flyingbugs commited on
Commit
5b61667
·
verified ·
1 Parent(s): 5c97647

Model save

Browse files
README.md ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: Qwen/Qwen2.5-Math-1.5B-Instruct
3
+ library_name: transformers
4
+ model_name: OpenR1-Qwen-1.5B-SFT-utility
5
+ tags:
6
+ - generated_from_trainer
7
+ - trl
8
+ - sft
9
+ licence: license
10
+ ---
11
+
12
+ # Model Card for OpenR1-Qwen-1.5B-SFT-utility
13
+
14
+ This model is a fine-tuned version of [Qwen/Qwen2.5-Math-1.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Math-1.5B-Instruct).
15
+ It has been trained using [TRL](https://github.com/huggingface/trl).
16
+
17
+ ## Quick start
18
+
19
+ ```python
20
+ from transformers import pipeline
21
+
22
+ question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
23
+ generator = pipeline("text-generation", model="flyingbugs/OpenR1-Qwen-1.5B-SFT-utility", device="cuda")
24
+ output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
25
+ print(output["generated_text"])
26
+ ```
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/jjh233/huggingface/runs/j7yfkyje)
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: 3.3.2
41
+ - Tokenizers: 0.21.0
42
+
43
+ ## Citations
44
+
45
+
46
+
47
+ Cite TRL as:
48
+
49
+ ```bibtex
50
+ @misc{vonwerra2022trl,
51
+ title = {{TRL: Transformer Reinforcement Learning}},
52
+ author = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallouédec},
53
+ year = 2020,
54
+ journal = {GitHub repository},
55
+ publisher = {GitHub},
56
+ howpublished = {\url{https://github.com/huggingface/trl}}
57
+ }
58
+ ```
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "total_flos": 42389716598784.0,
3
+ "train_loss": 6.370550432512837,
4
+ "train_runtime": 17767.6162,
5
+ "train_samples": 15806,
6
+ "train_samples_per_second": 0.083,
7
+ "train_steps_per_second": 0.005
8
+ }
generation_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 151643,
3
+ "eos_token_id": [
4
+ 151645,
5
+ 151643
6
+ ],
7
+ "pad_token_id": 151643,
8
+ "transformers_version": "4.49.0"
9
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "total_flos": 42389716598784.0,
3
+ "train_loss": 6.370550432512837,
4
+ "train_runtime": 17767.6162,
5
+ "train_samples": 15806,
6
+ "train_samples_per_second": 0.083,
7
+ "train_steps_per_second": 0.005
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,693 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "eval_steps": 500,
6
+ "global_step": 93,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.03225806451612903,
13
+ "grad_norm": 14.220000481200906,
14
+ "learning_rate": 5e-06,
15
+ "loss": 13.9265,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.06451612903225806,
20
+ "grad_norm": 14.702762596013635,
21
+ "learning_rate": 1e-05,
22
+ "loss": 13.95,
