ambrosfitz commited on
Commit
437a1d2
·
verified ·
1 Parent(s): e0b1d50

Upload 3 files

Browse files
events.out.tfevents.1728228710.c51c281a98ce.2055.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:782e98157db2efa561b1eacff6f1f9d70c8e0ebafc4c2eac592e8476242cf765
3
+ size 18372
events.out.tfevents.1728233913.c51c281a98ce.2055.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d7824addfc89ce5164e956f24f899b329793a5027481996a7ce134e1553cffec
3
+ size 359
trainer_state.json ADDED
@@ -0,0 +1,441 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.8340622782707214,
3
+ "best_model_checkpoint": "./results/checkpoint-1000",
4
+ "epoch": 2.9940436796823295,
5
+ "eval_steps": 500,
6
+ "global_step": 1131,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.05294506949040371,
13
+ "grad_norm": 62.576904296875,
14
+ "learning_rate": 2.0000000000000003e-06,
15
+ "loss": 16.7099,
16
+ "step": 20
17
+ },
18
+ {
19
+ "epoch": 0.10589013898080742,
20
+ "grad_norm": 35.65088653564453,
21
+ "learning_rate": 4.000000000000001e-06,
22
+ "loss": 14.7891,
23
+ "step": 40
24
+ },
25
+ {
26
+ "epoch": 0.15883520847121113,
27
+ "grad_norm": 30.480260848999023,
28
+ "learning_rate": 6e-06,
29
+ "loss": 11.6833,
30
+ "step": 60
31
+ },
32
+ {
33
+ "epoch": 0.21178027796161483,
34
+ "grad_norm": 24.958763122558594,
35
+ "learning_rate": 8.000000000000001e-06,
36
+ "loss": 7.878,
37
+ "step": 80
38
+ },
39
+ {
40
+ "epoch": 0.26472534745201853,
41
+ "grad_norm": 4.896631717681885,
42
+ "learning_rate": 1e-05,
43
+ "loss": 4.1299,
44
+ "step": 100
45
+ },
46
+ {
47
+ "epoch": 0.31767041694242226,
48
+ "grad_norm": 1.8821851015090942,
49
+ "learning_rate": 1.2e-05,
50
+ "loss": 2.512,
51
+ "step": 120
52
+ },
53
+ {
54
+ "epoch": 0.37061548643282594,
55
+ "grad_norm": 0.7400406002998352,
56
+ "learning_rate": 1.4000000000000001e-05,
57
+ "loss": 1.8807,
58
+ "step": 140
59
+ },
60
+ {
61
+ "epoch": 0.42356055592322966,
62
+ "grad_norm": 0.601634681224823,
63
+ "learning_rate": 1.6000000000000003e-05,
64
+ "loss": 1.5617,
65
+ "step": 160
66
+ },
67
+ {
68
+ "epoch": 0.47650562541363334,
69
+ "grad_norm": 0.5250083208084106,
70
+ "learning_rate": 1.8e-05,
71
+ "loss": 1.461,
72
+ "step": 180
73
+ },
74
+ {
75
+ "epoch": 0.5294506949040371,
76
+ "grad_norm": 0.5037821531295776,
77
+ "learning_rate": 2e-05,
78
+ "loss": 1.3703,
79
+ "step": 200
80
+ },
81
+ {
82
+ "epoch": 0.5823957643944407,
83
+ "grad_norm": 0.47873854637145996,
84
+ "learning_rate": 2.2000000000000003e-05,
85
+ "loss": 1.2966,
86
+ "step": 220
87
+ },
88
+ {
89
+ "epoch": 0.6353408338848445,
90
+ "grad_norm": 3.9664485454559326,
91
+ "learning_rate": 2.4e-05,
92
+ "loss": 1.2427,
93
+ "step": 240
94
+ },
95
+ {
96
+ "epoch": 0.6882859033752482,
97
+ "grad_norm": 0.45016345381736755,
98
+ "learning_rate": 2.6000000000000002e-05,
99
+ "loss": 1.2139,
