xxang commited on
Commit
3811f5a
·
verified ·
1 Parent(s): ca584b2
Files changed (1) hide show
  1. trainer_state.json +357 -0
trainer_state.json ADDED
@@ -0,0 +1,357 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.832,
5
+ "eval_steps": 500,
6
+ "global_step": 45,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.064,
13
+ "grad_norm": 2.67690372467041,
14
+ "learning_rate": 2.0000000000000003e-06,
15
+ "loss": 0.8906,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.128,
20
+ "grad_norm": 2.5836398601531982,
21
+ "learning_rate": 4.000000000000001e-06,
22
+ "loss": 0.8844,
23
+ "step": 2
24
+ },
25
+ {
26
+ "epoch": 0.192,
27
+ "grad_norm": 2.692220687866211,
28
+ "learning_rate": 6e-06,
29
+ "loss": 0.7934,
30
+ "step": 3
31
+ },
32
+ {
33
+ "epoch": 0.256,
34
+ "grad_norm": 2.424602508544922,
35
+ "learning_rate": 8.000000000000001e-06,
36
+ "loss": 0.8946,
37
+ "step": 4
38
+ },
39
+ {
40
+ "epoch": 0.32,
41
+ "grad_norm": 1.7447768449783325,
42
+ "learning_rate": 1e-05,
43
+ "loss": 0.7398,
44
+ "step": 5
45
+ },
46
+ {
47
+ "epoch": 0.384,
48
+ "grad_norm": 0.985548734664917,
49
+ "learning_rate": 9.984586668665641e-06,
50
+ "loss": 0.7794,
51
+ "step": 6
52
+ },
53
+ {
54
+ "epoch": 0.448,
55
+ "grad_norm": 0.9407026171684265,
56
+ "learning_rate": 9.938441702975689e-06,
57
+ "loss": 0.7591,
58
+ "step": 7
59
+ },
60
+ {
61
+ "epoch": 0.512,
62
+ "grad_norm": 1.5231335163116455,
63
+ "learning_rate": 9.861849601988384e-06,
64
+ "loss": 0.7895,
65
+ "step": 8
66
+ },
67
+ {
68
+ "epoch": 0.576,
69
+ "grad_norm": 1.7350199222564697,
70
+ "learning_rate": 9.755282581475769e-06,
71
+ "loss": 0.8721,
72
+ "step": 9
73
+ },
74
+ {
75
+ "epoch": 0.64,
76
+ "grad_norm": 1.5788978338241577,
77
+ "learning_rate": 9.619397662556434e-06,
78
+ "loss": 0.8114,
79
+ "step": 10
80
+ },
81
+ {
82
+ "epoch": 0.704,
83
+ "grad_norm": 1.2389469146728516,
84
+ "learning_rate": 9.45503262094184e-06,
85
+ "loss": 0.7008,
86
+ "step": 11
87
+ },
88
+ {
89
+ "epoch": 0.768,
90
+ "grad_norm": 1.423549771308899,
91
+ "learning_rate": 9.263200821770462e-06,
92
+ "loss": 0.7585,
93
+ "step": 12
94
+ },
95
+ {
96
+ "epoch": 0.832,
97
+ "grad_norm": 1.1815698146820068,
98
+ "learning_rate": 9.045084971874738e-06,
99
+ "loss": 0.6632,
100
+ "step": 13
101
+ },
102
+ {
103
+ "epoch": 0.896,
104
+ "grad_norm": 1.1034165620803833,
105
+ "learning_rate": 8.802029828000157e-06,
106
+ "loss": 0.7416,
107
+ "step": 14
108
+ },
109
+ {
110
+ "epoch": 0.96,
111
+ "grad_norm": 0.8475093245506287,
112
+ "learning_rate": 8.535533905932739e-06,
113
+ "loss": 0.7517,
114
+ "step": 15
115
+ },
116
+ {
117
+ "epoch": 1.0,
118
+ "grad_norm": 0.8475093245506287,
119
+ "learning_rate": 8.247240241650918e-06,
120
+ "loss": 0.7802,
121
+ "step": 16
122
+ },
123
