
1304
+ "step": 108000
1305
+ },
1306
+ {
1307
+ "epoch": 2.23,
1308
+ "learning_rate": 1.2760078804478401e-05,
1309
+ "loss": 1.0669,
1310
+ "step": 108500
1311
+ },
1312
+ {
1313
+ "epoch": 2.24,
1314
+ "learning_rate": 1.2588466264406873e-05,
1315
+ "loss": 1.0526,
1316
+ "step": 109000
1317
+ },
1318
+ {
1319
+ "epoch": 2.25,
1320
+ "learning_rate": 1.2416853724335345e-05,
1321
+ "loss": 1.0604,
1322
+ "step": 109500
1323
+ },
1324
+ {
1325
+ "epoch": 2.27,
1326
+ "learning_rate": 1.2245241184263816e-05,
1327
+ "loss": 1.0689,
1328
+ "step": 110000
1329
+ },
1330
+ {
1331
+ "epoch": 2.28,
1332
+ "learning_rate": 1.207362864419229e-05,
1333
+ "loss": 1.0626,
1334
+ "step": 110500
1335
+ },
1336
+ {
1337
+ "epoch": 2.29,
1338
+ "learning_rate": 1.1902016104120762e-05,
1339
+ "loss": 1.0637,
1340
+ "step": 111000
1341
+ },
1342
+ {
1343
+ "epoch": 2.3,
1344
+ "learning_rate": 1.1730403564049233e-05,
1345
+ "loss": 1.0511,
1346
+ "step": 111500
1347
+ },
1348
+ {
1349
+ "epoch": 2.31,
1350
+ "learning_rate": 1.1558791023977706e-05,
1351
+ "loss": 1.0449,
1352
+ "step": 112000
1353
+ },
1354
+ {
1355
+ "epoch": 2.32,
1356
+ "learning_rate": 1.1387178483906175e-05,
1357
+ "loss": 1.0675,
1358
+ "step": 112500
1359
+ },
1360
+ {
1361
+ "epoch": 2.33,
1362
+ "learning_rate": 1.1215565943834648e-05,
1363
+ "loss": 1.0524,
1364
+ "step": 113000
1365
+ },
1366
+ {
1367
+ "epoch": 2.34,
1368
+ "learning_rate": 1.1043953403763121e-05,
1369
+ "loss": 1.0554,
1370
+ "step": 113500
1371
+ },
1372
+ {
1373
+ "epoch": 2.35,
1374
+ "learning_rate": 1.0872340863691592e-05,
1375
+ "loss": 1.0476,
1376
+ "step": 114000
1377
+ },
1378
+ {
1379
+ "epoch": 2.36,
1380
+ "learning_rate": 1.0700728323620065e-05,
1381
+ "loss": 1.0516,
1382
+ "step": 114500
1383
+ },
1384
+ {
1385
+ "epoch": 2.37,
1386
+ "learning_rate": 1.0529115783548536e-05,
1387
+ "loss": 1.0472,
1388
+ "step": 115000
1389
+ },
1390
+ {
1391
+ "epoch": 2.38,
1392
+ "learning_rate": 1.0357503243477009e-05,
1393
+ "loss": 1.0581,
1394
+ "step": 115500
1395
+ },
1396
+ {
1397
+ "epoch": 2.39,
1398
+ "learning_rate": 1.018589070340548e-05,
1399
+ "loss": 1.0515,
1400
+ "step": 116000
1401
+ },
1402
+ {
1403
+ "epoch": 2.4,
1404
+ "learning_rate": 1.0014278163333951e-05,
1405
+ "loss": 1.0549,
1406
+ "step": 116500
1407
+ },
1408
+ {
1409
+ "epoch": 2.41,
1410
+ "learning_rate": 9.842665623262424e-06,
1411
+ "loss": 1.0411,
1412
+ "step": 117000
1413
+ },
1414
+ {
1415
+ "epoch": 2.42,
1416
+ "learning_rate": 9.671053083190895e-06,
1417
+ "loss": 1.0534,
1418
+ "step": 117500
1419
+ },
1420
+ {
1421
+ "epoch": 2.43,
1422
