Zheng Li
commited on
End of training
Browse files- README.md +4 -3
- all_results.json +11 -11
- eval_results.json +6 -6
- runs/May14_15-54-09_cs-Precision-7960-Tower/events.out.tfevents.1747253123.cs-Precision-7960-Tower.127892.1 +3 -0
- train_results.json +6 -6
- trainer_state.json +715 -968
README.md
CHANGED
@@ -3,6 +3,7 @@ library_name: transformers
|
|
3 |
license: apache-2.0
|
4 |
base_model: facebook/wav2vec2-base
|
5 |
tags:
|
|
|
6 |
- generated_from_trainer
|
7 |
datasets:
|
8 |
- superb
|
@@ -23,7 +24,7 @@ model-index:
|
|
23 |
metrics:
|
24 |
- name: Accuracy
|
25 |
type: accuracy
|
26 |
-
value: 0.
|
27 |
---
|
28 |
|
29 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
@@ -33,8 +34,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
33 |
|
34 |
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the superb dataset.
|
35 |
It achieves the following results on the evaluation set:
|
36 |
-
- Loss: 0.
|
37 |
-
- Accuracy: 0.
|
38 |
|
39 |
## Model description
|
40 |
|
|
|
3 |
license: apache-2.0
|
4 |
base_model: facebook/wav2vec2-base
|
5 |
tags:
|
6 |
+
- audio-classification
|
7 |
- generated_from_trainer
|
8 |
datasets:
|
9 |
- superb
|
|
|
24 |
metrics:
|
25 |
- name: Accuracy
|
26 |
type: accuracy
|
27 |
+
value: 0.9830832597822889
|
28 |
---
|
29 |
|
30 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
34 |
|
35 |
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the superb dataset.
|
36 |
It achieves the following results on the evaluation set:
|
37 |
+
- Loss: 0.0956
|
38 |
+
- Accuracy: 0.9831
|
39 |
|
40 |
## Model description
|
41 |
|
all_results.json
CHANGED
@@ -1,13 +1,13 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_loss": 0.
|
5 |
-
"eval_runtime": 5.
|
6 |
-
"eval_samples_per_second":
|
7 |
-
"eval_steps_per_second":
|
8 |
-
"total_flos":
|
9 |
-
"train_loss": 0.
|
10 |
-
"train_runtime":
|
11 |
-
"train_samples_per_second":
|
12 |
-
"train_steps_per_second":
|
13 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 7.996245306633291,
|
3 |
+
"eval_accuracy": 0.9830832597822889,
|
4 |
+
"eval_loss": 0.0956372618675232,
|
5 |
+
"eval_runtime": 5.6145,
|
6 |
+
"eval_samples_per_second": 1210.799,
|
7 |
+
"eval_steps_per_second": 37.938,
|
8 |
+
"total_flos": 3.777723239743488e+18,
|
9 |
+
"train_loss": 0.596273283347787,
|
10 |
+
"train_runtime": 640.7753,
|
11 |
+
"train_samples_per_second": 637.902,
|
12 |
+
"train_steps_per_second": 2.484
|
13 |
}
|
eval_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_loss": 0.
|
5 |
-
"eval_runtime": 5.
|
6 |
-
"eval_samples_per_second":
|
7 |
-
"eval_steps_per_second":
|
8 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 7.996245306633291,
|
3 |
+
"eval_accuracy": 0.9830832597822889,
|
4 |
+
"eval_loss": 0.0956372618675232,
|
5 |
+
"eval_runtime": 5.6145,
|
6 |
+
"eval_samples_per_second": 1210.799,
|
7 |
+
"eval_steps_per_second": 37.938
|
8 |
}
|
runs/May14_15-54-09_cs-Precision-7960-Tower/events.out.tfevents.1747253123.cs-Precision-7960-Tower.127892.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:15845e83e76e501dc7170d7b668f0ca3730974474a167df2d31342c8c5348fb8
|
3 |
+
size 411
|
train_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"total_flos":
|
4 |
-
"train_loss": 0.
|
5 |
-
"train_runtime":
|
6 |
-
"train_samples_per_second":
|
7 |
-
"train_steps_per_second":
|
8 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 7.996245306633291,
|
3 |
+
"total_flos": 3.777723239743488e+18,
|
4 |
+
"train_loss": 0.596273283347787,
|
5 |
+
"train_runtime": 640.7753,
|
6 |
+
"train_samples_per_second": 637.902,
|
7 |
+
"train_steps_per_second": 2.484
|
8 |
}
|
trainer_state.json
CHANGED
@@ -1,1465 +1,1212 @@
|
|
1 |
{
|
2 |
-
"best_metric": 0.
|
3 |
-
"best_model_checkpoint": "wav2vec2-base-ft-keyword-spotting/checkpoint-
|
4 |
-
"epoch":
|
5 |
"eval_steps": 500,
|
6 |
-
"global_step":
|
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.
|
13 |
-
"grad_norm": 2.