+ {
124
+ "epoch": 1.064,
125
+ "grad_norm": 1.1045087575912476,
126
+ "learning_rate": 7.938926261462366e-06,
127
+ "loss": 0.6967,
128
+ "step": 17
129
+ },
130
+ {
131
+ "epoch": 1.1280000000000001,
132
+ "grad_norm": 0.6255977153778076,
133
+ "learning_rate": 7.612492823579744e-06,
134
+ "loss": 0.6597,
135
+ "step": 18
136
+ },
137
+ {
138
+ "epoch": 1.192,
139
+ "grad_norm": 0.7395212054252625,
140
+ "learning_rate": 7.269952498697734e-06,
141
+ "loss": 0.6748,
142
+ "step": 19
143
+ },
144
+ {
145
+ "epoch": 1.256,
146
+ "grad_norm": 0.6993595957756042,
147
+ "learning_rate": 6.913417161825449e-06,
148
+ "loss": 0.7681,
149
+ "step": 20
150
+ },
151
+ {
152
+ "epoch": 1.32,
153
+ "grad_norm": 0.7346011996269226,
154
+ "learning_rate": 6.545084971874738e-06,
155
+ "loss": 0.7262,
156
+ "step": 21
157
+ },
158
+ {
159
+ "epoch": 1.384,
160
+ "grad_norm": 0.6753884553909302,
161
+ "learning_rate": 6.1672268192795285e-06,
162
+ "loss": 0.6249,
163
+ "step": 22
164
+ },
165
+ {
166
+ "epoch": 1.448,
167
+ "grad_norm": 0.7035785913467407,
168
+ "learning_rate": 5.782172325201155e-06,
169
+ "loss": 0.6907,
170
+ "step": 23
171
+ },
172
+ {
173
+ "epoch": 1.512,
174
+ "grad_norm": 0.6587271094322205,
175
+ "learning_rate": 5.392295478639226e-06,
176
+ "loss": 0.6754,
177
+ "step": 24
178
+ },
179
+ {
180
+ "epoch": 1.576,
181
+ "grad_norm": 0.5424773097038269,
182
+ "learning_rate": 5e-06,
183
+ "loss": 0.6626,
184
+ "step": 25
185
+ },
186
+ {
187
+ "epoch": 1.6400000000000001,
188
+ "grad_norm": 0.5312824845314026,
189
+ "learning_rate": 4.6077045213607765e-06,
190
+ "loss": 0.7369,
191
+ "step": 26
192
+ },
193
+ {
194
+ "epoch": 1.704,
195
+ "grad_norm": 0.5092198848724365,
196
+ "learning_rate": 4.217827674798845e-06,
197
+ "loss": 0.766,
198
+ "step": 27
199
+ },
200
+ {
201
+ "epoch": 1.768,
202
+ "grad_norm": 0.49842214584350586,
203
+ "learning_rate": 3.832773180720475e-06,
204
+ "loss": 0.7003,
205
+ "step": 28
206
+ },
207
+ {
208
+ "epoch": 1.8319999999999999,
209
+ "grad_norm": 0.5175044536590576,
210
+ "learning_rate": 3.4549150281252635e-06,
211
+ "loss": 0.7393,
212
+ "step": 29
213
+ },
214
+ {
215
+ "epoch": 1.896,
216
+ "grad_norm": 0.48302003741264343,
217
+ "learning_rate": 3.0865828381745515e-06,
218
+ "loss": 0.7018,
219
+ "step": 30
220
+ },
221
+ {
222
+ "epoch": 1.96,
223
+ "grad_norm": 0.4566391110420227,
224
+ "learning_rate": 2.7300475013022666e-06,
225
+ "loss": 0.6598,
226
+ "step": 31
227
+ },
228
+ {
229
+ "epoch": 2.0,
230
+ "grad_norm": 0.6593987941741943,
231
+ "learning_rate": 2.387507176420256e-06,
232
+ "loss": 0.651,
233
+ "step": 32
234
+ },
235
+ {
236
+ "epoch": 2.064,
237
+ "grad_norm": 0.6568493843078613,
238
+ "learning_rate": 2.061073738537635e-06,
239
+ "loss": 0.7002,
240
+ "step": 33