100
+ "step": 260
101
+ },
102
+ {
103
+ "epoch": 0.7412309728656519,
104
+ "grad_norm": 0.5136398077011108,
105
+ "learning_rate": 2.8000000000000003e-05,
106
+ "loss": 1.1679,
107
+ "step": 280
108
+ },
109
+ {
110
+ "epoch": 0.7941760423560555,
111
+ "grad_norm": 0.3736862242221832,
112
+ "learning_rate": 3e-05,
113
+ "loss": 1.1604,
114
+ "step": 300
115
+ },
116
+ {
117
+ "epoch": 0.8471211118464593,
118
+ "grad_norm": 0.38698282837867737,
119
+ "learning_rate": 3.2000000000000005e-05,
120
+ "loss": 1.1407,
121
+ "step": 320
122
+ },
123
+ {
124
+ "epoch": 0.900066181336863,
125
+ "grad_norm": 0.4257580637931824,
126
+ "learning_rate": 3.4000000000000007e-05,
127
+ "loss": 1.1177,
128
+ "step": 340
129
+ },
130
+ {
131
+ "epoch": 0.9530112508272667,
132
+ "grad_norm": 0.3982521891593933,
133
+ "learning_rate": 3.6e-05,
134
+ "loss": 1.1199,
135
+ "step": 360
136
+ },
137
+ {
138
+ "epoch": 1.0059563203176705,
139
+ "grad_norm": 0.3849237859249115,
140
+ "learning_rate": 3.8e-05,
141
+ "loss": 1.0925,
142
+ "step": 380
143
+ },
144
+ {
145
+ "epoch": 1.0589013898080741,
146
+ "grad_norm": 0.3753887414932251,
147
+ "learning_rate": 4e-05,
148
+ "loss": 1.0605,
149
+ "step": 400
150
+ },
151
+ {
152
+ "epoch": 1.1118464592984778,
153
+ "grad_norm": 0.3810591697692871,
154
+ "learning_rate": 4.2e-05,
155
+ "loss": 1.0541,
156
+ "step": 420
157
+ },
158
+ {
159
+ "epoch": 1.1647915287888815,
160
+ "grad_norm": 0.3707886040210724,
161
+ "learning_rate": 4.4000000000000006e-05,
162
+ "loss": 1.0334,
163
+ "step": 440
164
+ },
165
+ {
166
+ "epoch": 1.2177365982792852,
167
+ "grad_norm": 0.36902502179145813,
168
+ "learning_rate": 4.600000000000001e-05,
169
+ "loss": 1.0249,
170
+ "step": 460
171
+ },
172
+ {
173
+ "epoch": 1.270681667769689,
174
+ "grad_norm": 0.3862062692642212,
175
+ "learning_rate": 4.8e-05,
176
+ "loss": 1.0141,
177
+ "step": 480
178
+ },
179
+ {
180
+ "epoch": 1.3236267372600927,
181
+ "grad_norm": 0.36468520760536194,
182
+ "learning_rate": 5e-05,
183
+ "loss": 1.063,
184
+ "step": 500
185
+ },
186
+ {
187
+ "epoch": 1.3236267372600927,
188
+ "eval_loss": 0.9045532941818237,
189
+ "eval_runtime": 74.2771,
190
+ "eval_samples_per_second": 10.165,
191
+ "eval_steps_per_second": 2.545,
192
+ "step": 500
193
+ },
194
+ {
195
+ "epoch": 1.3765718067504964,
196
+ "grad_norm": 0.3426459729671478,
197
+ "learning_rate": 4.8415213946117275e-05,
198
+ "loss": 1.01,
199
+ "step": 520
200
+ },
201
+ {
202
+ "epoch": 1.4295168762409,
203
+ "grad_norm": 0.3736313581466675,
204
+ "learning_rate": 4.6830427892234554e-05,
205
+ "loss": 0.984,
206
+ "step": 540
207
+ },
208
+ {
209
+ "epoch": 1.4824619457313037,
210
+ "grad_norm": 0.36285898089408875,
211
+ "learning_rate": 4.524564183835183e-05,
212
+ "loss": 0.9816,
213
+ "step": 560
214
+ },
215
+ {
216
+ "epoch": 1.5354070152217076,
217