+ "learning_rate": 9.499440543119368e-06,
1423
+ "loss": 1.0509,
1424
+ "step": 118000
1425
+ },
1426
+ {
1427
+ "epoch": 2.44,
1428
+ "learning_rate": 9.327828003047839e-06,
1429
+ "loss": 1.0508,
1430
+ "step": 118500
1431
+ },
1432
+ {
1433
+ "epoch": 2.45,
1434
+ "learning_rate": 9.156215462976311e-06,
1435
+ "loss": 1.0392,
1436
+ "step": 119000
1437
+ },
1438
+ {
1439
+ "epoch": 2.46,
1440
+ "learning_rate": 8.984602922904784e-06,
1441
+ "loss": 1.0254,
1442
+ "step": 119500
1443
+ },
1444
+ {
1445
+ "epoch": 2.47,
1446
+ "learning_rate": 8.812990382833254e-06,
1447
+ "loss": 1.0562,
1448
+ "step": 120000
1449
+ },
1450
+ {
1451
+ "epoch": 2.48,
1452
+ "learning_rate": 8.641377842761726e-06,
1453
+ "loss": 1.0678,
1454
+ "step": 120500
1455
+ },
1456
+ {
1457
+ "epoch": 2.49,
1458
+ "learning_rate": 8.469765302690198e-06,
1459
+ "loss": 1.0201,
1460
+ "step": 121000
1461
+ },
1462
+ {
1463
+ "epoch": 2.5,
1464
+ "learning_rate": 8.29815276261867e-06,
1465
+ "loss": 1.0377,
1466
+ "step": 121500
1467
+ },
1468
+ {
1469
+ "epoch": 2.51,
1470
+ "learning_rate": 8.126540222547143e-06,
1471
+ "loss": 1.056,
1472
+ "step": 122000
1473
+ },
1474
+ {
1475
+ "epoch": 2.52,
1476
+ "learning_rate": 7.954927682475614e-06,
1477
+ "loss": 1.0411,
1478
+ "step": 122500
1479
+ },
1480
+ {
1481
+ "epoch": 2.53,
1482
+ "learning_rate": 7.783315142404087e-06,
1483
+ "loss": 1.0324,
1484
+ "step": 123000
1485
+ },
1486
+ {
1487
+ "epoch": 2.54,
1488
+ "learning_rate": 7.611702602332559e-06,
1489
+ "loss": 1.0235,
1490
+ "step": 123500
1491
+ },
1492
+ {
1493
+ "epoch": 2.55,
1494
+ "learning_rate": 7.440090062261029e-06,
1495
+ "loss": 1.0511,
1496
+ "step": 124000
1497
+ },
1498
+ {
1499
+ "epoch": 2.56,
1500
+ "learning_rate": 7.268477522189501e-06,
1501
+ "loss": 1.0382,
1502
+ "step": 124500
1503
+ },
1504
+ {
1505
+ "epoch": 2.57,
1506
+ "learning_rate": 7.096864982117973e-06,
1507
+ "loss": 1.0166,
1508
+ "step": 125000
1509
+ },
1510
+ {
1511
+ "epoch": 2.58,
1512
+ "learning_rate": 6.925252442046445e-06,
1513
+ "loss": 1.0437,
1514
+ "step": 125500
1515
+ },
1516
+ {
1517
+ "epoch": 2.59,
1518
+ "learning_rate": 6.753639901974918e-06,
1519
+ "loss": 1.0575,
1520
+ "step": 126000
1521
+ },
1522
+ {
1523
+ "epoch": 2.61,
1524
+ "learning_rate": 6.58202736190339e-06,
1525
+ "loss": 1.0355,
1526
+ "step": 126500
1527
+ },
1528
+ {
1529
+ "epoch": 2.62,
1530
+ "learning_rate": 6.410414821831862e-06,
1531
+ "loss": 1.0414,
1532
+ "step": 127000
1533
+ },
1534
+ {
1535
+ "epoch": 2.63,
1536
+ "learning_rate": 6.238802281760333e-06,
1537
+ "loss": 1.0213,
1538
+ "step": 127500
1539
+ },
