Blancy commited on
Commit
f212069
·
verified ·
1 Parent(s): d7b3b5c

Model save

Browse files
Files changed (5) hide show
  1. README.md +2 -2
  2. all_results.json +6 -6
  3. model.safetensors +1 -1
  4. train_results.json +6 -6
  5. trainer_state.json +641 -221
README.md CHANGED
@@ -27,14 +27,14 @@ print(output["generated_text"])
27
 
28
  ## Training procedure
29
 
30
- [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/224015062-chinese-university-of-hong-kong-shenzhen/huggingface/runs/7ehmga5f)
31
 
32
 
33
  This model was trained with SFT.
34
 
35
  ### Framework versions
36
 
37
- - TRL: 0.16.0.dev0
38
  - Transformers: 4.49.0
39
  - Pytorch: 2.5.1
40
  - Datasets: 3.5.0
 
27
 
28
  ## Training procedure
29
 
30
+ [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/224015062-chinese-university-of-hong-kong-shenzhen/huggingface/runs/atj7im6p)
31
 
32
 
33
  This model was trained with SFT.
34
 
35
  ### Framework versions
36
 
37
+ - TRL: 0.15.2
38
  - Transformers: 4.49.0
39
  - Pytorch: 2.5.1
40
  - Datasets: 3.5.0
all_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "total_flos": 4.4183709150420664e+18,
3
- "train_loss": 0.07813258930595952,
4
- "train_runtime": 6469.3467,
5
- "train_samples": 93733,
6
- "train_samples_per_second": 5.302,
7
- "train_steps_per_second": 0.041
8
  }
 
1
  {
2
+ "total_flos": 5.828408442271826e+17,
3
+ "train_loss": 0.10050818478102819,
4
+ "train_runtime": 2227.4528,
5
+ "train_samples": 10000,
6
+ "train_samples_per_second": 32.502,
7
+ "train_steps_per_second": 0.254
8
  }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:dd57186f0f3c3d60c23b32b82bb722b21fff3a9ce7706b12c3ef34c640da7c4d
3
  size 3087467144
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5d1fa244f2edfa87d04dc19f1c8464d82f54dcf5b0ace83e087b6dcc1ee87df0
3
  size 3087467144
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "total_flos": 4.4183709150420664e+18,
3
- "train_loss": 0.07813258930595952,
4
- "train_runtime": 6469.3467,
5
- "train_samples": 93733,
6
- "train_samples_per_second": 5.302,
7
- "train_steps_per_second": 0.041
8
  }
 