23
+ "step": 2
24
+ },
25
+ {
26
+ "epoch": 0.0967741935483871,
27
+ "grad_norm": 14.297305924795248,
28
+ "learning_rate": 1.5e-05,
29
+ "loss": 13.9116,
30
+ "step": 3
31
+ },
32
+ {
33
+ "epoch": 0.12903225806451613,
34
+ "grad_norm": 14.484767601202535,
35
+ "learning_rate": 2e-05,
36
+ "loss": 13.753,
37
+ "step": 4
38
+ },
39
+ {
40
+ "epoch": 0.16129032258064516,
41
+ "grad_norm": 14.480323510985746,
42
+ "learning_rate": 2.5e-05,
43
+ "loss": 13.222,
44
+ "step": 5
45
+ },
46
+ {
47
+ "epoch": 0.1935483870967742,
48
+ "grad_norm": 14.002264166506258,
49
+ "learning_rate": 3e-05,
50
+ "loss": 12.1647,
51
+ "step": 6
52
+ },
53
+ {
54
+ "epoch": 0.22580645161290322,
55
+ "grad_norm": 7.226601449948956,
56
+ "learning_rate": 3.5e-05,
57
+ "loss": 11.1642,
58
+ "step": 7
59
+ },
60
+ {
61
+ "epoch": 0.25806451612903225,
62
+ "grad_norm": 13.608621977990989,
63
+ "learning_rate": 4e-05,
64
+ "loss": 10.8504,
65
+ "step": 8
66
+ },
67
+ {
68
+ "epoch": 0.2903225806451613,
69
+ "grad_norm": 6.436610267622404,
70
+ "learning_rate": 4.5e-05,
71
+ "loss": 10.4279,
72
+ "step": 9
73
+ },
74
+ {
75
+ "epoch": 0.3225806451612903,
76
+ "grad_norm": 4.922039728837155,
77
+ "learning_rate": 5e-05,
78
+ "loss": 10.1287,
79
+ "step": 10
80
+ },
81
+ {
82
+ "epoch": 0.3548387096774194,
83
+ "grad_norm": 12.033200559896779,
84
+ "learning_rate": 4.9397590361445786e-05,
85
+ "loss": 9.9607,
86
+ "step": 11
87
+ },
88
+ {
89
+ "epoch": 0.3870967741935484,
90
+ "grad_norm": 3.5396711201156807,
91
+ "learning_rate": 4.879518072289157e-05,
92
+ "loss": 9.6616,
93
+ "step": 12
94
+ },
95
+ {
96
+ "epoch": 0.41935483870967744,
97
+ "grad_norm": 3.4751992441084534,
98
+ "learning_rate": 4.8192771084337354e-05,
99
+ "loss": 9.3783,
100
+ "step": 13
101
+ },
102
+ {
103
+ "epoch": 0.45161290322580644,
104
+ "grad_norm": 3.9005409440562735,
105
+ "learning_rate": 4.759036144578313e-05,
106
+ "loss": 9.1355,
107
+ "step": 14
108
+ },
109
+ {
110
+ "epoch": 0.4838709677419355,
111
+ "grad_norm": 2.8756405244423915,
112
+ "learning_rate": 4.698795180722892e-05,
113
+ "loss": 8.8203,
114
+ "step": 15
115
+ },
116
+ {
117
+ "epoch": 0.5161290322580645,
118
+ "grad_norm": 3.9300784221492027,
119
+ "learning_rate": 4.63855421686747e-05,
120
+ "loss": 8.5065,
121
+ "step": 16
122
+ },
123
+ {
124
+ "epoch": 0.5483870967741935,
125
+ "grad_norm": 5.84606546682727,
126
+ "learning_rate": 4.578313253012048e-05,
127
+ "loss": 8.3534,
128
+ "step": 17
129
+ },
130
+ {
131
+ "epoch": 0.5806451612903226,
132
+ "grad_norm": 5.025102319247832,
133
+ "learning_rate": 4.5180722891566266e-05,
134
+ "loss": 8.1611,
135
+ "step": 18
136
+ },
137
+ {
138
+ "epoch": 0.6129032258064516,
139
+ "grad_norm": 3.5261942209741486,
140
+ "learning_rate": 4.457831325301205e-05,