1540
+ {
1541
+ "epoch": 2.64,
1542
+ "learning_rate": 6.067189741688805e-06,
1543
+ "loss": 1.0432,
1544
+ "step": 128000
1545
+ },
1546
+ {
1547
+ "epoch": 2.65,
1548
+ "learning_rate": 5.895577201617277e-06,
1549
+ "loss": 1.0309,
1550
+ "step": 128500
1551
+ },
1552
+ {
1553
+ "epoch": 2.66,
1554
+ "learning_rate": 5.723964661545749e-06,
1555
+ "loss": 1.0199,
1556
+ "step": 129000
1557
+ },
1558
+ {
1559
+ "epoch": 2.67,
1560
+ "learning_rate": 5.5523521214742206e-06,
1561
+ "loss": 1.0388,
1562
+ "step": 129500
1563
+ },
1564
+ {
1565
+ "epoch": 2.68,
1566
+ "learning_rate": 5.3807395814026925e-06,
1567
+ "loss": 1.0346,
1568
+ "step": 130000
1569
+ },
1570
+ {
1571
+ "epoch": 2.69,
1572
+ "learning_rate": 5.2091270413311644e-06,
1573
+ "loss": 1.0467,
1574
+ "step": 130500
1575
+ },
1576
+ {
1577
+ "epoch": 2.7,
1578
+ "learning_rate": 5.037514501259636e-06,
1579
+ "loss": 1.035,
1580
+ "step": 131000
1581
+ },
1582
+ {
1583
+ "epoch": 2.71,
1584
+ "learning_rate": 4.865901961188108e-06,
1585
+ "loss": 1.0314,
1586
+ "step": 131500
1587
+ },
1588
+ {
1589
+ "epoch": 2.72,
1590
+ "learning_rate": 4.69428942111658e-06,
1591
+ "loss": 1.0416,
1592
+ "step": 132000
1593
+ },
1594
+ {
1595
+ "epoch": 2.73,
1596
+ "learning_rate": 4.522676881045052e-06,
1597
+ "loss": 1.0085,
1598
+ "step": 132500
1599
+ },
1600
+ {
1601
+ "epoch": 2.74,
1602
+ "learning_rate": 4.351064340973523e-06,
1603
+ "loss": 1.043,
1604
+ "step": 133000
1605
+ },
1606
+ {
1607
+ "epoch": 2.75,
1608
+ "learning_rate": 4.179451800901995e-06,
1609
+ "loss": 1.0191,
1610
+ "step": 133500
1611
+ },
1612
+ {
1613
+ "epoch": 2.76,
1614
+ "learning_rate": 4.007839260830468e-06,
1615
+ "loss": 1.0402,
1616
+ "step": 134000
1617
+ },
1618
+ {
1619
+ "epoch": 2.77,
1620
+ "learning_rate": 3.83622672075894e-06,
1621
+ "loss": 1.0454,
1622
+ "step": 134500
1623
+ },
1624
+ {
1625
+ "epoch": 2.78,
1626
+ "learning_rate": 3.664614180687411e-06,
1627
+ "loss": 1.019,
1628
+ "step": 135000
1629
+ },
1630
+ {
1631
+ "epoch": 2.79,
1632
+ "learning_rate": 3.493001640615883e-06,
1633
+ "loss": 1.0326,
1634
+ "step": 135500
1635
+ },
1636
+ {
1637
+ "epoch": 2.8,
1638
+ "learning_rate": 3.3213891005443554e-06,
1639
+ "loss": 1.033,
1640
+ "step": 136000
1641
+ },
1642
+ {
1643
+ "epoch": 2.81,
1644
+ "learning_rate": 3.1497765604728273e-06,
1645
+ "loss": 1.0224,
1646
+ "step": 136500
1647
+ },
1648
+ {
1649
+ "epoch": 2.82,
1650
+ "learning_rate": 2.978164020401299e-06,
1651
+ "loss": 1.0207,
1652
+ "step": 137000
1653
+ },
1654
+ {
1655
+ "epoch": 2.83,
1656
+ "learning_rate": 2.806551480329771e-06,
1657
