sparrow8i8 commited on
Commit
67f9670
·
verified ·
1 Parent(s): 16c048a

add trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +993 -0
trainer_state.json ADDED
@@ -0,0 +1,993 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 600,
3
+ "best_metric": 0.2658693790435791,
4
+ "best_model_checkpoint": "/workspace/anchor/qwen-compassion-web-s3407/train/checkpoint-600",
5
+ "epoch": 1.0,
6
+ "eval_steps": 50,
7
+ "global_step": 614,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.008143322475570033,
14
+ "grad_norm": 8.803375244140625,
15
+ "learning_rate": 2.5806451612903226e-05,
16
+ "loss": 1.2056,
17
+ "step": 5
18
+ },
19
+ {
20
+ "epoch": 0.016286644951140065,
21
+ "grad_norm": 3.7129921913146973,
22
+ "learning_rate": 5.8064516129032266e-05,
23
+ "loss": 0.9384,
24
+ "step": 10
25
+ },
26
+ {
27
+ "epoch": 0.024429967426710098,
28
+ "grad_norm": 2.6070480346679688,
29
+ "learning_rate": 9.032258064516129e-05,
30
+ "loss": 0.8639,
31
+ "step": 15
32
+ },
33
+ {
34
+ "epoch": 0.03257328990228013,
35
+ "grad_norm": 2.043976068496704,
36
+ "learning_rate": 0.00012258064516129034,
37
+ "loss": 0.82,
38
+ "step": 20
39
+ },
40
+ {
41
+ "epoch": 0.04071661237785016,
42
+ "grad_norm": 3.751563787460327,
43
+ "learning_rate": 0.00015483870967741937,
44
+ "loss": 0.7608,
45
+ "step": 25
46
+ },
47
+ {
48
+ "epoch": 0.048859934853420196,
49
+ "grad_norm": 2.8326261043548584,
50
+ "learning_rate": 0.0001870967741935484,
51
+ "loss": 0.6016,
52
+ "step": 30
53
+ },
54
+ {
55
+ "epoch": 0.057003257328990226,
56
+ "grad_norm": 2.2605817317962646,
57
+ "learning_rate": 0.00019998693330688282,
58
+ "loss": 0.6963,
59
+ "step": 35
60
+ },
61
+ {
62
+ "epoch": 0.06514657980456026,
63
+ "grad_norm": 2.8328440189361572,
64
+ "learning_rate": 0.00019990709365929677,
65
+ "loss": 0.5558,
66
+ "step": 40
67
+ },
68
+ {
69
+ "epoch": 0.0732899022801303,
70
+ "grad_norm": 2.2280285358428955,
71
+ "learning_rate": 0.00019975473152335726,
72
+ "loss": 0.5615,
73
+ "step": 45
74
+ },
75
+ {
76
+ "epoch": 0.08143322475570032,
77
+ "grad_norm": 1.7250114679336548,
78
+ "learning_rate": 0.0001995299574986306,
79
+ "loss": 0.5391,
80
+ "step": 50
81
+ },
82
+ {
83
+ "epoch": 0.08143322475570032,
84
+ "eval_loss": 0.374595582485199,
85
+ "eval_runtime": 163.9265,
86
+ "eval_samples_per_second": 1.537,
87
+ "eval_steps_per_second": 0.769,
88
+ "step": 50
89
+ },
90
+ {
91
+ "epoch": 0.08957654723127036,
92
+ "grad_norm": 1.6075690984725952,
93
+ "learning_rate": 0.00019923293474842174,
94
+ "loss": 0.5888,
95
+ "step": 55
96
+ },
97
+ {
98
+ "epoch": 0.09771986970684039,
99
+ "grad_norm": 1.2532232999801636,
100
+ "learning_rate": 0.00019886387888133413,
101
+ "loss": 0.4921,
102
+ "step": 60
103
+ },
104
+ {
105
+ "epoch": 0.10586319218241043,
106
+ "grad_norm": 1.2344268560409546,
107
+ "learning_rate": 0.00019842305779475968,
108
+ "loss": 0.3895,
109
+ "step": 65
110
+ },
111
+ {
112
+ "epoch": 0.11400651465798045,
113
+ "grad_norm": 1.2183916568756104,
114
+ "learning_rate": 0.0001979107914804122,
115
+ "loss": 0.5535,
116
+ "step": 70
117
+ },
118
+ {
119
+ "epoch": 0.12214983713355049,
120
+ "grad_norm": 0.927330493927002,
121