141
+ "loss": 8.0223,
142
+ "step": 19
143
+ },
144
+ {
145
+ "epoch": 0.6451612903225806,
146
+ "grad_norm": 8.46799283954891,
147
+ "learning_rate": 4.3975903614457834e-05,
148
+ "loss": 7.8982,
149
+ "step": 20
150
+ },
151
+ {
152
+ "epoch": 0.6774193548387096,
153
+ "grad_norm": 3.926166227638816,
154
+ "learning_rate": 4.337349397590362e-05,
155
+ "loss": 7.6319,
156
+ "step": 21
157
+ },
158
+ {
159
+ "epoch": 0.7096774193548387,
160
+ "grad_norm": 3.358443458515684,
161
+ "learning_rate": 4.27710843373494e-05,
162
+ "loss": 7.4718,
163
+ "step": 22
164
+ },
165
+ {
166
+ "epoch": 0.7419354838709677,
167
+ "grad_norm": 2.7982062504728487,
168
+ "learning_rate": 4.2168674698795186e-05,
169
+ "loss": 7.3488,
170
+ "step": 23
171
+ },
172
+ {
173
+ "epoch": 0.7741935483870968,
174
+ "grad_norm": 3.1113638072919216,
175
+ "learning_rate": 4.156626506024097e-05,
176
+ "loss": 7.15,
177
+ "step": 24
178
+ },
179
+ {
180
+ "epoch": 0.8064516129032258,
181
+ "grad_norm": 3.1595449829251945,
182
+ "learning_rate": 4.0963855421686746e-05,
183
+ "loss": 7.0563,
184
+ "step": 25
185
+ },
186
+ {
187
+ "epoch": 0.8387096774193549,
188
+ "grad_norm": 2.108212146858006,
189
+ "learning_rate": 4.036144578313254e-05,
190
+ "loss": 6.9565,
191
+ "step": 26
192
+ },
193
+ {
194
+ "epoch": 0.8709677419354839,
195
+ "grad_norm": 2.461122936407565,
196
+ "learning_rate": 3.9759036144578314e-05,
197
+ "loss": 6.8685,
198
+ "step": 27
199
+ },
200
+ {
201
+ "epoch": 0.9032258064516129,
202
+ "grad_norm": 1.6920822731314087,
203
+ "learning_rate": 3.91566265060241e-05,
204
+ "loss": 6.7245,
205
+ "step": 28
206
+ },
207
+ {
208
+ "epoch": 0.9354838709677419,
209
+ "grad_norm": 2.2327133352179276,
210
+ "learning_rate": 3.855421686746988e-05,
211
+ "loss": 6.6712,
212
+ "step": 29
213
+ },
214
+ {
215
+ "epoch": 0.967741935483871,
216
+ "grad_norm": 2.398221999231477,
217
+ "learning_rate": 3.7951807228915666e-05,
218
+ "loss": 6.5071,
219
+ "step": 30
220
+ },
221
+ {
222
+ "epoch": 1.0,
223
+ "grad_norm": 1.9848914786193277,
224
+ "learning_rate": 3.734939759036144e-05,
225
+ "loss": 6.3974,
226
+ "step": 31
227
+ },
228
+ {
229
+ "epoch": 1.032258064516129,
230
+ "grad_norm": 1.8070671857380076,
231
+ "learning_rate": 3.674698795180723e-05,
232
+ "loss": 6.3319,
233
+ "step": 32
234
+ },
235
+ {
236
+ "epoch": 1.064516129032258,
237
+ "grad_norm": 1.6370922118025715,
238
+ "learning_rate": 3.614457831325301e-05,
239
+ "loss": 6.1884,
240
+ "step": 33
241
+ },
242
+ {
243
+ "epoch": 1.096774193548387,
244
+ "grad_norm": 1.3131279804848621,
245
+ "learning_rate": 3.5542168674698794e-05,
246
+ "loss": 6.1036,
247
+ "step": 34
248
+ },
249
+ {
250
+ "epoch": 1.129032258064516,
251
+ "grad_norm": 1.340531475202813,
