mikhail-panzo commited on
Commit
a382caa
1 Parent(s): 08e8f19

Training in progress, step 2500, checkpoint

Browse files
checkpoint-2500/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:32f3a1545c6b7f8058c4294b66661bbd9e3d4dec0f402db560eba66003a6bb19
3
  size 577789320
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7493ad5637e6ccf113491f9e3ebc087d34378e0d2a945336aceb41119bc02077
3
  size 577789320
checkpoint-2500/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:02dad2ed8984c6eca1db04c9d3e2c43ae29372f7956bcc4fc7c2e307548955de
3
  size 1155772233
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4f370d362c714dd35f5e4e0a87163395f9a19b0a5b2aae1e343376a83b30b2b0
3
  size 1155772233
checkpoint-2500/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e12cffd706a7a40f72df497f7556e19c62f5295a943d4c2d84107cc3757766d8
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a806455f7b2323ada1257a2ccc10634eadf1cae42d8679a695b446f85f9dcda8
3
  size 14244
checkpoint-2500/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1cdca242b8590cd159290da20e22d5d5d903dee7a387d8e3f7463cb40559acf6
3
  size 1064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5b56c909bdfd359503d6bcec618fbe35b9e2276a74764820065954757bce0f5
3
  size 1064
checkpoint-2500/trainer_state.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
- "best_metric": 0.42134132981300354,
3
  "best_model_checkpoint": "mikhail-panzo/fil_b32_le5_s4000/checkpoint-2500",
4
- "epoch": 55.55555555555556,
5
  "eval_steps": 500,
6
  "global_step": 2500,
7
  "is_hyper_param_search": false,
@@ -9,400 +9,400 @@
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 1.1111111111111112,
13
- "grad_norm": 4.398683071136475,
14
- "learning_rate": 2.4500000000000004e-07,
15
- "loss": 0.8108,
16
  "step": 50
17
  },
18
  {
19
- "epoch": 2.2222222222222223,
20
- "grad_norm": 3.726436138153076,
21
- "learning_rate": 4.95e-07,
22
- "loss": 0.792,
23
  "step": 100
24
  },
25
  {
26
- "epoch": 3.3333333333333335,
27
- "grad_norm": 2.0379676818847656,
28
- "learning_rate": 7.450000000000001e-07,
29
- "loss": 0.7413,
30
  "step": 150
31
  },
32
  {
33
- "epoch": 4.444444444444445,
34
- "grad_norm": 2.0460364818573,
35
- "learning_rate": 9.950000000000002e-07,
36
- "loss": 0.7342,
37
  "step": 200
38
  },
39
  {
40
- "epoch": 5.555555555555555,
41
- "grad_norm": 2.803539276123047,
42
- "learning_rate": 1.2450000000000002e-06,
43
- "loss": 0.7147,
44
  "step": 250
45
  },
46
  {
47
- "epoch": 6.666666666666667,
48
- "grad_norm": 1.9818241596221924,
49
- "learning_rate": 1.495e-06,
50
- "loss": 0.7072,
51
  "step": 300
52
  },
53
  {
54
- "epoch": 7.777777777777778,
55
- "grad_norm": 3.081399917602539,
56
- "learning_rate": 1.745e-06,
57
- "loss": 0.6843,
58
  "step": 350
59
  },
60
  {
61
- "epoch": 8.88888888888889,
62
- "grad_norm": 2.4465785026550293,
63
- "learning_rate": 1.9950000000000004e-06,
64
- "loss": 0.657,
65
  "step": 400
66
  },
67
  {
68
- "epoch": 10.0,
69
- "grad_norm": 2.929323434829712,
70
- "learning_rate": 2.245e-06,
71
- "loss": 0.6597,
72
  "step": 450
73
  },
74
  {
75
- "epoch": 11.11111111111111,