+ "learning_rate": 0.00019732745179204552,
122
+ "loss": 0.4052,
123
+ "step": 75
124
+ },
125
+ {
126
+ "epoch": 0.13029315960912052,
127
+ "grad_norm": 1.6917295455932617,
128
+ "learning_rate": 0.00019667346217552527,
129
+ "loss": 0.4075,
130
+ "step": 80
131
+ },
132
+ {
133
+ "epoch": 0.13843648208469056,
134
+ "grad_norm": 1.3769302368164062,
135
+ "learning_rate": 0.00019594929736144976,
136
+ "loss": 0.5872,
137
+ "step": 85
138
+ },
139
+ {
140
+ "epoch": 0.1465798045602606,
141
+ "grad_norm": 1.2509673833847046,
142
+ "learning_rate": 0.00019515548302054335,
143
+ "loss": 0.4777,
144
+ "step": 90
145
+ },
146
+ {
147
+ "epoch": 0.15472312703583063,
148
+ "grad_norm": 1.104792594909668,
149
+ "learning_rate": 0.0001942925953820725,
150
+ "loss": 0.4456,
151
+ "step": 95
152
+ },
153
+ {
154
+ "epoch": 0.16286644951140064,
155
+ "grad_norm": 1.119893193244934,
156
+ "learning_rate": 0.00019336126081556134,
157
+ "loss": 0.5064,
158
+ "step": 100
159
+ },
160
+ {
161
+ "epoch": 0.16286644951140064,
162
+ "eval_loss": 0.3101080060005188,
163
+ "eval_runtime": 83.9108,
164
+ "eval_samples_per_second": 3.003,
165
+ "eval_steps_per_second": 1.502,
166
+ "step": 100
167
+ },
168
+ {
169
+ "epoch": 0.17100977198697068,
170
+ "grad_norm": 1.2182846069335938,
171
+ "learning_rate": 0.00019236215537611046,
172
+ "loss": 0.4342,
173
+ "step": 105
174
+ },
175
+ {
176
+ "epoch": 0.1791530944625407,
177
+ "grad_norm": 1.2241324186325073,
178
+ "learning_rate": 0.00019129600431364897,
179
+ "loss": 0.4511,
180
+ "step": 110
181
+ },
182
+ {
183
+ "epoch": 0.18729641693811075,
184
+ "grad_norm": 1.2518731355667114,
185
+ "learning_rate": 0.00019016358154647618,
186
+ "loss": 0.4454,
187
+ "step": 115
188
+ },
189
+ {
190
+ "epoch": 0.19543973941368079,
191
+ "grad_norm": 1.3357152938842773,
192
+ "learning_rate": 0.00018896570909947475,
193
+ "loss": 0.4239,
194
+ "step": 120
195
+ },
196
+ {
197
+ "epoch": 0.20358306188925082,
198
+ "grad_norm": 1.1552698612213135,
199
+ "learning_rate": 0.00018770325650740345,
200
+ "loss": 0.418,
201
+ "step": 125
202
+ },
203
+ {
204
+ "epoch": 0.21172638436482086,
205
+ "grad_norm": 1.177544355392456,
206
+ "learning_rate": 0.00018637714018370253,
207
+ "loss": 0.4731,
208
+ "step": 130
209
+ },
210
+ {
211
+ "epoch": 0.21986970684039087,
212
+ "grad_norm": 0.8246702551841736,
213
+ "learning_rate": 0.00018498832275526988,
214
+ "loss": 0.4305,
215
+ "step": 135
216
+ },
217
+ {
218
+ "epoch": 0.2280130293159609,
219
+ "grad_norm": 0.6699772477149963,
220
+ "learning_rate": 0.00018353781236369064,
221
+ "loss": 0.4319,
222
+ "step": 140
223
+ },
224
+ {
225
+ "epoch": 0.23615635179153094,
226
+ "grad_norm": 1.1820833683013916,
227
+ "learning_rate": 0.00018202666193342833,
228
+ "loss": 0.4795,
229
+ "step": 145
230
+ },
231
+ {
232
+ "epoch": 0.24429967426710097,
233
+ "grad_norm": 1.3192893266677856,
234
+ "learning_rate": 0.00018045596840750723,
235
+ "loss": 0.4758,
236
+ "step": 150
237
+ },
238
+ {
239
+ "epoch": 0.24429967426710097,
240
+ "eval_loss": 0.2977350652217865,
241
+ "eval_runtime": 83.4415,
242
+ "eval_samples_per_second": 3.02,
243
+ "eval_steps_per_second": 1.51,
244