252
+ "learning_rate": 3.4939759036144585e-05,
253
+ "loss": 5.9671,
254
+ "step": 35
255
+ },
256
+ {
257
+ "epoch": 1.1612903225806452,
258
+ "grad_norm": 2.005454028622789,
259
+ "learning_rate": 3.433734939759036e-05,
260
+ "loss": 5.8771,
261
+ "step": 36
262
+ },
263
+ {
264
+ "epoch": 1.1935483870967742,
265
+ "grad_norm": 1.4142674915557123,
266
+ "learning_rate": 3.3734939759036146e-05,
267
+ "loss": 5.8714,
268
+ "step": 37
269
+ },
270
+ {
271
+ "epoch": 1.2258064516129032,
272
+ "grad_norm": 1.2956798900198585,
273
+ "learning_rate": 3.313253012048193e-05,
274
+ "loss": 5.7898,
275
+ "step": 38
276
+ },
277
+ {
278
+ "epoch": 1.2580645161290323,
279
+ "grad_norm": 1.6593235984258043,
280
+ "learning_rate": 3.253012048192771e-05,
281
+ "loss": 5.6873,
282
+ "step": 39
283
+ },
284
+ {
285
+ "epoch": 1.2903225806451613,
286
+ "grad_norm": 1.380163595182406,
287
+ "learning_rate": 3.192771084337349e-05,
288
+ "loss": 5.7097,
289
+ "step": 40
290
+ },
291
+ {
292
+ "epoch": 1.3225806451612903,
293
+ "grad_norm": 1.0401057745638052,
294
+ "learning_rate": 3.132530120481928e-05,
295
+ "loss": 5.6556,
296
+ "step": 41
297
+ },
298
+ {
299
+ "epoch": 1.3548387096774195,
300
+ "grad_norm": 0.8473367054648464,
301
+ "learning_rate": 3.072289156626506e-05,
302
+ "loss": 5.4949,
303
+ "step": 42
304
+ },
305
+ {
306
+ "epoch": 1.3870967741935485,
307
+ "grad_norm": 1.1572154596782513,
308
+ "learning_rate": 3.012048192771085e-05,
309
+ "loss": 5.4822,
310
+ "step": 43
311
+ },
312
+ {
313
+ "epoch": 1.4193548387096775,
314
+ "grad_norm": 1.1443583002498423,
315
+ "learning_rate": 2.951807228915663e-05,
316
+ "loss": 5.4141,
317
+ "step": 44
318
+ },
319
+ {
320
+ "epoch": 1.4516129032258065,
321
+ "grad_norm": 1.1052929131923555,
322
+ "learning_rate": 2.891566265060241e-05,
323
+ "loss": 5.3635,
324
+ "step": 45
325
+ },
326
+ {
327
+ "epoch": 1.4838709677419355,
328
+ "grad_norm": 0.8970565525154285,
329
+ "learning_rate": 2.8313253012048197e-05,
330
+ "loss": 5.2672,
331
+ "step": 46
332
+ },
333
+ {
334
+ "epoch": 1.5161290322580645,
335
+ "grad_norm": 0.756063967066019,
336
+ "learning_rate": 2.7710843373493977e-05,
337
+ "loss": 5.2908,
338
+ "step": 47
339
+ },
340
+ {
341
+ "epoch": 1.5483870967741935,
342
+ "grad_norm": 0.7264373375167542,
343
+ "learning_rate": 2.7108433734939758e-05,
344
+ "loss": 5.2698,
345
+ "step": 48
346
+ },
347
+ {
348
+ "epoch": 1.5806451612903225,
349
+ "grad_norm": 0.7011676067381816,
350
+ "learning_rate": 2.6506024096385545e-05,
351
+ "loss": 5.2454,
352
+ "step": 49
353
+ },
354
+ {
355
+ "epoch": 1.6129032258064515,
356
+ "grad_norm": 0.6927754707271319,
357
+ "learning_rate": 2.5903614457831325e-05,
358
+ "loss": 5.1169,
359
+ "step": 50
360
+ },
361
+ {
362