76
- "grad_norm": 1.419311285018921,
77
- "learning_rate": 2.4950000000000003e-06,
78
- "loss": 0.6246,
79
  "step": 500
80
  },
81
  {
82
- "epoch": 11.11111111111111,
83
- "eval_loss": 0.5375720262527466,
84
- "eval_runtime": 9.2568,
85
- "eval_samples_per_second": 17.177,
86
- "eval_steps_per_second": 2.161,
87
  "step": 500
88
  },
89
  {
90
- "epoch": 12.222222222222221,
91
- "grad_norm": 2.3438374996185303,
92
- "learning_rate": 2.7450000000000004e-06,
93
- "loss": 0.6018,
94
  "step": 550
95
  },
96
  {
97
- "epoch": 13.333333333333334,
98
- "grad_norm": 1.7441222667694092,
99
- "learning_rate": 2.995e-06,
100
- "loss": 0.5547,
101
  "step": 600
102
  },
103
  {
104
- "epoch": 14.444444444444445,
105
- "grad_norm": 1.5592021942138672,
106
- "learning_rate": 3.2450000000000003e-06,
107
- "loss": 0.5466,
108
  "step": 650
109
  },
110
  {
111
- "epoch": 15.555555555555555,
112
- "grad_norm": 1.305444359779358,
113
- "learning_rate": 3.495e-06,
114
- "loss": 0.535,
115
  "step": 700
116
  },
117
  {
118
- "epoch": 16.666666666666668,
119
- "grad_norm": 2.2067477703094482,
120
- "learning_rate": 3.745e-06,
121
- "loss": 0.5343,
122
  "step": 750
123
  },
124
  {
125
- "epoch": 17.77777777777778,
126
- "grad_norm": 1.667480707168579,
127
- "learning_rate": 3.995000000000001e-06,
128
- "loss": 0.5296,
129
  "step": 800
130
  },
131
  {
132
- "epoch": 18.88888888888889,
133
- "grad_norm": 1.5465797185897827,
134
- "learning_rate": 4.245e-06,
135
- "loss": 0.5198,
136
  "step": 850
137
  },
138
  {
139
- "epoch": 20.0,
140
- "grad_norm": 2.760589361190796,
141
- "learning_rate": 4.495e-06,
142
- "loss": 0.5193,
143
  "step": 900
144
  },
145
  {
146
- "epoch": 21.11111111111111,
147
- "grad_norm": 1.4570050239562988,
148
- "learning_rate": 4.745e-06,
149
- "loss": 0.5124,
150
  "step": 950
151
  },
152
  {
153
- "epoch": 22.22222222222222,
154
- "grad_norm": 2.1914265155792236,
155
- "learning_rate": 4.9950000000000005e-06,
156
- "loss": 0.5034,
157
  "step": 1000
158
  },
159
  {
160
- "epoch": 22.22222222222222,
161
- "eval_loss": 0.45661377906799316,
162
- "eval_runtime": 9.2158,
163
- "eval_samples_per_second": 17.253,
164
- "eval_steps_per_second": 2.17,
165
  "step": 1000
166
  },
167
  {
168
- "epoch": 23.333333333333332,
169
- "grad_norm": 1.941751480102539,
170
- "learning_rate": 5.245e-06,
171
- "loss": 0.5046,
172
  "step": 1050
173
  },
174
  {
175
- "epoch": 24.444444444444443,
176
- "grad_norm": 1.9293475151062012,
177
- "learning_rate": 5.495000000000001e-06,
178
- "loss": 0.4948,
179
  "step": 1100
180
  },
181
  {
182
- "epoch": 25.555555555555557,
183
- "grad_norm": 1.3278849124908447,
184
- "learning_rate": 5.745000000000001e-06,
185
- "loss": 0.4931,
186
  "step": 1150
187
  },
188
  {
189
- "epoch": 26.666666666666668,
190
- "grad_norm": 1.1657841205596924,
191
- "learning_rate": 5.995000000000001e-06,
192
- "loss": 0.496,
193
  "step": 1200
194
  },
195
  {
196
- "epoch": 27.77777777777778,
197
- "grad_norm": 1.8730145692825317,
198
- "learning_rate": 6.245000000000001e-06,
199
- "loss": 0.4939,
200
  "step": 1250
201
  },
202
  {
203