+ "step": 150
245
+ },
246
+ {
247
+ "epoch": 0.252442996742671,
248
+ "grad_norm": 0.9550138711929321,
249
+ "learning_rate": 0.0001788268719512427,
250
+ "loss": 0.4393,
251
+ "step": 155
252
+ },
253
+ {
254
+ "epoch": 0.26058631921824105,
255
+ "grad_norm": 1.382379174232483,
256
+ "learning_rate": 0.00017714055512459565,
257
+ "loss": 0.4298,
258
+ "step": 160
259
+ },
260
+ {
261
+ "epoch": 0.2687296416938111,
262
+ "grad_norm": 0.9943409562110901,
263
+ "learning_rate": 0.0001753982420237533,
264
+ "loss": 0.5009,
265
+ "step": 165
266
+ },
267
+ {
268
+ "epoch": 0.2768729641693811,
269
+ "grad_norm": 1.2051012516021729,
270
+ "learning_rate": 0.00017360119739255852,
271
+ "loss": 0.3328,
272
+ "step": 170
273
+ },
274
+ {
275
+ "epoch": 0.28501628664495116,
276
+ "grad_norm": 0.7656549215316772,
277
+ "learning_rate": 0.00017175072570443312,
278
+ "loss": 0.4809,
279
+ "step": 175
280
+ },
281
+ {
282
+ "epoch": 0.2931596091205212,
283
+ "grad_norm": 0.8127760291099548,
284
+ "learning_rate": 0.00016984817021546177,
285
+ "loss": 0.37,
286
+ "step": 180
287
+ },
288
+ {
289
+ "epoch": 0.30130293159609123,
290
+ "grad_norm": 1.5806375741958618,
291
+ "learning_rate": 0.000167894911989323,
292
+ "loss": 0.4262,
293
+ "step": 185
294
+ },
295
+ {
296
+ "epoch": 0.30944625407166126,
297
+ "grad_norm": 1.1315789222717285,
298
+ "learning_rate": 0.00016589236889477646,
299
+ "loss": 0.3342,
300
+ "step": 190
301
+ },
302
+ {
303
+ "epoch": 0.31758957654723124,
304
+ "grad_norm": 1.2172818183898926,
305
+ "learning_rate": 0.00016384199457643262,
306
+ "loss": 0.4647,
307
+ "step": 195
308
+ },
309
+ {
310
+ "epoch": 0.3257328990228013,
311
+ "grad_norm": 1.0508732795715332,
312
+ "learning_rate": 0.00016174527739955342,
313
+ "loss": 0.5813,
314
+ "step": 200
315
+ },
316
+ {
317
+ "epoch": 0.3257328990228013,
318
+ "eval_loss": 0.29167938232421875,
319
+ "eval_runtime": 83.6844,
320
+ "eval_samples_per_second": 3.011,
321
+ "eval_steps_per_second": 1.506,
322
+ "step": 200
323
+ },
324
+ {
325
+ "epoch": 0.3338762214983713,
326
+ "grad_norm": 0.6504319310188293,
327
+ "learning_rate": 0.00015960373936964892,
328
+ "loss": 0.385,
329
+ "step": 205
330
+ },
331
+ {
332
+ "epoch": 0.34201954397394135,
333
+ "grad_norm": 1.1721259355545044,
334
+ "learning_rate": 0.0001574189350276545,
335
+ "loss": 0.4746,
336
+ "step": 210
337
+ },
338
+ {
339
+ "epoch": 0.3501628664495114,
340
+ "grad_norm": 1.2938706874847412,
341
+ "learning_rate": 0.00015519245032149083,
342
+ "loss": 0.5205,
343
+ "step": 215
344
+ },
345
+ {
346
+ "epoch": 0.3583061889250814,
347
+ "grad_norm": 0.9688608646392822,
348
+ "learning_rate": 0.0001529259014548253,
349
+ "loss": 0.4305,
350
+ "step": 220
351
+ },
352
+ {
353
+ "epoch": 0.36644951140065146,
354
+ "grad_norm": 1.1749860048294067,
355
+ "learning_rate": 0.00015062093371387097,
356
+ "loss": 0.4359,
357
+ "step": 225
358
+ },
359
+ {
360
+ "epoch": 0.3745928338762215,
361
+ "grad_norm": 1.3086587190628052,
362
+ "learning_rate": 0.00014827922027307451,
363
+ "loss": 0.4028,
364
+ "step": 230
365
+ },
366
+ {
367
+ "epoch": 0.38273615635179153,
368
+ "grad_norm": 1.0187360048294067,
369