1
  {
2
+ "total_flos": 5.828408442271826e+17,
3
+ "train_loss": 0.10050818478102819,
4
+ "train_runtime": 2227.4528,
5
+ "train_samples": 10000,
6
+ "train_samples_per_second": 32.502,
7
+ "train_steps_per_second": 0.254
8
  }
trainer_state.json CHANGED
@@ -3,394 +3,814 @@
3
  "best_model_checkpoint": null,
4
  "epoch": 1.0,
5
  "eval_steps": 500,
6
- "global_step": 268,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 0.018656716417910446,
13
- "grad_norm": 0.169921875,
14
- "learning_rate": 1.785714285714286e-05,
15
- "loss": 0.1062,
16
  "step": 5
17
  },
18
  {
19
- "epoch": 0.03731343283582089,
20
- "grad_norm": 0.126953125,
21
- "learning_rate": 3.571428571428572e-05,
22
- "loss": 0.0978,
23
  "step": 10
24
  },
25
  {
26
- "epoch": 0.055970149253731345,
27
- "grad_norm": 0.07666015625,
28
- "learning_rate": 4.999827900623038e-05,
29
- "loss": 0.0911,
30
  "step": 15
31
  },
32
  {
33
- "epoch": 0.07462686567164178,
34
- "grad_norm": 0.05322265625,
35
- "learning_rate": 4.993807186343243e-05,
36
- "loss": 0.0873,
37
  "step": 20
38
  },
39
  {
40
- "epoch": 0.09328358208955224,
41
- "grad_norm": 0.039794921875,
42
- "learning_rate": 4.979207812402531e-05,
43
- "loss": 0.0838,
44
  "step": 25
45
  },
46
  {
47
- "epoch": 0.11194029850746269,
48
- "grad_norm": 0.030029296875,
49
- "learning_rate": 4.956085596012407e-05,
50
- "loss": 0.0832,
51
  "step": 30
52
  },
53
  {
54
- "epoch": 0.13059701492537312,
55
- "grad_norm": 0.0252685546875,
56
- "learning_rate": 4.924528939432311e-05,
57
- "loss": 0.0814,
58
  "step": 35
59
  },
60
  {
61
- "epoch": 0.14925373134328357,
62
- "grad_norm": 0.021484375,
63
- "learning_rate": 4.884658491984735e-05,
64
- "loss": 0.0806,
65
  "step": 40
66
  },
67
  {
68
- "epoch": 0.16791044776119404,
69
- "grad_norm": 0.020751953125,
70
- "learning_rate": 4.8366266887814235e-05,
71
- "loss": 0.0809,
72
  "step": 45
73
  },
74
  {
75
- "epoch": 0.1865671641791045,
76
- "grad_norm": 0.01953125,
77
- "learning_rate": 4.780617167924209e-05,
78
- "loss": 0.0788,
79
  "step": 50
80
  },
81
  {
82
- "epoch": 0.20522388059701493,
83
- "grad_norm": 0.0205078125,
84
- "learning_rate": 4.716844068408693e-05,
85
- "loss": 0.0793,
86
  "step": 55
87
  },
88
  {
89
- "epoch": 0.22388059701492538,
90
- "grad_norm": 0.019775390625,
91
- "learning_rate": 4.6455512114150546e-05,
92
- "loss": 0.0789,
93
  "step": 60
94
  },
95
  {
96
- "epoch": 0.24253731343283583,
97
- "grad_norm": 0.01904296875,
98
- "learning_rate": 4.5670111681161296e-05,
99
- "loss": 0.0778,
100
  "step": 65
101
  },
102
  {
103
- "epoch": 0.26119402985074625,
104
- "grad_norm": 0.01953125,
105
- "learning_rate": 4.481524217566783e-05,
106
- "loss": 0.0775,
107
  "step": 70
108
  },
109
  {
110
- "epoch": 0.2798507462686567,
111
- "grad_norm": 0.01953125,
112
- "learning_rate": 4.3894171986588217e-05,
113
- "loss": 0.0773,
114
  "step": 75
115
  },
116
  {
117
- "epoch": 0.29850746268656714,
118
- "grad_norm": 0.0194091796875,
119
- "learning_rate": 4.29104226053073e-05,
120
- "loss": 0.0773,
121
  "step": 80
122
  },
123
  {
124
- "epoch": 0.31716417910447764,
125
- "grad_norm": 0.0184326171875,
126
- "learning_rate": 4.186775516209732e-05,
127
- "loss": 0.0768,
128
  "step": 85
129
  },
130
  {
131
- "epoch": 0.3358208955223881,
132
- "grad_norm": 0.0194091796875,
133
- "learning_rate": 4.077015604633669e-05,
134
- "loss": 0.0773,
135
  "step": 90
136
  },
137
  {
138
- "epoch": 0.35447761194029853,
139
- "grad_norm": 0.0191650390625,
140
- "learning_rate": 3.962182166550441e-05,
141
- "loss": 0.0772,
142
  "step": 95
143
  },
144
  {
145
- "epoch": 0.373134328358209,
146
- "grad_norm": 0.0185546875,
147
- "learning_rate": 3.8427142401220634e-05,
148
- "loss": 0.0768,
149
  "step": 100
150
  },
151
  {
152
- "epoch": 0.3917910447761194,
153
- "grad_norm": 0.0186767578125,
154
- "learning_rate": 3.71906858236735e-05,
155
- "loss": 0.0764,
156
  "step": 105
157
  },
158
  {
159
- "epoch": 0.41044776119402987,
160
- "grad_norm": 0.0179443359375,
161
- "learning_rate": 3.591717922860785e-05,
162
- "loss": 0.0773,
163
  "step": 110
164
  },
165
  {
166
- "epoch": 0.4291044776119403,
167
- "grad_norm": 0.017822265625,
168
- "learning_rate": 3.46114915636416e-05,
169
- "loss": 0.0762,
170
  "step": 115
171
  },
172
  {
173
- "epoch": 0.44776119402985076,
174
- "grad_norm": 0.0194091796875,
175
- "learning_rate": 3.3278614813010034e-05,
176
- "loss": 0.0765,
177
  "step": 120
178
  },
179
  {
180
- "epoch": 0.4664179104477612,
181
- "grad_norm": 0.0179443359375,
182
- "learning_rate": 3.1923644911909e-05,
183
- "loss": 0.0762,
184
  "step": 125
185
  },
186
  {
187
- "epoch": 0.48507462686567165,
188
- "grad_norm": 0.017578125,
189
- "learning_rate": 3.0551762263406576e-05,
190
- "loss": 0.0761,
191
  "step": 130
192
  },
193
  {
194
- "epoch": 0.503731343283582,
195
- "grad_norm": 0.0186767578125,
196
- "learning_rate": 2.9168211932412042e-05,
197
- "loss": 0.0758,
198
  "step": 135
199
  },
200
  {
201
- "epoch": 0.5223880597014925,
202
- "grad_norm": 0.0174560546875,
203