- "epoch": 28.88888888888889,
204
- "grad_norm": 2.1156740188598633,
205
- "learning_rate": 6.4950000000000005e-06,
206
- "loss": 0.4908,
207
  "step": 1300
208
  },
209
  {
210
- "epoch": 30.0,
211
- "grad_norm": 1.799806833267212,
212
- "learning_rate": 6.745000000000001e-06,
213
- "loss": 0.4875,
214
  "step": 1350
215
  },
216
  {
217
- "epoch": 31.11111111111111,
218
- "grad_norm": 1.762253761291504,
219
- "learning_rate": 6.995000000000001e-06,
220
- "loss": 0.4836,
221
  "step": 1400
222
  },
223
  {
224
- "epoch": 32.22222222222222,
225
- "grad_norm": 1.305191993713379,
226
- "learning_rate": 7.245000000000001e-06,
227
- "loss": 0.4811,
228
  "step": 1450
229
  },
230
  {
231
- "epoch": 33.333333333333336,
232
- "grad_norm": 1.5870022773742676,
233
- "learning_rate": 7.495000000000001e-06,
234
- "loss": 0.4807,
235
  "step": 1500
236
  },
237
  {
238
- "epoch": 33.333333333333336,
239
- "eval_loss": 0.43601250648498535,
240
- "eval_runtime": 9.5435,
241
- "eval_samples_per_second": 16.66,
242
- "eval_steps_per_second": 2.096,
243
  "step": 1500
244
  },
245
  {
246
- "epoch": 34.44444444444444,
247
- "grad_norm": 2.632877826690674,
248
- "learning_rate": 7.745e-06,
249
- "loss": 0.4782,
250
  "step": 1550
251
  },
252
  {
253
- "epoch": 35.55555555555556,
254
- "grad_norm": 1.3705276250839233,
255
- "learning_rate": 7.995e-06,
256
- "loss": 0.48,
257
  "step": 1600
258
  },
259
  {
260
- "epoch": 36.666666666666664,
261
- "grad_norm": 1.9232587814331055,
262
- "learning_rate": 8.245000000000002e-06,
263
- "loss": 0.47,
264
  "step": 1650
265
  },
266
  {
267
- "epoch": 37.77777777777778,
268
- "grad_norm": 1.3885018825531006,
269
- "learning_rate": 8.495e-06,
270
- "loss": 0.4711,
271
  "step": 1700
272
  },
273
  {
274
- "epoch": 38.888888888888886,
275
- "grad_norm": 2.4857094287872314,
276
- "learning_rate": 8.745000000000002e-06,
277
- "loss": 0.4721,
278
  "step": 1750
279
  },
280
  {
281
- "epoch": 40.0,
282
- "grad_norm": 1.7662663459777832,
283
- "learning_rate": 8.995000000000001e-06,
284
- "loss": 0.4656,
285
  "step": 1800
286
  },
287
  {
288
- "epoch": 41.111111111111114,
289
- "grad_norm": 1.906548261642456,
290
- "learning_rate": 9.245e-06,
291
- "loss": 0.463,
292
  "step": 1850
293
  },
294
  {
295
- "epoch": 42.22222222222222,
296
- "grad_norm": 1.7204883098602295,
297
- "learning_rate": 9.495000000000001e-06,
298
- "loss": 0.4592,
299
  "step": 1900
300
  },
301
  {
302
- "epoch": 43.333333333333336,
303
- "grad_norm": 1.3313167095184326,
304
- "learning_rate": 9.745e-06,
305
- "loss": 0.4581,
306
  "step": 1950
307
  },
308
  {
309
- "epoch": 44.44444444444444,
310
- "grad_norm": 1.94256591796875,
311
- "learning_rate": 9.995000000000002e-06,
312
- "loss": 0.4582,
313
  "step": 2000
314
  },
315
  {
316
- "epoch": 44.44444444444444,
317
- "eval_loss": 0.4244868755340576,
318
- "eval_runtime": 9.1316,
319
- "eval_samples_per_second": 17.412,
320
- "eval_steps_per_second": 2.19,
321
  "step": 2000
322
  },
323
  {
324
- "epoch": 45.55555555555556,
325
- "grad_norm": 2.019158124923706,
326
- "learning_rate": 9.755e-06,
327
- "loss": 0.4573,