|
14 |
-
"learning_rate": 1.
|
15 |
-
"loss":
|
16 |
"step": 10
|
17 |
},
|
18 |
{
|
19 |
-
"epoch": 0.
|
20 |
-
"grad_norm": 3.
|
21 |
-
"learning_rate":
|
22 |
-
"loss": 4.
|
23 |
"step": 20
|
24 |
},
|
25 |
{
|
26 |
-
"epoch": 0.
|
27 |
-
"grad_norm": 3.
|
28 |
-
"learning_rate":
|
29 |
-
"loss": 4.
|
30 |
"step": 30
|
31 |
},
|
32 |
{
|
33 |
-
"epoch": 0.
|
34 |
-
"grad_norm": 4.
|
35 |
-
"learning_rate":
|
36 |
-
"loss": 3.
|
37 |
"step": 40
|
38 |
},
|
39 |
{
|
40 |
-
"epoch": 0.
|
41 |
-
"grad_norm": 5.
|
42 |
-
"learning_rate":
|
43 |
-
"loss": 3.
|
44 |
"step": 50
|
45 |
},
|
46 |
{
|
47 |
-
"epoch": 0.
|
48 |
-
"grad_norm":
|
49 |
-
"learning_rate":
|
50 |
-
"loss":
|
51 |
"step": 60
|
52 |
},
|
53 |
{
|
54 |
-
"epoch": 0.
|
55 |
-
"grad_norm": 5.
|
56 |
-
"learning_rate": 1.
|
57 |
-
"loss": 2.
|
58 |
"step": 70
|
59 |
},
|
60 |
{
|
61 |
-
"epoch": 0.
|
62 |
-
"grad_norm": 5.
|
63 |
-
"learning_rate": 1.
|
64 |
-
"loss": 2.
|
65 |
"step": 80
|
66 |
},
|
67 |
{
|
68 |
-
"epoch": 0.
|
69 |
-
"grad_norm": 4.
|
70 |
-
"learning_rate": 1.
|
71 |
-
"loss": 2.
|
72 |
"step": 90
|
73 |
},
|
74 |
{
|
75 |
-
"epoch": 0.
|
76 |
-
"grad_norm":
|
77 |
-
"learning_rate": 1.
|
78 |
-
"loss":
|
79 |
"step": 100
|
80 |
},
|
81 |
{
|
82 |
-
"epoch": 0.
|
83 |
-
"grad_norm":
|
84 |
-
"learning_rate":
|
85 |
-
"loss":
|
86 |
"step": 110
|
87 |
},
|
88 |
{
|
89 |
-
"epoch": 0.
|
90 |
-
"grad_norm":
|
91 |
-
"learning_rate":
|
92 |
-
"loss": 1.
|
93 |
"step": 120
|
94 |
},
|
95 |
{
|
96 |
-
"epoch": 0.
|
97 |
-
"grad_norm":
|
98 |
-
"learning_rate":
|
99 |
-
"loss": 1.
|
100 |
"step": 130
|
101 |
},
|
102 |
{
|
103 |
-
"epoch": 0.
|
104 |
-
"grad_norm":
|
105 |
-
"learning_rate": 2.
|
106 |
-
"loss": 1.
|
107 |
"step": 140
|
108 |
},
|
109 |
{
|
110 |
-
"epoch": 0.
|
111 |
-
"grad_norm":
|
112 |
-
"learning_rate": 2.
|
113 |
-
"loss": 1.
|
114 |
"step": 150
|
115 |
},
|
116 |
{
|
117 |
-
"epoch": 0.
|
118 |
-
"grad_norm":
|
119 |
-
"learning_rate":
|
120 |
-
"loss": 1.
|
121 |
"step": 160
|
122 |
},
|
123 |
{
|
124 |
-
"epoch": 0.
|
125 |
-
"grad_norm":
|
126 |
-
"learning_rate": 2.
|
127 |
-
"loss": 1.
|
128 |
"step": 170
|
129 |
},
|
130 |
{
|
131 |
-
"epoch": 0.
|
132 |
-
"grad_norm":
|
133 |
-
"learning_rate": 2.
|
134 |
-
"loss": 1.
|
135 |
"step": 180
|
136 |
},
|
137 |
{
|
138 |
-
"epoch": 0.
|
139 |
-
"grad_norm":
|
140 |
-
"learning_rate": 2.
|
141 |
-
"loss": 1.
|
142 |
"step": 190
|
143 |
},
|
144 |
{
|
145 |
-
"epoch": 0.
|
146 |
-
"
|
147 |
-
"
|
148 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
149 |
"step": 200
|
150 |
},
|
151 |
{
|
152 |
-
"epoch":
|
153 |
-
"grad_norm":
|
154 |
-
"learning_rate": 2.
|
155 |
-
"loss": 1.
|
156 |
"step": 210
|
157 |
},
|
158 |
{
|
159 |
-
"epoch":
|
160 |
-
"grad_norm": 2.