+ "epoch": 1.6451612903225805,
363
+ "grad_norm": 0.6274016321054947,
364
+ "learning_rate": 2.530120481927711e-05,
365
+ "loss": 5.079,
366
+ "step": 51
367
+ },
368
+ {
369
+ "epoch": 1.6774193548387095,
370
+ "grad_norm": 0.6260513529033087,
371
+ "learning_rate": 2.4698795180722893e-05,
372
+ "loss": 5.0318,
373
+ "step": 52
374
+ },
375
+ {
376
+ "epoch": 1.7096774193548387,
377
+ "grad_norm": 0.6122295776355817,
378
+ "learning_rate": 2.4096385542168677e-05,
379
+ "loss": 5.0809,
380
+ "step": 53
381
+ },
382
+ {
383
+ "epoch": 1.7419354838709677,
384
+ "grad_norm": 0.6219482405443587,
385
+ "learning_rate": 2.349397590361446e-05,
386
+ "loss": 4.9977,
387
+ "step": 54
388
+ },
389
+ {
390
+ "epoch": 1.7741935483870968,
391
+ "grad_norm": 0.7127219759114208,
392
+ "learning_rate": 2.289156626506024e-05,
393
+ "loss": 4.9713,
394
+ "step": 55
395
+ },
396
+ {
397
+ "epoch": 1.8064516129032258,
398
+ "grad_norm": 0.6728184125308878,
399
+ "learning_rate": 2.2289156626506025e-05,
400
+ "loss": 4.9564,
401
+ "step": 56
402
+ },
403
+ {
404
+ "epoch": 1.838709677419355,
405
+ "grad_norm": 0.6658692158816993,
406
+ "learning_rate": 2.168674698795181e-05,
407
+ "loss": 4.8953,
408
+ "step": 57
409
+ },
410
+ {
411
+ "epoch": 1.870967741935484,
412
+ "grad_norm": 0.6480363491872242,
413
+ "learning_rate": 2.1084337349397593e-05,
414
+ "loss": 4.8271,
415
+ "step": 58
416
+ },
417
+ {
418
+ "epoch": 1.903225806451613,
419
+ "grad_norm": 0.6798514027495196,
420
+ "learning_rate": 2.0481927710843373e-05,
421
+ "loss": 4.856,
422
+ "step": 59
423
+ },
424
+ {
425
+ "epoch": 1.935483870967742,
426
+ "grad_norm": 0.5868913336163518,
427
+ "learning_rate": 1.9879518072289157e-05,
428
+ "loss": 4.7935,
429
+ "step": 60
430
+ },
431
+ {
432
+ "epoch": 1.967741935483871,
433
+ "grad_norm": 0.5793335785631702,
434
+ "learning_rate": 1.927710843373494e-05,
435
+ "loss": 4.8151,
436
+ "step": 61
437
+ },
438
+ {
439
+ "epoch": 2.0,
440
+ "grad_norm": 0.5816955191304313,
441
+ "learning_rate": 1.867469879518072e-05,
442
+ "loss": 4.7828,
443
+ "step": 62
444
+ },
445
+ {
446
+ "epoch": 2.032258064516129,
447
+ "grad_norm": 0.5845389898807549,
448
+ "learning_rate": 1.8072289156626505e-05,
449
+ "loss": 4.6459,
450
+ "step": 63
451
+ },
452
+ {
453
+ "epoch": 2.064516129032258,
454
+ "grad_norm": 0.5946763020956095,
455
+ "learning_rate": 1.7469879518072292e-05,
456
+ "loss": 4.7297,
457
+ "step": 64
458
+ },
459
+ {
460
+ "epoch": 2.096774193548387,
461
+ "grad_norm": 0.5627139324820443,
462
+ "learning_rate": 1.6867469879518073e-05,
463
+ "loss": 4.6863,
464
+ "step": 65
465
+ },
466
+ {
467
+ "epoch": 2.129032258064516,
468
+ "grad_norm": 0.5636104401228851,
469
+ "learning_rate": 1.6265060240963857e-05,
470
+ "loss": 4.6437,
471
+ "step": 66