328
  "step": 2050
329
  },
330
  {
331
- "epoch": 46.666666666666664,
332
- "grad_norm": 1.7636100053787231,
333
- "learning_rate": 9.505000000000001e-06,
334
- "loss": 0.4642,
335
  "step": 2100
336
  },
337
  {
338
- "epoch": 47.77777777777778,
339
- "grad_norm": 2.0878355503082275,
340
- "learning_rate": 9.260000000000001e-06,
341
- "loss": 0.454,
342
  "step": 2150
343
  },
344
  {
345
- "epoch": 48.888888888888886,
346
- "grad_norm": 1.5207270383834839,
347
- "learning_rate": 9.01e-06,
348
- "loss": 0.4573,
349
  "step": 2200
350
  },
351
  {
352
- "epoch": 50.0,
353
- "grad_norm": 2.592440605163574,
354
- "learning_rate": 8.76e-06,
355
- "loss": 0.4557,
356
  "step": 2250
357
  },
358
  {
359
- "epoch": 51.111111111111114,
360
- "grad_norm": 4.970732688903809,
361
- "learning_rate": 8.51e-06,
362
- "loss": 0.4565,
363
  "step": 2300
364
  },
365
  {
366
- "epoch": 52.22222222222222,
367
- "grad_norm": 1.2565674781799316,
368
- "learning_rate": 8.26e-06,
369
- "loss": 0.4492,
370
  "step": 2350
371
  },
372
  {
373
- "epoch": 53.333333333333336,
374
- "grad_norm": 1.402048110961914,
375
- "learning_rate": 8.010000000000001e-06,
376
- "loss": 0.4477,
377
  "step": 2400
378
  },
379
  {
380
- "epoch": 54.44444444444444,
381
- "grad_norm": 2.7496023178100586,
382
- "learning_rate": 7.76e-06,
383
- "loss": 0.4484,
384
  "step": 2450
385
  },
386
  {
387
- "epoch": 55.55555555555556,
388
- "grad_norm": 2.6707046031951904,
389
- "learning_rate": 7.510000000000001e-06,
390
- "loss": 0.4541,
391
  "step": 2500
392
  },
393
  {
394
- "epoch": 55.55555555555556,
395
- "eval_loss": 0.42134132981300354,
396
- "eval_runtime": 9.5109,
397
- "eval_samples_per_second": 16.718,
398
- "eval_steps_per_second": 2.103,
399
  "step": 2500
400
  }
401
  ],
402
  "logging_steps": 50,
403
  "max_steps": 4000,
404
  "num_input_tokens_seen": 0,
405
- "num_train_epochs": 89,
406
  "save_steps": 500,
407
  "stateful_callbacks": {
408
  "TrainerControl": {
@@ -416,7 +416,7 @@
416
  "attributes": {}
417
  }
418
  },
419
- "total_flos": 1.7199264863629728e+16,
420
  "train_batch_size": 16,
421
  "trial_name": null,
422
  "trial_params": null
 
1
  {
2
+ "best_metric": 0.41214704513549805,
3
  "best_model_checkpoint": "mikhail-panzo/fil_b32_le5_s4000/checkpoint-2500",
4
+ "epoch": 222.22222222222223,
5
  "eval_steps": 500,
6
  "global_step": 2500,
7
  "is_hyper_param_search": false,
 
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 4.444444444444445,
13
+ "grad_norm": 4.288165092468262,
14
+ "learning_rate": 2.5000000000000004e-07,
15
+ "loss": 0.7989,
16
  "step": 50
17
  },
18
  {
19
+ "epoch": 8.88888888888889,
20
+ "grad_norm": 2.4493463039398193,
21
+ "learning_rate": 5.000000000000001e-07,
22
+ "loss": 0.7701,
23
  "step": 100
24
  },
25
  {
26
+ "epoch": 13.333333333333334,
27
+ "grad_norm": 1.3798907995224,
28
+ "learning_rate": 7.5e-07,
29
+ "loss": 0.7374,
30
  "step": 150
31
  },
32
  {
33
+ "epoch": 17.77777777777778,
34
+ "grad_norm": 2.6173877716064453,
35
+ "learning_rate": 1.0000000000000002e-06,
36
+ "loss": 0.7101,
37
  "step": 200
38
  },
39
  {
40