- "learning_rate": 2.777828359242567e-05,
204
- "loss": 0.0766,
205
  "step": 140
206
  },
207
  {
208
- "epoch": 0.5410447761194029,
209
- "grad_norm": 0.0179443359375,
210
- "learning_rate": 2.6387291301738377e-05,
211
- "loss": 0.0756,
212
  "step": 145
213
  },
214
  {
215
- "epoch": 0.5597014925373134,
216
- "grad_norm": 0.01708984375,
217
- "learning_rate": 2.50005531864019e-05,
218
- "loss": 0.0755,
219
  "step": 150
220
  },
221
  {
222
- "epoch": 0.5783582089552238,
223
- "grad_norm": 0.0169677734375,
224
- "learning_rate": 2.362337110764688e-05,
225
- "loss": 0.0755,
226
  "step": 155
227
  },
228
  {
229
- "epoch": 0.5970149253731343,
230
- "grad_norm": 0.0181884765625,
231
- "learning_rate": 2.226101039148557e-05,
232
- "loss": 0.0754,
233
  "step": 160
234
  },
235
  {
236
- "epoch": 0.6156716417910447,
237
- "grad_norm": 0.0169677734375,
238
- "learning_rate": 2.0918679697998252e-05,
239
- "loss": 0.0753,
240
  "step": 165
241
  },
242
  {
243
- "epoch": 0.6343283582089553,
244
- "grad_norm": 0.017578125,
245
- "learning_rate": 1.9601511107268255e-05,
246
- "loss": 0.0755,
247
  "step": 170
248
  },
249
  {
250
- "epoch": 0.6529850746268657,
251
- "grad_norm": 0.0181884765625,
252
- "learning_rate": 1.8314540498102216e-05,
253
- "loss": 0.0755,
254
  "step": 175
255
  },
256
  {
257
- "epoch": 0.6716417910447762,
258
- "grad_norm": 0.016357421875,
259
- "learning_rate": 1.7062688294552992e-05,
260
- "loss": 0.0745,
261
  "step": 180
262
  },
263
  {
264
- "epoch": 0.6902985074626866,
265
- "grad_norm": 0.0167236328125,
266
- "learning_rate": 1.5850740653856096e-05,
267
- "loss": 0.075,
268
  "step": 185
269
  },
270
  {
271
- "epoch": 0.7089552238805971,
272
- "grad_norm": 0.0172119140625,
273
- "learning_rate": 1.4683331167703218e-05,
274
- "loss": 0.0756,
275
  "step": 190
276
  },
277
  {
278
- "epoch": 0.7276119402985075,
279
- "grad_norm": 0.016845703125,
280
- "learning_rate": 1.356492314681356e-05,
281
- "loss": 0.076,
282
  "step": 195
283
  },
284
  {
285
- "epoch": 0.746268656716418,
286
- "grad_norm": 0.0169677734375,
287
- "learning_rate": 1.2499792556533716e-05,
288
- "loss": 0.0753,
289
  "step": 200
290
  },
291
  {
292
- "epoch": 0.7649253731343284,
293
- "grad_norm": 0.0167236328125,
294
- "learning_rate": 1.1492011668707753e-05,
295
- "loss": 0.075,
296
  "step": 205
297
  },
298
  {
299
- "epoch": 0.7835820895522388,
300
- "grad_norm": 0.0169677734375,
301
- "learning_rate": 1.0545433492320603e-05,
302
- "loss": 0.0757,
303
  "step": 210
304
  },
305
  {
306
- "epoch": 0.8022388059701493,
307
- "grad_norm": 0.0167236328125,
308
- "learning_rate": 9.663677042440537e-06,
309
- "loss": 0.075,
310
  "step": 215
311
  },
312
  {
313
- "epoch": 0.8208955223880597,
314
- "grad_norm": 0.0167236328125,
315
- "learning_rate": 8.850113503781367e-06,
316
- "loss": 0.0751,
317
  "step": 220
318
  },
319
  {
320
- "epoch": 0.8395522388059702,
321
- "grad_norm": 0.0167236328125,
322
- "learning_rate": 8.107853341784671e-06,
323
- "loss": 0.076,
324
  "step": 225
325
  },
326
  {
327
- "epoch": 0.8582089552238806,
328
- "grad_norm": 0.01708984375,
329
- "learning_rate": 7.439734410499752e-06,
330
- "loss": 0.0755,
331
  "step": 230
332
  },
333
  {
334
- "epoch": 0.8768656716417911,
335
- "grad_norm": 0.016357421875,
336
- "learning_rate": 6.848311102728011e-06,
337
- "loss": 0.0756,
338
  "step": 235
339
  },
340
  {
341
- "epoch": 0.8955223880597015,
342
- "grad_norm": 0.0167236328125,
343
- "learning_rate": 6.335844583913515e-06,
344
- "loss": 0.0751,
345
  "step": 240
346
  },
347
  {
348
- "epoch": 0.914179104477612,
349
- "grad_norm": 0.016357421875,
350
- "learning_rate": 5.904294147118193e-06,
351
- "loss": 0.0757,
352
  "step": 245
353
  },
354
  {
355
- "epoch": 0.9328358208955224,
356
- "grad_norm": 0.0164794921875,
357
- "learning_rate": 5.555309722133842e-06,
358
- "loss": 0.0753,
359
  "step": 250
360
  },
361
  {
362
- "epoch": 0.9514925373134329,
363
- "grad_norm": 0.0166015625,
364
- "learning_rate": 5.290225567370509e-06,
365
- "loss": 0.0747,
366
  "step": 255
367
  },
368
  {
369
- "epoch": 0.9701492537313433,
370
- "grad_norm": 0.016845703125,
371
- "learning_rate": 5.110055168638854e-06,
372
- "loss": 0.0753,
373
  "step": 260
374
  },
375
  {
376
- "epoch": 0.9888059701492538,
377
- "grad_norm": 0.016845703125,
378
- "learning_rate": 5.0154873643297575e-06,
379
- "loss": 0.0757,
380
  "step": 265
381
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
382
  {
383
  "epoch": 1.0,
384
- "step": 268,
385
- "total_flos": 4.4183709150420664e+18,
386
- "train_loss": 0.07813258930595952,
387
- "train_runtime": 6469.3467,
388
- "train_samples_per_second": 5.302,
389
- "train_steps_per_second": 0.041
390
  }
391
  ],
392
  "logging_steps": 5,
393
- "max_steps": 268,
394
  "num_input_tokens_seen": 0,
395
  "num_train_epochs": 1,
396
  "save_steps": 100,
@@ -406,7 +826,7 @@
406
  "attributes": {}
407
  }
408
  },
409
- "total_flos": 4.4183709150420664e+18,
410
  "train_batch_size": 128,
411
  "trial_name": null,
412
  "trial_params": null
 