472
+ },
473
+ {
474
+ "epoch": 2.161290322580645,
475
+ "grad_norm": 0.5966948986254044,
476
+ "learning_rate": 1.566265060240964e-05,
477
+ "loss": 4.6102,
478
+ "step": 67
479
+ },
480
+ {
481
+ "epoch": 2.193548387096774,
482
+ "grad_norm": 0.5635496928050323,
483
+ "learning_rate": 1.5060240963855424e-05,
484
+ "loss": 4.56,
485
+ "step": 68
486
+ },
487
+ {
488
+ "epoch": 2.225806451612903,
489
+ "grad_norm": 0.6201573994580546,
490
+ "learning_rate": 1.4457831325301205e-05,
491
+ "loss": 4.5522,
492
+ "step": 69
493
+ },
494
+ {
495
+ "epoch": 2.258064516129032,
496
+ "grad_norm": 0.5283310806265233,
497
+ "learning_rate": 1.3855421686746989e-05,
498
+ "loss": 4.5434,
499
+ "step": 70
500
+ },
501
+ {
502
+ "epoch": 2.2903225806451615,
503
+ "grad_norm": 0.5830200387240102,
504
+ "learning_rate": 1.3253012048192772e-05,
505
+ "loss": 4.5213,
506
+ "step": 71
507
+ },
508
+ {
509
+ "epoch": 2.3225806451612905,
510
+ "grad_norm": 0.5366398051729224,
511
+ "learning_rate": 1.2650602409638555e-05,
512
+ "loss": 4.4882,
513
+ "step": 72
514
+ },
515
+ {
516
+ "epoch": 2.3548387096774195,
517
+ "grad_norm": 0.5278747342683787,
518
+ "learning_rate": 1.2048192771084338e-05,
519
+ "loss": 4.4285,
520
+ "step": 73
521
+ },
522
+ {
523
+ "epoch": 2.3870967741935485,
524
+ "grad_norm": 0.5187375955155756,
525
+ "learning_rate": 1.144578313253012e-05,
526
+ "loss": 4.4649,
527
+ "step": 74
528
+ },
529
+ {
530
+ "epoch": 2.4193548387096775,
531
+ "grad_norm": 0.49022439094573594,
532
+ "learning_rate": 1.0843373493975904e-05,
533
+ "loss": 4.4827,
534
+ "step": 75
535
+ },
536
+ {
537
+ "epoch": 2.4516129032258065,
538
+ "grad_norm": 0.4832270774924177,
539
+ "learning_rate": 1.0240963855421687e-05,
540
+ "loss": 4.4521,
541
+ "step": 76
542
+ },
543
+ {
544
+ "epoch": 2.4838709677419355,
545
+ "grad_norm": 0.5104329783830706,
546
+ "learning_rate": 9.63855421686747e-06,
547
+ "loss": 4.4009,
548
+ "step": 77
549
+ },
550
+ {
551
+ "epoch": 2.5161290322580645,
552
+ "grad_norm": 0.46199522908208795,
553
+ "learning_rate": 9.036144578313253e-06,
554
+ "loss": 4.3962,
555
+ "step": 78
556
+ },
557
+ {
558
+ "epoch": 2.5483870967741935,
559
+ "grad_norm": 0.4752588993383047,
560
+ "learning_rate": 8.433734939759036e-06,
561
+ "loss": 4.4422,
562
+ "step": 79
563
+ },
564
+ {
565
+ "epoch": 2.5806451612903225,
566
+ "grad_norm": 0.48368008070516033,
567
+ "learning_rate": 7.83132530120482e-06,
568
+ "loss": 4.2568,
569
+ "step": 80
570
+ },
571
+ {
572
+ "epoch": 2.6129032258064515,
573
+ "grad_norm": 0.47323578510274494,
574
+ "learning_rate": 7.228915662650602e-06,
575
+ "loss": 4.3716,
576
+ "step": 81
577
+ },
578
+ {
579
+ "epoch": 2.6451612903225805,
580
+ "grad_norm": 0.47907919622898615,
581