+ "epoch": 22.22222222222222,
41
+ "grad_norm": 1.3756701946258545,
42
+ "learning_rate": 1.25e-06,
43
+ "loss": 0.6887,
44
  "step": 250
45
  },
46
  {
47
+ "epoch": 26.666666666666668,
48
+ "grad_norm": 1.0456843376159668,
49
+ "learning_rate": 1.5e-06,
50
+ "loss": 0.672,
51
  "step": 300
52
  },
53
  {
54
+ "epoch": 31.11111111111111,
55
+ "grad_norm": 0.9518398642539978,
56
+ "learning_rate": 1.75e-06,
57
+ "loss": 0.6583,
58
  "step": 350
59
  },
60
  {
61
+ "epoch": 35.55555555555556,
62
+ "grad_norm": 1.305978775024414,
63
+ "learning_rate": 2.0000000000000003e-06,
64
+ "loss": 0.6258,
65
  "step": 400
66
  },
67
  {
68
+ "epoch": 40.0,
69
+ "grad_norm": 0.8857623338699341,
70
+ "learning_rate": 2.25e-06,
71
+ "loss": 0.5835,
72
  "step": 450
73
  },
74
  {
75
+ "epoch": 44.44444444444444,
76
+ "grad_norm": 1.052182912826538,
77
+ "learning_rate": 2.5e-06,
78
+ "loss": 0.5539,
79
  "step": 500
80
  },
81
  {
82
+ "epoch": 44.44444444444444,
83
+ "eval_loss": 0.48496538400650024,
84
+ "eval_runtime": 9.4574,
85
+ "eval_samples_per_second": 16.812,
86
+ "eval_steps_per_second": 2.115,
87
  "step": 500
88
  },
89
  {
90
+ "epoch": 48.888888888888886,
91
+ "grad_norm": 1.249458909034729,
92
+ "learning_rate": 2.7500000000000004e-06,
93
+ "loss": 0.5387,
94
  "step": 550
95
  },
96
  {
97
+ "epoch": 53.333333333333336,
98
+ "grad_norm": 0.801243245601654,
99
+ "learning_rate": 3e-06,
100
+ "loss": 0.5344,
101
  "step": 600
102
  },
103
  {
104
+ "epoch": 57.77777777777778,
105
+ "grad_norm": 1.1922223567962646,
106
+ "learning_rate": 3.2500000000000002e-06,
107
+ "loss": 0.5259,
108
  "step": 650
109
  },
110
  {
111
+ "epoch": 62.22222222222222,
112
+ "grad_norm": 0.9428519606590271,
113
+ "learning_rate": 3.5e-06,
114
+ "loss": 0.5179,
115
  "step": 700
116
  },
117
  {
118
+ "epoch": 66.66666666666667,
119
+ "grad_norm": 0.755260705947876,
120
+ "learning_rate": 3.7500000000000005e-06,
121
+ "loss": 0.5082,
122
  "step": 750
123
  },
124
  {
125
+ "epoch": 71.11111111111111,
126
+ "grad_norm": 1.8587590456008911,
127
+ "learning_rate": 4.000000000000001e-06,
128
+ "loss": 0.5096,
129
  "step": 800
130
  },
131
  {
132
+ "epoch": 75.55555555555556,
133
+ "grad_norm": 0.6854829788208008,
134
+ "learning_rate": 4.25e-06,
135
+ "loss": 0.4952,
136
  "step": 850
137
  },
138
  {
139
+ "epoch": 80.0,
140
+ "grad_norm": 0.9088854789733887,
141
+ "learning_rate": 4.5e-06,
142
+ "loss": 0.4941,
143
  "step": 900
144
  },
145
  {
146
+ "epoch": 84.44444444444444,
147
+ "grad_norm": 0.8932746648788452,
148
+ "learning_rate": 4.75e-06,
149
+ "loss": 0.4936,
150
  "step": 950
151
  },
152
  {
153
+ "epoch": 88.88888888888889,
154
+ "grad_norm": 0.9161014556884766,
155
+ "learning_rate": 5e-06,
156
+ "loss": 0.4867,
157
  "step": 1000
158
  },
159
  {
160
+ "epoch": 88.88888888888889,
161
+ "eval_loss": 0.4438903331756592,
162
+ "eval_runtime": 9.2831,
163
+ "eval_samples_per_second": 17.128,
164
+ "eval_steps_per_second": 2.154,
165
  "step": 1000
166
  },
167
  {
168
+ "epoch": 93.33333333333333,
169
+ "grad_norm": 0.6858564615249634,
170
+ "learning_rate": 5.2500000000000006e-06,