3
  "best_model_checkpoint": null,
4
  "epoch": 1.0,
5
  "eval_steps": 500,
6
+ "global_step": 566,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 0.0088339222614841,
13
+ "grad_norm": 0.318359375,
14
+ "learning_rate": 8.620689655172414e-06,
15
+ "loss": 0.145,
16
  "step": 5
17
  },
18
  {
19
+ "epoch": 0.0176678445229682,
20
+ "grad_norm": 0.1845703125,
21
+ "learning_rate": 1.7241379310344828e-05,
22
+ "loss": 0.1351,
23
  "step": 10
24
  },
25
  {
26
+ "epoch": 0.026501766784452298,
27
+ "grad_norm": 0.1513671875,
28
+ "learning_rate": 2.5862068965517244e-05,
29
+ "loss": 0.1249,
30
  "step": 15
31
  },
32
  {
33
+ "epoch": 0.0353356890459364,
34
+ "grad_norm": 0.12109375,
35
+ "learning_rate": 3.4482758620689657e-05,
36
+ "loss": 0.1189,
37
  "step": 20
38
  },
39
  {
40
+ "epoch": 0.044169611307420496,
41
+ "grad_norm": 0.09619140625,
42
+ "learning_rate": 4.3103448275862066e-05,
43
+ "loss": 0.1181,
44
  "step": 25
45
  },
46
  {
47
+ "epoch": 0.053003533568904596,
48
+ "grad_norm": 0.0908203125,
49
+ "learning_rate": 4.999961496300632e-05,
50
+ "loss": 0.1127,
51
  "step": 30
52
  },
53
  {
54
+ "epoch": 0.061837455830388695,
55
+ "grad_norm": 0.0830078125,
56
+ "learning_rate": 4.9986140051876094e-05,
57
+ "loss": 0.1102,
58
  "step": 35
59
  },
60
  {
61
+ "epoch": 0.0706713780918728,
62
+ "grad_norm": 0.08447265625,
63
+ "learning_rate": 4.995342646712217e-05,
64
+ "loss": 0.1093,
65
  "step": 40
66
  },
67
  {
68
+ "epoch": 0.07950530035335689,
69
+ "grad_norm": 0.07958984375,
70
+ "learning_rate": 4.9901502197807084e-05,
71
+ "loss": 0.109,
72
  "step": 45
73
  },
74
  {
75
+ "epoch": 0.08833922261484099,
76
+ "grad_norm": 0.07958984375,
77
+ "learning_rate": 4.9830411669255416e-05,
78
+ "loss": 0.107,
79
  "step": 50
80
  },
81
  {
82
+ "epoch": 0.09717314487632508,
83
+ "grad_norm": 0.07763671875,
84
+ "learning_rate": 4.974021570504443e-05,
85
+ "loss": 0.1041,
86
  "step": 55
87
  },
88
  {
89
+ "epoch": 0.10600706713780919,
90
+ "grad_norm": 0.07421875,
91
+ "learning_rate": 4.963099147496465e-05,
92
+ "loss": 0.1029,
93
  "step": 60
94
  },
95
  {
96
+ "epoch": 0.11484098939929328,
97
+ "grad_norm": 0.0810546875,
98
+ "learning_rate": 4.9502832428995005e-05,
99
+ "loss": 0.1047,
100
  "step": 65
101
  },
102
  {
103
+ "epoch": 0.12367491166077739,
104
+ "grad_norm": 0.0771484375,
105
+ "learning_rate": 4.935584821734901e-05,
106
+ "loss": 0.1044,
107
  "step": 70
108
  },
109
  {
110
+ "epoch": 0.13250883392226148,
111
+ "grad_norm": 0.07568359375,
112
+ "learning_rate": 4.919016459666026e-05,
113
+ "loss": 0.1063,
114
  "step": 75
115
  },
116
  {
117
+ "epoch": 0.1413427561837456,
118
+ "grad_norm": 0.0830078125,
119
+ "learning_rate": 4.9005923322387706e-05,
120
+ "loss": 0.1025,
121
  "step": 80
122
  },
123
  {
124
+ "epoch": 0.1501766784452297,
125
+ "grad_norm": 0.0859375,
126
+ "learning_rate": 4.880328202753264e-05,
127
+ "loss": 0.1022,
128
  "step": 85
129
  },
130
  {
131
+ "epoch": 0.15901060070671377,
132
+ "grad_norm": 0.0791015625,
133
+ "learning_rate": 4.858241408777117e-05,
134
+ "loss": 0.1031,
135
  "step": 90
136
  },
137
  {
138
+ "epoch": 0.16784452296819788,
139
+ "grad_norm": 0.07421875,
140
+ "learning_rate": 4.834350847311758e-05,
141
+ "loss": 0.1016,
142
  "step": 95
143
  },
144
  {
145
+ "epoch": 0.17667844522968199,
146
+ "grad_norm": 0.076171875,
147
+ "learning_rate": 4.8086769586245554e-05,
148
+ "loss": 0.1008,
149
  "step": 100
150
  },
151
  {
152
+ "epoch": 0.1855123674911661,
153
+ "grad_norm": 0.0751953125,
154
+ "learning_rate": 4.7812417087605456e-05,
155
+ "loss": 0.1043,
156
  "step": 105
157
  },
158
  {
159
+ "epoch": 0.19434628975265017,
160
+ "grad_norm": 0.07275390625,
161
+ "learning_rate": 4.752068570748746e-05,
162
+ "loss": 0.1038,
163
  "step": 110
164
  },
165
  {
166
+ "epoch": 0.20318021201413428,
167
+ "grad_norm": 0.078125,
168
+ "learning_rate": 4.721182504519118e-05,
169
+ "loss": 0.1014,
170
  "step": 115
171
  },
172
  {
173