+ "learning_rate": 6.626506024096386e-06,
582
+ "loss": 4.4467,
583
+ "step": 82
584
+ },
585
+ {
586
+ "epoch": 2.6774193548387095,
587
+ "grad_norm": 0.5132262686017864,
588
+ "learning_rate": 6.024096385542169e-06,
589
+ "loss": 4.353,
590
+ "step": 83
591
+ },
592
+ {
593
+ "epoch": 2.709677419354839,
594
+ "grad_norm": 0.4512163239687551,
595
+ "learning_rate": 5.421686746987952e-06,
596
+ "loss": 4.3904,
597
+ "step": 84
598
+ },
599
+ {
600
+ "epoch": 2.741935483870968,
601
+ "grad_norm": 0.4613864288737105,
602
+ "learning_rate": 4.819277108433735e-06,
603
+ "loss": 4.3699,
604
+ "step": 85
605
+ },
606
+ {
607
+ "epoch": 2.774193548387097,
608
+ "grad_norm": 0.4834904636291854,
609
+ "learning_rate": 4.216867469879518e-06,
610
+ "loss": 4.3382,
611
+ "step": 86
612
+ },
613
+ {
614
+ "epoch": 2.806451612903226,
615
+ "grad_norm": 0.45137040545065016,
616
+ "learning_rate": 3.614457831325301e-06,
617
+ "loss": 4.3651,
618
+ "step": 87
619
+ },
620
+ {
621
+ "epoch": 2.838709677419355,
622
+ "grad_norm": 0.44625483484835893,
623
+ "learning_rate": 3.0120481927710846e-06,
624
+ "loss": 4.4046,
625
+ "step": 88
626
+ },
627
+ {
628
+ "epoch": 2.870967741935484,
629
+ "grad_norm": 0.4306895833715559,
630
+ "learning_rate": 2.4096385542168676e-06,
631
+ "loss": 4.3594,
632
+ "step": 89
633
+ },
634
+ {
635
+ "epoch": 2.903225806451613,
636
+ "grad_norm": 0.46583862244193375,
637
+ "learning_rate": 1.8072289156626506e-06,
638
+ "loss": 4.3189,
639
+ "step": 90
640
+ },
641
+ {
642
+ "epoch": 2.935483870967742,
643
+ "grad_norm": 0.4698345242151301,
644
+ "learning_rate": 1.2048192771084338e-06,
645
+ "loss": 4.3348,
646
+ "step": 91
647
+ },
648
+ {
649
+ "epoch": 2.967741935483871,
650
+ "grad_norm": 0.4355936118573381,
651
+ "learning_rate": 6.024096385542169e-07,
652
+ "loss": 4.3989,
653
+ "step": 92
654
+ },
655
+ {
656
+ "epoch": 3.0,
657
+ "grad_norm": 0.4741459705347069,
658
+ "learning_rate": 0.0,
659
+ "loss": 4.3099,
660
+ "step": 93
661
+ },
662
+ {
663
+ "epoch": 3.0,
664
+ "step": 93,
665
+ "total_flos": 42389716598784.0,
666
+ "train_loss": 6.370550432512837,
667
+ "train_runtime": 17767.6162,
668
+ "train_samples_per_second": 0.083,
669
+ "train_steps_per_second": 0.005
670
+ }
671
+ ],
672
+ "logging_steps": 1,
673
+ "max_steps": 93,
674
+ "num_input_tokens_seen": 0,
675
+ "num_train_epochs": 3,
676
+ "save_steps": 500,
677
+ "stateful_callbacks": {
678
+ "TrainerControl": {
679
+ "args": {
680
+ "should_epoch_stop": false,
681
+ "should_evaluate": false,
682
+ "should_log": false,
683
+ "should_save": true,
684
+ "should_training_stop": true
685
+ },
686
+ "attributes": {}
687
+ }
688
+ },
689
+ "total_flos": 42389716598784.0,
690
+ "train_batch_size": 4,
691
+ "trial_name": null,
692
+ "trial_params": null
693
+ }