171
+ "loss": 0.486,
172
  "step": 1050
173
  },
174
  {
175
+ "epoch": 97.77777777777777,
176
+ "grad_norm": 1.2040996551513672,
177
+ "learning_rate": 5.500000000000001e-06,
178
+ "loss": 0.4789,
179
  "step": 1100
180
  },
181
  {
182
+ "epoch": 102.22222222222223,
183
+ "grad_norm": 1.0067086219787598,
184
+ "learning_rate": 5.75e-06,
185
+ "loss": 0.477,
186
  "step": 1150
187
  },
188
  {
189
+ "epoch": 106.66666666666667,
190
+ "grad_norm": 1.4230589866638184,
191
+ "learning_rate": 6e-06,
192
+ "loss": 0.473,
193
  "step": 1200
194
  },
195
  {
196
+ "epoch": 111.11111111111111,
197
+ "grad_norm": 0.916904091835022,
198
+ "learning_rate": 6.25e-06,
199
+ "loss": 0.4698,
200
  "step": 1250
201
  },
202
  {
203
+ "epoch": 115.55555555555556,
204
+ "grad_norm": 0.9870907664299011,
205
+ "learning_rate": 6.5000000000000004e-06,
206
+ "loss": 0.4693,
207
  "step": 1300
208
  },
209
  {
210
+ "epoch": 120.0,
211
+ "grad_norm": 1.0495694875717163,
212
+ "learning_rate": 6.750000000000001e-06,
213
+ "loss": 0.4651,
214
  "step": 1350
215
  },
216
  {
217
+ "epoch": 124.44444444444444,
218
+ "grad_norm": 1.6615877151489258,
219
+ "learning_rate": 7e-06,
220
+ "loss": 0.4648,
221
  "step": 1400
222
  },
223
  {
224
+ "epoch": 128.88888888888889,
225
+ "grad_norm": 0.89169842004776,
226
+ "learning_rate": 7.25e-06,
227
+ "loss": 0.4643,
228
  "step": 1450
229
  },
230
  {
231
+ "epoch": 133.33333333333334,
232
+ "grad_norm": 0.8503329157829285,
233
+ "learning_rate": 7.500000000000001e-06,
234
+ "loss": 0.4595,
235
  "step": 1500
236
  },
237
  {
238
+ "epoch": 133.33333333333334,
239
+ "eval_loss": 0.4245499074459076,
240
+ "eval_runtime": 8.4972,
241
+ "eval_samples_per_second": 18.712,
242
+ "eval_steps_per_second": 2.354,
243
  "step": 1500
244
  },
245
  {
246
+ "epoch": 137.77777777777777,
247
+ "grad_norm": 0.9252403974533081,
248
+ "learning_rate": 7.75e-06,
249
+ "loss": 0.4617,
250
  "step": 1550
251
  },
252
  {
253
+ "epoch": 142.22222222222223,
254
+ "grad_norm": 0.8366873860359192,
255
+ "learning_rate": 8.000000000000001e-06,
256
+ "loss": 0.4557,
257
  "step": 1600
258
  },
259
  {
260
+ "epoch": 146.66666666666666,
261
+ "grad_norm": 0.7907727360725403,
262
+ "learning_rate": 8.25e-06,
263
+ "loss": 0.4565,
264
  "step": 1650
265
  },
266
  {
267
+ "epoch": 151.11111111111111,
268
+ "grad_norm": 0.8415725231170654,
269
+ "learning_rate": 8.5e-06,
270
+ "loss": 0.4544,
271
  "step": 1700
272
  },
273
  {
274
+ "epoch": 155.55555555555554,
275
+ "grad_norm": 0.8423948884010315,
276
+ "learning_rate": 8.750000000000001e-06,
277
+ "loss": 0.4528,
278
  "step": 1750
279
  },
280
  {
281
+ "epoch": 160.0,
282
+ "grad_norm": 0.8236134648323059,
283
+ "learning_rate": 9e-06,
284
+ "loss": 0.4555,
285
  "step": 1800
286
  },
287
  {
288
+ "epoch": 164.44444444444446,
289
+ "grad_norm": 0.8070024251937866,
290
+ "learning_rate": 9.250000000000001e-06,
291
+ "loss": 0.4474,
292
  "step": 1850
293
  },
294
  {
295
+ "epoch": 168.88888888888889,
296
+ "grad_norm": 1.041704535484314,
297
+ "learning_rate": 9.5e-06,
298
+ "loss": 0.4478,
299
  "step": 1900