+ "epoch": 0.21201413427561838,
174
+ "grad_norm": 0.0712890625,
175
+ "learning_rate": 4.688609935547371e-05,
176
+ "loss": 0.1004,
177
  "step": 120
178
  },
179
  {
180
+ "epoch": 0.22084805653710246,
181
+ "grad_norm": 0.07177734375,
182
+ "learning_rate": 4.654378732245869e-05,
183
+ "loss": 0.1017,
184
  "step": 125
185
  },
186
  {
187
+ "epoch": 0.22968197879858657,
188
+ "grad_norm": 0.07421875,
189
+ "learning_rate": 4.618518182120011e-05,
190
+ "loss": 0.1006,
191
  "step": 130
192
  },
193
  {
194
+ "epoch": 0.23851590106007067,
195
+ "grad_norm": 0.0751953125,
196
+ "learning_rate": 4.5810589667104347e-05,
197
+ "loss": 0.1008,
198
  "step": 135
199
  },
200
  {
201
+ "epoch": 0.24734982332155478,
202
+ "grad_norm": 0.08447265625,
203
+ "learning_rate": 4.542033135342537e-05,
204
+ "loss": 0.1012,
205
  "step": 140
206
  },
207
  {
208
+ "epoch": 0.25618374558303886,
209
+ "grad_norm": 0.07275390625,
210
+ "learning_rate": 4.5014740777057405e-05,
211
+ "loss": 0.1026,
212
  "step": 145
213
  },
214
  {
215
+ "epoch": 0.26501766784452296,
216
+ "grad_norm": 0.07861328125,
217
+ "learning_rate": 4.45941649528596e-05,
218
+ "loss": 0.1017,
219
  "step": 150
220
  },
221
  {
222
+ "epoch": 0.27385159010600707,
223
+ "grad_norm": 0.0771484375,
224
+ "learning_rate": 4.4158963716757444e-05,
225
+ "loss": 0.099,
226
  "step": 155
227
  },
228
  {
229
+ "epoch": 0.2826855123674912,
230
+ "grad_norm": 0.07666015625,
231
+ "learning_rate": 4.370950941787456e-05,
232
+ "loss": 0.1021,
233
  "step": 160
234
  },
235
  {
236
+ "epoch": 0.2915194346289753,
237
+ "grad_norm": 0.0712890625,
238
+ "learning_rate": 4.324618659995855e-05,
239
+ "loss": 0.1006,
240
  "step": 165
241
  },
242
  {
243
+ "epoch": 0.3003533568904594,
244
+ "grad_norm": 0.076171875,
245
+ "learning_rate": 4.27693916723734e-05,
246
+ "loss": 0.1006,
247
  "step": 170
248
  },
249
  {
250
+ "epoch": 0.30918727915194344,
251
+ "grad_norm": 0.07275390625,
252
+ "learning_rate": 4.227953257093985e-05,
253
+ "loss": 0.0995,
254
  "step": 175
255
  },
256
  {
257
+ "epoch": 0.31802120141342755,
258
+ "grad_norm": 0.07763671875,
259
+ "learning_rate": 4.1777028408913985e-05,
260
+ "loss": 0.1005,
261
  "step": 180
262
  },
263
  {
264
+ "epoch": 0.32685512367491165,
265
+ "grad_norm": 0.07373046875,
266
+ "learning_rate": 4.126230911840269e-05,
267
+ "loss": 0.0995,
268
  "step": 185
269
  },
270
  {
271
+ "epoch": 0.33568904593639576,
272
+ "grad_norm": 0.07666015625,
273
+ "learning_rate": 4.07358150825226e-05,
274
+ "loss": 0.0984,
275
  "step": 190
276
  },
277
  {
278
+ "epoch": 0.34452296819787986,
279
+ "grad_norm": 0.07568359375,
280
+ "learning_rate": 4.0197996758617594e-05,
281
+ "loss": 0.0979,
282
  "step": 195
283
  },
284
  {
285
+ "epoch": 0.35335689045936397,
286
+ "grad_norm": 0.076171875,
287
+ "learning_rate": 3.964931429285675e-05,
288
+ "loss": 0.0998,
289
  "step": 200
290
  },
291
  {
292
+ "epoch": 0.3621908127208481,
293
+ "grad_norm": 0.07958984375,
294
+ "learning_rate": 3.909023712654291e-05,
295
+ "loss": 0.1012,
296
  "step": 205
297
  },
298
  {
299
+ "epoch": 0.3710247349823322,
300
+ "grad_norm": 0.07373046875,
301
+ "learning_rate": 3.852124359446845e-05,
302
+ "loss": 0.0987,
303
  "step": 210
304
  },
305
  {
306
+ "epoch": 0.37985865724381623,
307
+ "grad_norm": 0.0732421875,
308
+ "learning_rate": 3.794282051566199e-05,
309
+ "loss": 0.0982,
310
  "step": 215
311
  },
312
  {
313
+ "epoch": 0.38869257950530034,
314
+ "grad_norm": 0.0751953125,
315
+ "learning_rate": 3.7355462776876184e-05,
316
+ "loss": 0.0984,
317
  "step": 220
318
  },
319
  {
320
+ "epoch": 0.39752650176678445,
321
+ "grad_norm": 0.06982421875,
322
+ "learning_rate": 3.6759672909172846e-05,
323
+ "loss": 0.0973,
324
  "step": 225
325
  },
326
  {
327
+ "epoch": 0.40636042402826855,
328
+ "grad_norm": 0.07080078125,
329
+ "learning_rate": 3.615596065796791e-05,
330
+ "loss": 0.1007,
331
  "step": 230
332
  },
333
  {
334
+ "epoch": 0.41519434628975266,
335
+ "grad_norm": 0.07421875,