+ "learning_rate": 0.00014590246098055996,
370
+ "loss": 0.4615,
371
+ "step": 235
372
+ },
373
+ {
374
+ "epoch": 0.39087947882736157,
375
+ "grad_norm": 0.8567041754722595,
376
+ "learning_rate": 0.00014349238112421024,
377
+ "loss": 0.3113,
378
+ "step": 240
379
+ },
380
+ {
381
+ "epoch": 0.3990228013029316,
382
+ "grad_norm": 0.9721431136131287,
383
+ "learning_rate": 0.000141050730179282,
384
+ "loss": 0.4088,
385
+ "step": 245
386
+ },
387
+ {
388
+ "epoch": 0.40716612377850164,
389
+ "grad_norm": 1.1704206466674805,
390
+ "learning_rate": 0.0001385792805384625,
391
+ "loss": 0.5045,
392
+ "step": 250
393
+ },
394
+ {
395
+ "epoch": 0.40716612377850164,
396
+ "eval_loss": 0.2854733467102051,
397
+ "eval_runtime": 83.6144,
398
+ "eval_samples_per_second": 3.014,
399
+ "eval_steps_per_second": 1.507,
400
+ "step": 250
401
+ },
402
+ {
403
+ "epoch": 0.4153094462540717,
404
+ "grad_norm": 1.115260124206543,
405
+ "learning_rate": 0.00013607982622529133,
406
+ "loss": 0.4098,
407
+ "step": 255
408
+ },
409
+ {
410
+ "epoch": 0.4234527687296417,
411
+ "grad_norm": 1.2716691493988037,
412
+ "learning_rate": 0.00013355418159187985,
413
+ "loss": 0.4753,
414
+ "step": 260
415
+ },
416
+ {
417
+ "epoch": 0.43159609120521175,
418
+ "grad_norm": 0.6622514724731445,
419
+ "learning_rate": 0.00013100418000187419,
420
+ "loss": 0.4589,
421
+ "step": 265
422
+ },
423
+ {
424
+ "epoch": 0.43973941368078173,
425
+ "grad_norm": 1.056097149848938,
426
+ "learning_rate": 0.0001284316724996181,
427
+ "loss": 0.4588,
428
+ "step": 270
429
+ },
430
+ {
431
+ "epoch": 0.44788273615635177,
432
+ "grad_norm": 1.1712679862976074,
433
+ "learning_rate": 0.00012583852646648095,
434
+ "loss": 0.5739,
435
+ "step": 275
436
+ },
437
+ {
438
+ "epoch": 0.4560260586319218,
439
+ "grad_norm": 1.0975291728973389,
440
+ "learning_rate": 0.00012322662426532708,
441
+ "loss": 0.4602,
442
+ "step": 280
443
+ },
444
+ {
445
+ "epoch": 0.46416938110749184,
446
+ "grad_norm": 0.7166972756385803,
447
+ "learning_rate": 0.00012059786187410984,
448
+ "loss": 0.3027,
449
+ "step": 285
450
+ },
451
+ {
452
+ "epoch": 0.4723127035830619,
453
+ "grad_norm": 0.63871830701828,
454
+ "learning_rate": 0.00011795414750958265,
455
+ "loss": 0.2792,
456
+ "step": 290
457
+ },
458
+ {
459
+ "epoch": 0.4804560260586319,
460
+ "grad_norm": 1.0939736366271973,
461
+ "learning_rate": 0.00011529740024212565,
462
+ "loss": 0.4039,
463
+ "step": 295
464
+ },
465
+ {
466
+ "epoch": 0.48859934853420195,
467
+ "grad_norm": 1.2159818410873413,
468
+ "learning_rate": 0.00011262954860269399,
469
+ "loss": 0.5488,
470
+ "step": 300
471
+ },
472
+ {
473
+ "epoch": 0.48859934853420195,
474
+ "eval_loss": 0.2815026044845581,
475
+ "eval_runtime": 84.5088,
476
+ "eval_samples_per_second": 2.982,
477
+ "eval_steps_per_second": 1.491,
478
+ "step": 300
479
+ },
480
+ {
481
+ "epoch": 0.496742671009772,
482
+ "grad_norm": 1.18336021900177,
483
+ "learning_rate": 0.0001099525291828986,
484
+ "loss": 0.3673,
485
+ "step": 305
486
+ },
487
+ {
488
+ "epoch": 0.504885993485342,
489
+ "grad_norm": 1.2917401790618896,
490
+ "learning_rate": 0.00010726828522923562,
491
+ "loss": 0.402,
492
+ "step": 310