+ "loss": 1.0219,
1658
+ "step": 137500
1659
+ },
1660
+ {
1661
+ "epoch": 2.84,
1662
+ "learning_rate": 2.6349389402582427e-06,
1663
+ "loss": 1.0244,
1664
+ "step": 138000
1665
+ },
1666
+ {
1667
+ "epoch": 2.85,
1668
+ "learning_rate": 2.4633264001867147e-06,
1669
+ "loss": 1.0157,
1670
+ "step": 138500
1671
+ },
1672
+ {
1673
+ "epoch": 2.86,
1674
+ "learning_rate": 2.2917138601151866e-06,
1675
+ "loss": 1.008,
1676
+ "step": 139000
1677
+ },
1678
+ {
1679
+ "epoch": 2.87,
1680
+ "learning_rate": 2.120101320043658e-06,
1681
+ "loss": 1.0381,
1682
+ "step": 139500
1683
+ },
1684
+ {
1685
+ "epoch": 2.88,
1686
+ "learning_rate": 1.9484887799721305e-06,
1687
+ "loss": 1.0198,
1688
+ "step": 140000
1689
+ },
1690
+ {
1691
+ "epoch": 2.89,
1692
+ "learning_rate": 1.776876239900602e-06,
1693
+ "loss": 1.0076,
1694
+ "step": 140500
1695
+ },
1696
+ {
1697
+ "epoch": 2.9,
1698
+ "learning_rate": 1.6052636998290742e-06,
1699
+ "loss": 1.0175,
1700
+ "step": 141000
1701
+ },
1702
+ {
1703
+ "epoch": 2.91,
1704
+ "learning_rate": 1.433651159757546e-06,
1705
+ "loss": 1.0199,
1706
+ "step": 141500
1707
+ },
1708
+ {
1709
+ "epoch": 2.92,
1710
+ "learning_rate": 1.2620386196860177e-06,
1711
+ "loss": 1.026,
1712
+ "step": 142000
1713
+ },
1714
+ {
1715
+ "epoch": 2.93,
1716
+ "learning_rate": 1.0904260796144896e-06,
1717
+ "loss": 1.0359,
1718
+ "step": 142500
1719
+ },
1720
+ {
1721
+ "epoch": 2.94,
1722
+ "learning_rate": 9.188135395429614e-07,
1723
+ "loss": 1.0409,
1724
+ "step": 143000
1725
+ },
1726
+ {
1727
+ "epoch": 2.96,
1728
+ "learning_rate": 7.472009994714334e-07,
1729
+ "loss": 1.0028,
1730
+ "step": 143500
1731
+ },
1732
+ {
1733
+ "epoch": 2.97,
1734
+ "learning_rate": 5.755884593999052e-07,
1735
+ "loss": 1.0147,
1736
+ "step": 144000
1737
+ },
1738
+ {
1739
+ "epoch": 2.98,
1740
+ "learning_rate": 4.0397591932837716e-07,
1741
+ "loss": 1.0111,
1742
+ "step": 144500
1743
+ },
1744
+ {
1745
+ "epoch": 2.99,
1746
+ "learning_rate": 2.3236337925684908e-07,
1747
+ "loss": 1.0241,
1748
+ "step": 145000
1749
+ },
1750
+ {
1751
+ "epoch": 3.0,
1752
+ "learning_rate": 6.075083918532095e-08,
1753
+ "loss": 1.0178,
1754
+ "step": 145500
1755
+ },
1756
+ {
1757
+ "epoch": 3.0,
1758
+ "step": 145677,
1759
+ "total_flos": 2.385792939319296e+17,
1760
+ "train_loss": 1.349638056625609,
1761
+ "train_runtime": 40215.8311,
1762
+ "train_samples_per_second": 36.223,
1763
+ "train_steps_per_second": 3.622
1764
+ }
1765
+ ],
1766
+ "max_steps": 145677,
1767
+ "num_train_epochs": 3,
1768
+ "total_flos": 2.385792939319296e+17,
1769
+ "trial_name": null,
1770
+ "trial_params": null
1771
+ }