336
+ "learning_rate": 3.554484254690379e-05,
337
+ "loss": 0.0972,
338
  "step": 235
339
  },
340
  {
341
+ "epoch": 0.42402826855123676,
342
+ "grad_norm": 0.07080078125,
343
+ "learning_rate": 3.492684143592252e-05,
344
+ "loss": 0.0974,
345
  "step": 240
346
  },
347
  {
348
+ "epoch": 0.43286219081272087,
349
+ "grad_norm": 0.076171875,
350
+ "learning_rate": 3.4302486073917686e-05,
351
+ "loss": 0.0991,
352
  "step": 245
353
  },
354
  {
355
+ "epoch": 0.4416961130742049,
356
+ "grad_norm": 0.076171875,
357
+ "learning_rate": 3.3672310646347844e-05,
358
+ "loss": 0.0979,
359
  "step": 250
360
  },
361
  {
362
+ "epoch": 0.450530035335689,
363
+ "grad_norm": 0.07275390625,
364
+ "learning_rate": 3.3036854318198575e-05,
365
+ "loss": 0.0987,
366
  "step": 255
367
  },
368
  {
369
+ "epoch": 0.45936395759717313,
370
+ "grad_norm": 0.07470703125,
371
+ "learning_rate": 3.2396660772684114e-05,
372
+ "loss": 0.0999,
373
  "step": 260
374
  },
375
  {
376
+ "epoch": 0.46819787985865724,
377
+ "grad_norm": 0.07373046875,
378
+ "learning_rate": 3.1752277746083325e-05,
379
+ "loss": 0.0979,
380
  "step": 265
381
  },
382
+ {
383
+ "epoch": 0.47703180212014135,
384
+ "grad_norm": 0.07373046875,
385
+ "learning_rate": 3.110425655910795e-05,
386
+ "loss": 0.0983,
387
+ "step": 270
388
+ },
389
+ {
390
+ "epoch": 0.48586572438162545,
391
+ "grad_norm": 0.07080078125,
392
+ "learning_rate": 3.045315164520405e-05,
393
+ "loss": 0.0981,
394
+ "step": 275
395
+ },
396
+ {
397
+ "epoch": 0.49469964664310956,
398
+ "grad_norm": 0.07568359375,
399
+ "learning_rate": 2.9799520076190268e-05,
400
+ "loss": 0.0987,
401
+ "step": 280
402
+ },
403
+ {
404
+ "epoch": 0.5035335689045937,
405
+ "grad_norm": 0.07080078125,
406
+ "learning_rate": 2.914392108563883e-05,
407
+ "loss": 0.0963,
408
+ "step": 285
409
+ },
410
+ {
411
+ "epoch": 0.5123674911660777,
412
+ "grad_norm": 0.07470703125,
413
+ "learning_rate": 2.848691559040687e-05,
414
+ "loss": 0.0977,
415
+ "step": 290
416
+ },
417
+ {
418
+ "epoch": 0.5212014134275619,
419
+ "grad_norm": 0.072265625,
420
+ "learning_rate": 2.7829065710727682e-05,
421
+ "loss": 0.0959,
422
+ "step": 295
423
+ },
424
+ {
425
+ "epoch": 0.5300353356890459,
426
+ "grad_norm": 0.07275390625,
427
+ "learning_rate": 2.7170934289272327e-05,
428
+ "loss": 0.0983,
429
+ "step": 300
430
+ },
431
+ {
432
+ "epoch": 0.5388692579505301,
433
+ "grad_norm": 0.07177734375,
434
+ "learning_rate": 2.6513084409593137e-05,
435
+ "loss": 0.0981,
436
+ "step": 305
437
+ },
438
+ {
439
+ "epoch": 0.5477031802120141,
440
+ "grad_norm": 0.07373046875,
441
+ "learning_rate": 2.585607891436118e-05,
442
+ "loss": 0.0972,
443
+ "step": 310
444
+ },
445
+ {
446
+ "epoch": 0.5565371024734982,
447
+ "grad_norm": 0.078125,
448
+ "learning_rate": 2.5200479923809738e-05,
449
+ "loss": 0.0964,
450
+ "step": 315
451
+ },
452
+ {
453
+ "epoch": 0.5653710247349824,
454
+ "grad_norm": 0.06787109375,
455
+ "learning_rate": 2.4546848354795954e-05,
456
+ "loss": 0.0966,
457
+ "step": 320
458
+ },
459
+ {
460
+ "epoch": 0.5742049469964664,
461
+ "grad_norm": 0.068359375,
462
+ "learning_rate": 2.3895743440892053e-05,
463
+ "loss": 0.0983,
464
+ "step": 325
465
+ },
466
+ {
467
+ "epoch": 0.5830388692579506,
468
+ "grad_norm": 0.0712890625,
469
+ "learning_rate": 2.3247722253916677e-05,
470
+ "loss": 0.0983,
471
+ "step": 330
472
+ },
473
+ {
474
+ "epoch": 0.5918727915194346,
475
+ "grad_norm": 0.07177734375,
476
+ "learning_rate": 2.2603339227315902e-05,
477
+ "loss": 0.0982,
478
+ "step": 335
479
+ },
480
+ {
481
+ "epoch": 0.6007067137809188,
482
+ "grad_norm": 0.06787109375,
483
+ "learning_rate": 2.1963145681801434e-05,
484
+ "loss": 0.0968,
485
+ "step": 340
486
+ },
487
+ {
488
+ "epoch": 0.6095406360424028,
489
+ "grad_norm": 0.0712890625,
490
+ "learning_rate": 2.132768935365215e-05,
491
+ "loss": 0.0976,
492
+ "step": 345
493
+ },
494
+ {
495
+ "epoch": 0.6183745583038869,
496
+ "grad_norm": 0.0693359375,
497
+ "learning_rate": 2.069751392608232e-05,