493
+ },
494
+ {
495
+ "epoch": 0.5130293159609121,
496
+ "grad_norm": 1.1676180362701416,
497
+ "learning_rate": 0.00010457876523248518,
498
+ "loss": 0.4644,
499
+ "step": 315
500
+ },
501
+ {
502
+ "epoch": 0.5211726384364821,
503
+ "grad_norm": 0.6200851202011108,
504
+ "learning_rate": 0.00010188592151330343,
505
+ "loss": 0.3415,
506
+ "step": 320
507
+ },
508
+ {
509
+ "epoch": 0.5293159609120521,
510
+ "grad_norm": 0.6668900847434998,
511
+ "learning_rate": 9.919170880503415e-05,
512
+ "loss": 0.4088,
513
+ "step": 325
514
+ },
515
+ {
516
+ "epoch": 0.5374592833876222,
517
+ "grad_norm": 1.1132478713989258,
518
+ "learning_rate": 9.649808283476941e-05,
519
+ "loss": 0.2441,
520
+ "step": 330
521
+ },
522
+ {
523
+ "epoch": 0.5456026058631922,
524
+ "grad_norm": 0.7450593113899231,
525
+ "learning_rate": 9.38069989036886e-05,
526
+ "loss": 0.4173,
527
+ "step": 335
528
+ },
529
+ {
530
+ "epoch": 0.5537459283387622,
531
+ "grad_norm": 0.7038878798484802,
532
+ "learning_rate": 9.112041046770653e-05,
533
+ "loss": 0.2244,
534
+ "step": 340
535
+ },
536
+ {
537
+ "epoch": 0.5618892508143323,
538
+ "grad_norm": 1.1429181098937988,
539
+ "learning_rate": 8.844026771946147e-05,
540
+ "loss": 0.4247,
541
+ "step": 345
542
+ },
543
+ {
544
+ "epoch": 0.5700325732899023,
545
+ "grad_norm": 1.2036070823669434,
546
+ "learning_rate": 8.57685161726715e-05,
547
+ "loss": 0.4933,
548
+ "step": 350
549
+ },
550
+ {
551
+ "epoch": 0.5700325732899023,
552
+ "eval_loss": 0.2762308418750763,
553
+ "eval_runtime": 83.4746,
554
+ "eval_samples_per_second": 3.019,
555
+ "eval_steps_per_second": 1.509,
556
+ "step": 350
557
+ },
558
+ {
559
+ "epoch": 0.5781758957654723,
560
+ "grad_norm": 0.9277458786964417,
561
+ "learning_rate": 8.310709524988743e-05,
562
+ "loss": 0.4341,
563
+ "step": 355
564
+ },
565
+ {
566
+ "epoch": 0.5863192182410424,
567
+ "grad_norm": 1.0389032363891602,
568
+ "learning_rate": 8.045793687466757e-05,
569
+ "loss": 0.327,
570
+ "step": 360
571
+ },
572
+ {
573
+ "epoch": 0.5944625407166124,
574
+ "grad_norm": 1.003150463104248,
575
+ "learning_rate": 7.782296406919557e-05,
576
+ "loss": 0.4683,
577
+ "step": 365
578
+ },
579
+ {
580
+ "epoch": 0.6026058631921825,
581
+ "grad_norm": 1.0817468166351318,
582
+ "learning_rate": 7.520408955836007e-05,
583
+ "loss": 0.2979,
584
+ "step": 370
585
+ },
586
+ {
587
+ "epoch": 0.6107491856677525,
588
+ "grad_norm": 0.6353104114532471,
589
+ "learning_rate": 7.260321438130913e-05,
590
+ "loss": 0.3314,
591
+ "step": 375
592
+ },
593
+ {
594
+ "epoch": 0.6188925081433225,
595
+ "grad_norm": 0.8815907835960388,
596
+ "learning_rate": 7.002222651148714e-05,
597
+ "loss": 0.2524,
598
+ "step": 380
599
+ },
600
+ {
601
+ "epoch": 0.6270358306188925,
602
+ "grad_norm": 1.3402814865112305,
603
+ "learning_rate": 6.746299948615631e-05,
604
+ "loss": 0.5739,
605
+ "step": 385
606
+ },
607
+ {
608
+ "epoch": 0.6351791530944625,
609
+ "grad_norm": 0.8972781896591187,
610
+ "learning_rate": 6.492739104639727e-05,
611
+ "loss": 0.3207,
612
+ "step": 390
613
+ },
614
+ {
615
+ "epoch": 0.6433224755700325,
616
+ "grad_norm": 0.8222898840904236,
617
+ "learning_rate": 6.24172417885762e-05,