+ "grad_norm": 0.37838441133499146,
218
+ "learning_rate": 4.36608557844691e-05,
219
+ "loss": 0.9807,
220
+ "step": 580
221
+ },
222
+ {
223
+ "epoch": 1.588352084712111,
224
+ "grad_norm": 0.3449678421020508,
225
+ "learning_rate": 4.207606973058637e-05,
226
+ "loss": 0.982,
227
+ "step": 600
228
+ },
229
+ {
230
+ "epoch": 1.641297154202515,
231
+ "grad_norm": 0.3467804789543152,
232
+ "learning_rate": 4.0491283676703644e-05,
233
+ "loss": 0.9553,
234
+ "step": 620
235
+ },
236
+ {
237
+ "epoch": 1.6942422236929184,
238
+ "grad_norm": 0.3551880419254303,
239
+ "learning_rate": 3.8906497622820917e-05,
240
+ "loss": 0.9701,
241
+ "step": 640
242
+ },
243
+ {
244
+ "epoch": 1.7471872931833223,
245
+ "grad_norm": 0.3425547182559967,
246
+ "learning_rate": 3.7321711568938196e-05,
247
+ "loss": 0.973,
248
+ "step": 660
249
+ },
250
+ {
251
+ "epoch": 1.800132362673726,
252
+ "grad_norm": 0.32189810276031494,
253
+ "learning_rate": 3.573692551505547e-05,
254
+ "loss": 0.9599,
255
+ "step": 680
256
+ },
257
+ {
258
+ "epoch": 1.8530774321641297,
259
+ "grad_norm": 0.34214696288108826,
260
+ "learning_rate": 3.415213946117274e-05,
261
+ "loss": 0.952,
262
+ "step": 700
263
+ },
264
+ {
265
+ "epoch": 1.9060225016545336,
266
+ "grad_norm": 0.33412784337997437,
267
+ "learning_rate": 3.256735340729002e-05,
268
+ "loss": 0.9453,
269
+ "step": 720
270
+ },
271
+ {
272
+ "epoch": 1.958967571144937,
273
+ "grad_norm": 0.33273905515670776,
274
+ "learning_rate": 3.098256735340729e-05,
275
+ "loss": 0.9387,
276
+ "step": 740
277
+ },
278
+ {
279
+ "epoch": 2.011912640635341,
280
+ "grad_norm": 0.32698702812194824,
281
+ "learning_rate": 2.939778129952457e-05,
282
+ "loss": 0.9456,
283
+ "step": 760
284
+ },
285
+ {
286
+ "epoch": 2.0648577101257444,
287
+ "grad_norm": 0.36428529024124146,
288
+ "learning_rate": 2.7812995245641837e-05,
289
+ "loss": 0.9445,
290
+ "step": 780
291
+ },
292
+ {
293
+ "epoch": 2.1178027796161483,
294
+ "grad_norm": 0.3233266770839691,
295
+ "learning_rate": 2.6228209191759113e-05,
296
+ "loss": 0.9185,
297
+ "step": 800
298
+ },
299
+ {
300
+ "epoch": 2.1707478491065517,
301
+ "grad_norm": 0.3173067569732666,
302
+ "learning_rate": 2.4643423137876386e-05,
303
+ "loss": 0.9146,
304
+ "step": 820
305
+ },
306
+ {
307
+ "epoch": 2.2236929185969556,
308
+ "grad_norm": 0.33917301893234253,
309
+ "learning_rate": 2.305863708399366e-05,
310
+ "loss": 0.9195,
311
+ "step": 840
312
+ },
313
+ {
314
+ "epoch": 2.2766379880873595,
315
+ "grad_norm": 0.3438282608985901,
316
+ "learning_rate": 2.1473851030110938e-05,
317
+ "loss": 0.9356,
318
+ "step": 860
319
+ },
320
+ {
321
+ "epoch": 2.329583057577763,
322
+ "grad_norm": 0.33590319752693176,
323
+ "learning_rate": 1.988906497622821e-05,
324
+ "loss": 0.9319,
325
+ "step": 880
326
+ },
327
+ {
328
+ "epoch": 2.382528127068167,
329