498
+ "loss": 0.0974,
499
+ "step": 350
500
+ },
501
+ {
502
+ "epoch": 0.627208480565371,
503
+ "grad_norm": 0.0712890625,
504
+ "learning_rate": 2.0073158564077483e-05,
505
+ "loss": 0.0992,
506
+ "step": 355
507
+ },
508
+ {
509
+ "epoch": 0.6360424028268551,
510
+ "grad_norm": 0.0703125,
511
+ "learning_rate": 1.9455157453096225e-05,
512
+ "loss": 0.0992,
513
+ "step": 360
514
+ },
515
+ {
516
+ "epoch": 0.6448763250883393,
517
+ "grad_norm": 0.06884765625,
518
+ "learning_rate": 1.8844039342032095e-05,
519
+ "loss": 0.0961,
520
+ "step": 365
521
+ },
522
+ {
523
+ "epoch": 0.6537102473498233,
524
+ "grad_norm": 0.06884765625,
525
+ "learning_rate": 1.8240327090827153e-05,
526
+ "loss": 0.097,
527
+ "step": 370
528
+ },
529
+ {
530
+ "epoch": 0.6625441696113075,
531
+ "grad_norm": 0.068359375,
532
+ "learning_rate": 1.764453722312383e-05,
533
+ "loss": 0.0979,
534
+ "step": 375
535
+ },
536
+ {
537
+ "epoch": 0.6713780918727915,
538
+ "grad_norm": 0.0693359375,
539
+ "learning_rate": 1.705717948433801e-05,
540
+ "loss": 0.0963,
541
+ "step": 380
542
+ },
543
+ {
544
+ "epoch": 0.6802120141342756,
545
+ "grad_norm": 0.06787109375,
546
+ "learning_rate": 1.6478756405531564e-05,
547
+ "loss": 0.0969,
548
+ "step": 385
549
+ },
550
+ {
551
+ "epoch": 0.6890459363957597,
552
+ "grad_norm": 0.0673828125,
553
+ "learning_rate": 1.5909762873457096e-05,
554
+ "loss": 0.0963,
555
+ "step": 390
556
+ },
557
+ {
558
+ "epoch": 0.6978798586572438,
559
+ "grad_norm": 0.06787109375,
560
+ "learning_rate": 1.5350685707143258e-05,
561
+ "loss": 0.0973,
562
+ "step": 395
563
+ },
564
+ {
565
+ "epoch": 0.7067137809187279,
566
+ "grad_norm": 0.06591796875,
567
+ "learning_rate": 1.4802003241382406e-05,
568
+ "loss": 0.0963,
569
+ "step": 400
570
+ },
571
+ {
572
+ "epoch": 0.715547703180212,
573
+ "grad_norm": 0.0673828125,
574
+ "learning_rate": 1.4264184917477397e-05,
575
+ "loss": 0.0964,
576
+ "step": 405
577
+ },
578
+ {
579
+ "epoch": 0.7243816254416962,
580
+ "grad_norm": 0.0673828125,
581
+ "learning_rate": 1.3737690881597321e-05,
582
+ "loss": 0.0981,
583
+ "step": 410
584
+ },
585
+ {
586
+ "epoch": 0.7332155477031802,
587
+ "grad_norm": 0.068359375,
588
+ "learning_rate": 1.3222971591086014e-05,
589
+ "loss": 0.0977,
590
+ "step": 415
591
+ },
592
+ {
593
+ "epoch": 0.7420494699646644,
594
+ "grad_norm": 0.06689453125,
595
+ "learning_rate": 1.2720467429060156e-05,
596
+ "loss": 0.0975,
597
+ "step": 420
598
+ },
599
+ {
600
+ "epoch": 0.7508833922261484,
601
+ "grad_norm": 0.06787109375,
602
+ "learning_rate": 1.2230608327626608e-05,
603
+ "loss": 0.0978,
604
+ "step": 425
605
+ },
606
+ {
607
+ "epoch": 0.7597173144876325,
608
+ "grad_norm": 0.0712890625,
609
+ "learning_rate": 1.1753813400041453e-05,
610
+ "loss": 0.0954,
611
+ "step": 430
612
+ },
613
+ {
614
+ "epoch": 0.7685512367491166,
615
+ "grad_norm": 0.06640625,
616
+ "learning_rate": 1.1290490582125454e-05,
617
+ "loss": 0.0952,
618
+ "step": 435
619
+ },
620
+ {
621
+ "epoch": 0.7773851590106007,
622
+ "grad_norm": 0.068359375,
623
+ "learning_rate": 1.0841036283242558e-05,
624
+ "loss": 0.0971,
625
+ "step": 440
626
+ },
627
+ {
628
+ "epoch": 0.7862190812720848,
629
+ "grad_norm": 0.068359375,
630
+ "learning_rate": 1.0405835047140401e-05,
631
+ "loss": 0.0982,
632
+ "step": 445
633
+ },
634
+ {
635
+ "epoch": 0.7950530035335689,
636
+ "grad_norm": 0.06689453125,
637
+ "learning_rate": 9.985259222942602e-06,
638
+ "loss": 0.0952,
639
+ "step": 450
640
+ },
641
+ {
642
+ "epoch": 0.803886925795053,
643
+ "grad_norm": 0.0703125,
644
+ "learning_rate": 9.57966864657463e-06,
645
+ "loss": 0.0978,
646
+ "step": 455
647
+ },
648
+ {
649
+ "epoch": 0.8127208480565371,
650
+ "grad_norm": 0.06982421875,
651
+ "learning_rate": 9.189410332895662e-06,
652
+ "loss": 0.0989,
653
+ "step": 460
654
+ },
655
+ {
656
+ "epoch": 0.8215547703180212,
657
+ "grad_norm": 0.06787109375,
658
+ "learning_rate": 8.814818178799892e-06,
659
+ "loss": 0.0981,
660
+ "step": 465