618
+ "loss": 0.4004,
619
+ "step": 395
620
+ },
621
+ {
622
+ "epoch": 0.6514657980456026,
623
+ "grad_norm": 1.0864665508270264,
624
+ "learning_rate": 5.9934373828257106e-05,
625
+ "loss": 0.335,
626
+ "step": 400
627
+ },
628
+ {
629
+ "epoch": 0.6514657980456026,
630
+ "eval_loss": 0.27289000153541565,
631
+ "eval_runtime": 83.8616,
632
+ "eval_samples_per_second": 3.005,
633
+ "eval_steps_per_second": 1.502,
634
+ "step": 400
635
+ },
636
+ {
637
+ "epoch": 0.6596091205211726,
638
+ "grad_norm": 0.620107889175415,
639
+ "learning_rate": 5.7480589477529545e-05,
640
+ "loss": 0.3715,
641
+ "step": 405
642
+ },
643
+ {
644
+ "epoch": 0.6677524429967426,
645
+ "grad_norm": 0.9920016527175903,
646
+ "learning_rate": 5.505766993671162e-05,
647
+ "loss": 0.4575,
648
+ "step": 410
649
+ },
650
+ {
651
+ "epoch": 0.6758957654723127,
652
+ "grad_norm": 1.1352708339691162,
653
+ "learning_rate": 5.266737400137765e-05,
654
+ "loss": 0.2978,
655
+ "step": 415
656
+ },
657
+ {
658
+ "epoch": 0.6840390879478827,
659
+ "grad_norm": 1.303180456161499,
660
+ "learning_rate": 5.031143678565005e-05,
661
+ "loss": 0.3386,
662
+ "step": 420
663
+ },
664
+ {
665
+ "epoch": 0.6921824104234527,
666
+ "grad_norm": 0.6123512387275696,
667
+ "learning_rate": 4.799156846268095e-05,
668
+ "loss": 0.3016,
669
+ "step": 425
670
+ },
671
+ {
672
+ "epoch": 0.7003257328990228,
673
+ "grad_norm": 1.2777764797210693,
674
+ "learning_rate": 4.57094530232389e-05,
675
+ "loss": 0.4377,
676
+ "step": 430
677
+ },
678
+ {
679
+ "epoch": 0.7084690553745928,
680
+ "grad_norm": 0.9900017380714417,
681
+ "learning_rate": 4.346674705330117e-05,
682
+ "loss": 0.4437,
683
+ "step": 435
684
+ },
685
+ {
686
+ "epoch": 0.7166123778501629,
687
+ "grad_norm": 1.029066801071167,
688
+ "learning_rate": 4.1265078531538916e-05,
689
+ "loss": 0.3486,
690
+ "step": 440
691
+ },
692
+ {
693
+ "epoch": 0.7247557003257329,
694
+ "grad_norm": 1.181535005569458,
695
+ "learning_rate": 3.9106045647569e-05,
696
+ "loss": 0.4141,
697
+ "step": 445
698
+ },
699
+ {
700
+ "epoch": 0.7328990228013029,
701
+ "grad_norm": 1.3165513277053833,
702
+ "learning_rate": 3.69912156418289e-05,
703
+ "loss": 0.5187,
704
+ "step": 450
705
+ },
706
+ {
707
+ "epoch": 0.7328990228013029,
708
+ "eval_loss": 0.2702295184135437,
709
+ "eval_runtime": 83.3618,
710
+ "eval_samples_per_second": 3.023,
711
+ "eval_steps_per_second": 1.511,
712
+ "step": 450
713
+ },
714
+ {
715
+ "epoch": 0.741042345276873,
716
+ "grad_norm": 1.1030694246292114,
717
+ "learning_rate": 3.492212366791831e-05,
718
+ "loss": 0.3816,
719
+ "step": 455
720
+ },
721
+ {
722
+ "epoch": 0.749185667752443,
723
+ "grad_norm": 1.0479238033294678,
724
+ "learning_rate": 3.290027167823204e-05,
725
+ "loss": 0.3815,
726
+ "step": 460
727
+ },
728
+ {
729
+ "epoch": 0.757328990228013,
730
+ "grad_norm": 1.0574332475662231,
731
+ "learning_rate": 3.092712733369387e-05,
732
+ "loss": 0.3443,
733
+ "step": 465
734
+ },
735
+ {
736
+ "epoch": 0.7654723127035831,
737
+ "grad_norm": 0.7265905737876892,
738
+ "learning_rate": 2.9004122938382617e-05,
739
+ "loss": 0.3229,
740
+ "step": 470
741
+ },
742
+ {
743
+ "epoch": 0.7736156351791531,