300
  },
301
  {
302
+ "epoch": 173.33333333333334,
303
+ "grad_norm": 1.021083116531372,
304
+ "learning_rate": 9.75e-06,
305
+ "loss": 0.4433,
306
  "step": 1950
307
  },
308
  {
309
+ "epoch": 177.77777777777777,
310
+ "grad_norm": 1.097479224205017,
311
+ "learning_rate": 1e-05,
312
+ "loss": 0.4395,
313
  "step": 2000
314
  },
315
  {
316
+ "epoch": 177.77777777777777,
317
+ "eval_loss": 0.41554170846939087,
318
+ "eval_runtime": 9.3691,
319
+ "eval_samples_per_second": 16.971,
320
+ "eval_steps_per_second": 2.135,
321
  "step": 2000
322
  },
323
  {
324
+ "epoch": 182.22222222222223,
325
+ "grad_norm": 1.0658737421035767,
326
+ "learning_rate": 9.75e-06,
327
+ "loss": 0.4403,
328
  "step": 2050
329
  },
330
  {
331
+ "epoch": 186.66666666666666,
332
+ "grad_norm": 0.7741203308105469,
333
+ "learning_rate": 9.5e-06,
334
+ "loss": 0.4371,
335
  "step": 2100
336
  },
337
  {
338
+ "epoch": 191.11111111111111,
339
+ "grad_norm": 0.8409827947616577,
340
+ "learning_rate": 9.250000000000001e-06,
341
+ "loss": 0.4383,
342
  "step": 2150
343
  },
344
  {
345
+ "epoch": 195.55555555555554,
346
+ "grad_norm": 0.7340779900550842,
347
+ "learning_rate": 9e-06,
348
+ "loss": 0.4367,
349
  "step": 2200
350
  },
351
  {
352
+ "epoch": 200.0,
353
+ "grad_norm": 1.0130568742752075,
354
+ "learning_rate": 8.750000000000001e-06,
355
+ "loss": 0.4364,
356
  "step": 2250
357
  },
358
  {
359
+ "epoch": 204.44444444444446,
360
+ "grad_norm": 0.7156779170036316,
361
+ "learning_rate": 8.5e-06,
362
+ "loss": 0.4337,
363
  "step": 2300
364
  },
365
  {
366
+ "epoch": 208.88888888888889,
367
+ "grad_norm": 1.1859618425369263,
368
+ "learning_rate": 8.25e-06,
369
+ "loss": 0.4342,
370
  "step": 2350
371
  },
372
  {
373
+ "epoch": 213.33333333333334,
374
+ "grad_norm": 0.6573166847229004,
375
+ "learning_rate": 8.000000000000001e-06,
376
+ "loss": 0.432,
377
  "step": 2400
378
  },
379
  {
380
+ "epoch": 217.77777777777777,
381
+ "grad_norm": 1.0668668746948242,
382
+ "learning_rate": 7.75e-06,
383
+ "loss": 0.4338,
384
  "step": 2450
385
  },
386
  {
387
+ "epoch": 222.22222222222223,
388
+ "grad_norm": 0.8622764945030212,
389
+ "learning_rate": 7.500000000000001e-06,
390
+ "loss": 0.4327,
391
  "step": 2500
392
  },
393
  {
394
+ "epoch": 222.22222222222223,
395
+ "eval_loss": 0.41214704513549805,
396
+ "eval_runtime": 9.4587,
397
+ "eval_samples_per_second": 16.81,
398
+ "eval_steps_per_second": 2.114,
399
  "step": 2500
400
  }
401
  ],
402
  "logging_steps": 50,
403
  "max_steps": 4000,
404
  "num_input_tokens_seen": 0,
405
+ "num_train_epochs": 364,
406
  "save_steps": 500,
407
  "stateful_callbacks": {
408
  "TrainerControl": {
 
416
  "attributes": {}
417
  }
418
  },
419
+ "total_flos": 6.872445306432288e+16,
420
  "train_batch_size": 16,
421
  "trial_name": null,
422
  "trial_params": null
checkpoint-2500/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:05b7ecdded947540b9670cedf7cfe87afc7199b237efb642b915ffbb505d51dc
3
  size 5304
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2de69a241d012253ce4b1526ecd0f9fd0f7dfbedfea74bfff08144d42d9f523a
3
  size 5304