661
+ },
662
+ {
663
+ "epoch": 0.8303886925795053,
664
+ "grad_norm": 0.06591796875,
665
+ "learning_rate": 8.456212677541312e-06,
666
+ "loss": 0.0945,
667
+ "step": 470
668
+ },
669
+ {
670
+ "epoch": 0.8392226148409894,
671
+ "grad_norm": 0.0673828125,
672
+ "learning_rate": 8.113900644526301e-06,
673
+ "loss": 0.0988,
674
+ "step": 475
675
+ },
676
+ {
677
+ "epoch": 0.8480565371024735,
678
+ "grad_norm": 0.06689453125,
679
+ "learning_rate": 7.788174954808826e-06,
680
+ "loss": 0.0973,
681
+ "step": 480
682
+ },
683
+ {
684
+ "epoch": 0.8568904593639576,
685
+ "grad_norm": 0.06689453125,
686
+ "learning_rate": 7.479314292512542e-06,
687
+ "loss": 0.0972,
688
+ "step": 485
689
+ },
690
+ {
691
+ "epoch": 0.8657243816254417,
692
+ "grad_norm": 0.06689453125,
693
+ "learning_rate": 7.187582912394548e-06,
694
+ "loss": 0.0977,
695
+ "step": 490
696
+ },
697
+ {
698
+ "epoch": 0.8745583038869258,
699
+ "grad_norm": 0.06591796875,
700
+ "learning_rate": 6.913230413754452e-06,
701
+ "loss": 0.0962,
702
+ "step": 495
703
+ },
704
+ {
705
+ "epoch": 0.8833922261484098,
706
+ "grad_norm": 0.06982421875,
707
+ "learning_rate": 6.656491526882422e-06,
708
+ "loss": 0.0958,
709
+ "step": 500
710
+ },
711
+ {
712
+ "epoch": 0.892226148409894,
713
+ "grad_norm": 0.06494140625,
714
+ "learning_rate": 6.417585912228833e-06,
715
+ "loss": 0.0959,
716
+ "step": 505
717
+ },
718
+ {
719
+ "epoch": 0.901060070671378,
720
+ "grad_norm": 0.0703125,
721
+ "learning_rate": 6.196717972467361e-06,
722
+ "loss": 0.0978,
723
+ "step": 510
724
+ },
725
+ {
726
+ "epoch": 0.9098939929328622,
727
+ "grad_norm": 0.0673828125,
728
+ "learning_rate": 5.994076677612297e-06,
729
+ "loss": 0.0982,
730
+ "step": 515
731
+ },
732
+ {
733
+ "epoch": 0.9187279151943463,
734
+ "grad_norm": 0.06640625,
735
+ "learning_rate": 5.809835403339747e-06,
736
+ "loss": 0.0971,
737
+ "step": 520
738
+ },
739
+ {
740
+ "epoch": 0.9275618374558304,
741
+ "grad_norm": 0.06787109375,
742
+ "learning_rate": 5.644151782650993e-06,
743
+ "loss": 0.0953,
744
+ "step": 525
745
+ },
746
+ {
747
+ "epoch": 0.9363957597173145,
748
+ "grad_norm": 0.06982421875,
749
+ "learning_rate": 5.497167571004998e-06,
750
+ "loss": 0.0956,
751
+ "step": 530
752
+ },
753
+ {
754
+ "epoch": 0.9452296819787986,
755
+ "grad_norm": 0.06689453125,
756
+ "learning_rate": 5.36900852503536e-06,
757
+ "loss": 0.0963,
758
+ "step": 535
759
+ },
760
+ {
761
+ "epoch": 0.9540636042402827,
762
+ "grad_norm": 0.06494140625,
763
+ "learning_rate": 5.259784294955576e-06,
764
+ "loss": 0.0985,
765
+ "step": 540
766
+ },
767
+ {
768
+ "epoch": 0.9628975265017667,
769
+ "grad_norm": 0.06884765625,
770
+ "learning_rate": 5.169588330744585e-06,
771
+ "loss": 0.0982,
772
+ "step": 545
773
+ },
774
+ {
775
+ "epoch": 0.9717314487632509,
776
+ "grad_norm": 0.068359375,
777
+ "learning_rate": 5.098497802192923e-06,
778
+ "loss": 0.0979,
779
+ "step": 550
780
+ },
781
+ {
782
+ "epoch": 0.980565371024735,
783
+ "grad_norm": 0.06591796875,
784
+ "learning_rate": 5.046573532877835e-06,
785
+ "loss": 0.0977,
786
+ "step": 555
787
+ },
788
+ {
789
+ "epoch": 0.9893992932862191,
790
+ "grad_norm": 0.06689453125,
791
+ "learning_rate": 5.013859948123909e-06,
792
+ "loss": 0.0976,
793
+ "step": 560
794
+ },
795
+ {
796
+ "epoch": 0.9982332155477032,
797
+ "grad_norm": 0.06591796875,
798
+ "learning_rate": 5.000385036993684e-06,
799
+ "loss": 0.0977,
800
+ "step": 565
801
+ },
802
  {
803
  "epoch": 1.0,
804
+ "step": 566,
805
+ "total_flos": 5.828408442271826e+17,
806
+ "train_loss": 0.10050818478102819,
807
+ "train_runtime": 2227.4528,
808
+ "train_samples_per_second": 32.502,
809
+ "train_steps_per_second": 0.254
810
  }
811
  ],
812
  "logging_steps": 5,
813
+ "max_steps": 566,
814
  "num_input_tokens_seen": 0,
815
  "num_train_epochs": 1,
816
  "save_steps": 100,
 
826
  "attributes": {}
827
  }
828
  },
829
+ "total_flos": 5.828408442271826e+17,
830
  "train_batch_size": 128,
831
  "trial_name": null,
832
  "trial_params": null