744
+ "grad_norm": 0.9613872766494751,
745
+ "learning_rate": 2.7132654399823444e-05,
746
+ "loss": 0.4289,
747
+ "step": 475
748
+ },
749
+ {
750
+ "epoch": 0.7817589576547231,
751
+ "grad_norm": 0.8705881834030151,
752
+ "learning_rate": 2.531408021569982e-05,
753
+ "loss": 0.391,
754
+ "step": 480
755
+ },
756
+ {
757
+ "epoch": 0.7899022801302932,
758
+ "grad_norm": 0.9562520384788513,
759
+ "learning_rate": 2.3549720487720738e-05,
760
+ "loss": 0.4853,
761
+ "step": 485
762
+ },
763
+ {
764
+ "epoch": 0.7980456026058632,
765
+ "grad_norm": 1.064986228942871,
766
+ "learning_rate": 2.184085596336011e-05,
767
+ "loss": 0.4105,
768
+ "step": 490
769
+ },
770
+ {
771
+ "epoch": 0.8061889250814332,
772
+ "grad_norm": 1.1484014987945557,
773
+ "learning_rate": 2.0188727106162874e-05,
774
+ "loss": 0.4579,
775
+ "step": 495
776
+ },
777
+ {
778
+ "epoch": 0.8143322475570033,
779
+ "grad_norm": 0.9046717286109924,
780
+ "learning_rate": 1.8594533195293427e-05,
781
+ "loss": 0.4128,
782
+ "step": 500
783
+ },
784
+ {
785
+ "epoch": 0.8143322475570033,
786
+ "eval_loss": 0.26746508479118347,
787
+ "eval_runtime": 83.6362,
788
+ "eval_samples_per_second": 3.013,
789
+ "eval_steps_per_second": 1.507,
790
+ "step": 500
791
+ },
792
+ {
793
+ "epoch": 0.8224755700325733,
794
+ "grad_norm": 1.0069684982299805,
795
+ "learning_rate": 1.7059431454979824e-05,
796
+ "loss": 0.4211,
797
+ "step": 505
798
+ },
799
+ {
800
+ "epoch": 0.8306188925081434,
801
+ "grad_norm": 1.418688416481018,
802
+ "learning_rate": 1.5584536214485457e-05,
803
+ "loss": 0.3868,
804
+ "step": 510
805
+ },
806
+ {
807
+ "epoch": 0.8387622149837134,
808
+ "grad_norm": 1.0533849000930786,
809
+ "learning_rate": 1.4170918099218166e-05,
810
+ "loss": 0.3819,
811
+ "step": 515
812
+ },
813
+ {
814
+ "epoch": 0.8469055374592834,
815
+ "grad_norm": 0.9637581706047058,
816
+ "learning_rate": 1.2819603253564205e-05,
817
+ "loss": 0.3536,
818
+ "step": 520
819
+ },
820
+ {
821
+ "epoch": 0.8550488599348535,
822
+ "grad_norm": 1.0241345167160034,
823
+ "learning_rate": 1.153157259601062e-05,
824
+ "loss": 0.3699,
825
+ "step": 525
826
+ },
827
+ {
828
+ "epoch": 0.8631921824104235,
829
+ "grad_norm": 0.985444962978363,
830
+ "learning_rate": 1.030776110709718e-05,
831
+ "loss": 0.4363,
832
+ "step": 530
833
+ },
834
+ {
835
+ "epoch": 0.8713355048859935,
836
+ "grad_norm": 1.0319409370422363,
837
+ "learning_rate": 9.149057150714801e-06,
838
+ "loss": 0.4192,
839
+ "step": 535
840
+ },
841
+ {
842
+ "epoch": 0.8794788273615635,
843
+ "grad_norm": 1.0504857301712036,
844
+ "learning_rate": 8.056301829242784e-06,
845
+ "loss": 0.3702,
846
+ "step": 540
847
+ },
848
+ {
849
+ "epoch": 0.8876221498371335,
850
+ "grad_norm": 0.662237286567688,
851
+ "learning_rate": 7.030288372993066e-06,
852
+ "loss": 0.2473,
853
+ "step": 545
854
+ },
855
+ {
856
+ "epoch": 0.8957654723127035,
857
+ "grad_norm": 1.0782599449157715,
858
+ "learning_rate": 6.0717615644051206e-06,
859
+ "loss": 0.4462,
860
+ "step": 550
861
+ },
862
+ {
863
+ "epoch": 0.8957654723127035,
864
+ "eval_loss": 0.2662240266799927,
865
+ "eval_runtime": 83.3987,
866
+ "eval_samples_per_second": 3.022,