+ "grad_norm": 0.5414553880691528,
330
+ "learning_rate": 1.8304278922345483e-05,
331
+ "loss": 0.9208,
332
+ "step": 900
333
+ },
334
+ {
335
+ "epoch": 2.4354731965585703,
336
+ "grad_norm": 0.34509792923927307,
337
+ "learning_rate": 1.671949286846276e-05,
338
+ "loss": 0.9026,
339
+ "step": 920
340
+ },
341
+ {
342
+ "epoch": 2.488418266048974,
343
+ "grad_norm": 0.30984020233154297,
344
+ "learning_rate": 1.5134706814580033e-05,
345
+ "loss": 0.9066,
346
+ "step": 940
347
+ },
348
+ {
349
+ "epoch": 2.541363335539378,
350
+ "grad_norm": 0.31895536184310913,
351
+ "learning_rate": 1.3549920760697307e-05,
352
+ "loss": 0.9275,
353
+ "step": 960
354
+ },
355
+ {
356
+ "epoch": 2.5943084050297816,
357
+ "grad_norm": 0.3005692660808563,
358
+ "learning_rate": 1.1965134706814581e-05,
359
+ "loss": 0.9241,
360
+ "step": 980
361
+ },
362
+ {
363
+ "epoch": 2.6472534745201854,
364
+ "grad_norm": 0.325959712266922,
365
+ "learning_rate": 1.0380348652931855e-05,
366
+ "loss": 0.9181,
367
+ "step": 1000
368
+ },
369
+ {
370
+ "epoch": 2.6472534745201854,
371
+ "eval_loss": 0.8340622782707214,
372
+ "eval_runtime": 74.3137,
373
+ "eval_samples_per_second": 10.16,
374
+ "eval_steps_per_second": 2.543,
375
+ "step": 1000
376
+ },
377
+ {
378
+ "epoch": 2.700198544010589,
379
+ "grad_norm": 0.37024152278900146,
380
+ "learning_rate": 8.79556259904913e-06,
381
+ "loss": 0.9156,
382
+ "step": 1020
383
+ },
384
+ {
385
+ "epoch": 2.753143613500993,
386
+ "grad_norm": 0.3021298050880432,
387
+ "learning_rate": 7.2107765451664034e-06,
388
+ "loss": 0.9204,
389
+ "step": 1040
390
+ },
391
+ {
392
+ "epoch": 2.8060886829913967,
393
+ "grad_norm": 0.35478419065475464,
394
+ "learning_rate": 5.625990491283677e-06,
395
+ "loss": 0.9342,
396
+ "step": 1060
397
+ },
398
+ {
399
+ "epoch": 2.8590337524818,
400
+ "grad_norm": 0.34113916754722595,
401
+ "learning_rate": 4.041204437400952e-06,
402
+ "loss": 0.9187,
403
+ "step": 1080
404
+ },
405
+ {
406
+ "epoch": 2.9119788219722036,
407
+ "grad_norm": 0.33516696095466614,
408
+ "learning_rate": 2.456418383518225e-06,
409
+ "loss": 0.9238,
410
+ "step": 1100
411
+ },
412
+ {
413
+ "epoch": 2.9649238914626075,
414
+ "grad_norm": 0.2989369034767151,
415
+ "learning_rate": 8.716323296354993e-07,
416
+ "loss": 0.9188,
417
+ "step": 1120
418
+ }
419
+ ],
420
+ "logging_steps": 20,
421
+ "max_steps": 1131,
422
+ "num_input_tokens_seen": 0,
423
+ "num_train_epochs": 3,
424
+ "save_steps": 1000,
425
+ "stateful_callbacks": {
426
+ "TrainerControl": {
427
+ "args": {
428
+ "should_epoch_stop": false,
429
+ "should_evaluate": false,
430
+ "should_log": false,
431
+ "should_save": true,
432
+ "should_training_stop": true
433
+ },
434
+ "attributes": {}
435
+ }
436
+ },
437
+ "total_flos": 1.101970198757376e+16,
438
+ "train_batch_size": 4,
439
+ "trial_name": null,
440
+ "trial_params": null
441
+ }