241
+ },
242
+ {
243
+ "epoch": 2.128,
244
+ "grad_norm": 0.4518449306488037,
245
+ "learning_rate": 1.7527597583490825e-06,
246
+ "loss": 0.7008,
247
+ "step": 34
248
+ },
249
+ {
250
+ "epoch": 2.192,
251
+ "grad_norm": 0.4067769944667816,
252
+ "learning_rate": 1.4644660940672628e-06,
253
+ "loss": 0.6299,
254
+ "step": 35
255
+ },
256
+ {
257
+ "epoch": 2.2560000000000002,
258
+ "grad_norm": 0.41410645842552185,
259
+ "learning_rate": 1.1979701719998454e-06,
260
+ "loss": 0.6832,
261
+ "step": 36
262
+ },
263
+ {
264
+ "epoch": 2.32,
265
+ "grad_norm": 0.33175525069236755,
266
+ "learning_rate": 9.549150281252633e-07,
267
+ "loss": 0.5874,
268
+ "step": 37
269
+ },
270
+ {
271
+ "epoch": 2.384,
272
+ "grad_norm": 0.4822118580341339,
273
+ "learning_rate": 7.367991782295392e-07,
274
+ "loss": 0.6533,
275
+ "step": 38
276
+ },
277
+ {
278
+ "epoch": 2.448,
279
+ "grad_norm": 0.4443027079105377,
280
+ "learning_rate": 5.449673790581611e-07,
281
+ "loss": 0.7716,
282
+ "step": 39
283
+ },
284
+ {
285
+ "epoch": 2.512,
286
+ "grad_norm": 0.38352447748184204,
287
+ "learning_rate": 3.8060233744356634e-07,
288
+ "loss": 0.7014,
289
+ "step": 40
290
+ },
291
+ {
292
+ "epoch": 2.576,
293
+ "grad_norm": 0.39618387818336487,
294
+ "learning_rate": 2.447174185242324e-07,
295
+ "loss": 0.5842,
296
+ "step": 41
297
+ },
298
+ {
299
+ "epoch": 2.64,
300
+ "grad_norm": 0.3571045398712158,
301
+ "learning_rate": 1.3815039801161723e-07,
302
+ "loss": 0.6552,
303
+ "step": 42
304
+ },
305
+ {
306
+ "epoch": 2.7039999999999997,
307
+ "grad_norm": 0.44936269521713257,
308
+ "learning_rate": 6.15582970243117e-08,
309
+ "loss": 0.6133,
310
+ "step": 43
311
+ },
312
+ {
313
+ "epoch": 2.768,
314
+ "grad_norm": 0.40641671419143677,
315
+ "learning_rate": 1.541333133436018e-08,
316
+ "loss": 0.6315,
317
+ "step": 44
318
+ },
319
+ {
320
+ "epoch": 2.832,
321
+ "grad_norm": 0.4280812740325928,
322
+ "learning_rate": 0.0,
323
+ "loss": 0.6455,
324
+ "step": 45
325
+ },
326
+ {
327
+ "epoch": 2.832,
328
+ "step": 45,
329
+ "total_flos": 75420674818048.0,
330
+ "train_loss": 0.7178263836436801,
331
+ "train_runtime": 10468.2018,
332
+ "train_samples_per_second": 0.287,
333
+ "train_steps_per_second": 0.004
334
+ }
335
+ ],
336
+ "logging_steps": 1,
337
+ "max_steps": 45,
338
+ "num_input_tokens_seen": 0,
339
+ "num_train_epochs": 3,
340
+ "save_steps": 500,
341
+ "stateful_callbacks": {
342
+ "TrainerControl": {
343
+ "args": {
344
+ "should_epoch_stop": false,
345
+ "should_evaluate": false,
346
+ "should_log": false,
347
+ "should_save": true,
348
+ "should_training_stop": true
349
+ },
350
+ "attributes": {}
351
+ }
352
+ },
353
+ "total_flos": 75420674818048.0,
354
+ "train_batch_size": 1,
355
+ "trial_name": null,
356
+ "trial_params": null
357
+ }