867
+ "eval_steps_per_second": 1.511,
868
+ "step": 550
869
+ },
870
+ {
871
+ "epoch": 0.9039087947882736,
872
+ "grad_norm": 0.8521231412887573,
873
+ "learning_rate": 5.181417197408734e-06,
874
+ "loss": 0.4453,
875
+ "step": 555
876
+ },
877
+ {
878
+ "epoch": 0.9120521172638436,
879
+ "grad_norm": 0.8748440742492676,
880
+ "learning_rate": 4.359901572347758e-06,
881
+ "loss": 0.3904,
882
+ "step": 560
883
+ },
884
+ {
885
+ "epoch": 0.9201954397394136,
886
+ "grad_norm": 1.3463112115859985,
887
+ "learning_rate": 3.607811026831176e-06,
888
+ "loss": 0.4692,
889
+ "step": 565
890
+ },
891
+ {
892
+ "epoch": 0.9283387622149837,
893
+ "grad_norm": 1.050785779953003,
894
+ "learning_rate": 2.9256915028519573e-06,
895
+ "loss": 0.3126,
896
+ "step": 570
897
+ },
898
+ {
899
+ "epoch": 0.9364820846905537,
900
+ "grad_norm": 0.9636266231536865,
901
+ "learning_rate": 2.3140381504882737e-06,
902
+ "loss": 0.4212,
903
+ "step": 575
904
+ },
905
+ {
906
+ "epoch": 0.9446254071661238,
907
+ "grad_norm": 0.9108374714851379,
908
+ "learning_rate": 1.7732949684743594e-06,
909
+ "loss": 0.3299,
910
+ "step": 580
911
+ },
912
+ {
913
+ "epoch": 0.9527687296416938,
914
+ "grad_norm": 1.299425482749939,
915
+ "learning_rate": 1.303854481902067e-06,
916
+ "loss": 0.3926,
917
+ "step": 585
918
+ },
919
+ {
920
+ "epoch": 0.9609120521172638,
921
+ "grad_norm": 1.0949175357818604,
922
+ "learning_rate": 9.060574572873237e-07,
923
+ "loss": 0.3908,
924
+ "step": 590
925
+ },
926
+ {
927
+ "epoch": 0.9690553745928339,
928
+ "grad_norm": 1.1094610691070557,
929
+ "learning_rate": 5.801926552078563e-07,
930
+ "loss": 0.3617,
931
+ "step": 595
932
+ },
933
+ {
934
+ "epoch": 0.9771986970684039,
935
+ "grad_norm": 1.1713780164718628,
936
+ "learning_rate": 3.264966206921294e-07,
937
+ "loss": 0.3869,
938
+ "step": 600
939
+ },
940
+ {
941
+ "epoch": 0.9771986970684039,
942
+ "eval_loss": 0.2658693790435791,
943
+ "eval_runtime": 83.6731,
944
+ "eval_samples_per_second": 3.012,
945
+ "eval_steps_per_second": 1.506,
946
+ "step": 600
947
+ },
948
+ {
949
+ "epoch": 0.9853420195439739,
950
+ "grad_norm": 1.2683329582214355,
951
+ "learning_rate": 1.451535115114866e-07,
952
+ "loss": 0.4507,
953
+ "step": 605
954
+ },
955
+ {
956
+ "epoch": 0.993485342019544,
957
+ "grad_norm": 0.9197269678115845,
958
+ "learning_rate": 3.629496450011649e-08,
959
+ "loss": 0.3196,
960
+ "step": 610
961
+ }
962
+ ],
963
+ "logging_steps": 5,
964
+ "max_steps": 614,
965
+ "num_input_tokens_seen": 0,
966
+ "num_train_epochs": 1,
967
+ "save_steps": 100,
968
+ "stateful_callbacks": {
969
+ "EarlyStoppingCallback": {
970
+ "args": {
971
+ "early_stopping_patience": 2,
972
+ "early_stopping_threshold": 0.001
973
+ },
974
+ "attributes": {
975
+ "early_stopping_patience_counter": 1
976
+ }
977
+ },
978
+ "TrainerControl": {
979
+ "args": {
980
+ "should_epoch_stop": false,
981
+ "should_evaluate": false,
982
+ "should_log": false,
983
+ "should_save": true,
984
+ "should_training_stop": true
985
+ },
986
+ "attributes": {}
987
+ }
988
+ },
989
+ "total_flos": 2.5080656292484915e+18,
990
+ "train_batch_size": 2,
991
+ "trial_name": null,
992
+ "trial_params": null
993
+ }