|
161 |
-
"learning_rate": 2.
|
162 |
-
"loss": 1.
|
163 |
"step": 220
|
164 |
},
|
165 |
{
|
166 |
-
"epoch":
|
167 |
-
"grad_norm":
|
168 |
-
"learning_rate": 2.
|
169 |
-
"loss": 1.
|
170 |
"step": 230
|
171 |
},
|
172 |
{
|
173 |
-
"epoch":
|
174 |
-
"grad_norm": 4.
|
175 |
-
"learning_rate": 2.
|
176 |
-
"loss": 1.
|
177 |
"step": 240
|
178 |
},
|
179 |
{
|
180 |
-
"epoch":
|
181 |
-
"grad_norm": 3.
|
182 |
-
"learning_rate": 2.
|
183 |
-
"loss": 1.
|
184 |
"step": 250
|
185 |
},
|
186 |
{
|
187 |
-
"epoch":
|
188 |
-
"grad_norm": 5.
|
189 |
-
"learning_rate": 2.
|
190 |
-
"loss": 1.
|
191 |
"step": 260
|
192 |
},
|
193 |
{
|
194 |
-
"epoch":
|
195 |
-
"grad_norm":
|
196 |
-
"learning_rate": 2.
|
197 |
-
"loss": 1.
|
198 |
"step": 270
|
199 |
},
|
200 |
{
|
201 |
-
"epoch":
|
202 |
-
"grad_norm":
|
203 |
-
"learning_rate": 2.
|
204 |
-
"loss": 1.
|
205 |
"step": 280
|
206 |
},
|
207 |
{
|
208 |
-
"epoch":
|
209 |
-
"grad_norm":
|
210 |
-
"learning_rate": 2.
|
211 |
-
"loss":
|
212 |
"step": 290
|
213 |
},
|
214 |
{
|
215 |
-
"epoch":
|
216 |
-
"grad_norm":
|
217 |
-
"learning_rate": 2.
|
218 |
-
"loss":
|
219 |
"step": 300
|
220 |
},
|
221 |
{
|
222 |
-
"epoch":
|
223 |
-
"grad_norm":
|
224 |
-
"learning_rate": 2.
|
225 |
-
"loss": 1.
|
226 |
"step": 310
|
227 |
},
|
228 |
{
|
229 |
-
"epoch":
|
230 |
-
"grad_norm": 2.
|
231 |
-
"learning_rate": 2.
|
232 |
-
"loss":
|
233 |
"step": 320
|
234 |
},
|
235 |
{
|
236 |
-
"epoch":
|
237 |
-
"grad_norm":
|
238 |
-
"learning_rate": 2.
|
239 |
-
"loss":
|
240 |
"step": 330
|
241 |
},
|
242 |
{
|
243 |
-
"epoch":
|
244 |
-
"grad_norm":
|
245 |
-
"learning_rate": 2.
|
246 |
-
"loss":
|
247 |
"step": 340
|
248 |
},
|
249 |
{
|
250 |
-
"epoch":
|
251 |
-
"grad_norm":
|
252 |
-
"learning_rate": 2.
|
253 |
-
"loss":
|
254 |
"step": 350
|
255 |
},
|
256 |
{
|
257 |
-
"epoch":
|
258 |
-
"grad_norm":
|
259 |
-
"learning_rate": 2.
|
260 |
-
"loss":
|
261 |
"step": 360
|
262 |
},
|
263 |
{
|
264 |
-
"epoch":
|
265 |
-
"grad_norm":
|
266 |
-
"learning_rate": 2.
|
267 |
-
"loss":
|
268 |
"step": 370
|
269 |
},
|
270 |
{
|
271 |
-
"epoch":
|
272 |
-
"grad_norm":
|
273 |
-
"learning_rate": 2.
|
274 |
-
"loss":
|
275 |
"step": 380
|
276 |
},
|
277 |
{
|
278 |
-
"epoch":
|
279 |
-
"grad_norm": 2.
|
280 |
-
"learning_rate": 2.
|
281 |
-
"loss":
|
282 |
"step": 390
|
283 |
},
|
284 |
{
|
285 |
-
"epoch": 1.
|
286 |
-
"
|
287 |
-
"
|
288 |
-
"
|
289 |
-
"
|
|
|
|
|
290 |
},
|
291 |
{
|
292 |
-
"epoch":
|
293 |
-
"
|
294 |
-
"
|
295 |
-
"
|
296 |
-
"eval_samples_per_second": 1336.08,
|
297 |
-
"eval_steps_per_second": 41.863,
|
298 |
"step": 400
|
299 |
},
|
300 |
{
|
301 |
-
"epoch":
|
302 |
-
"grad_norm":
|
303 |
-
"learning_rate": 2.
|
304 |
-
"loss": 0.
|
305 |
"step": 410
|
306 |
},
|
307 |
{
|
308 |
-
"epoch":
|
309 |
-
"grad_norm":
|
310 |
-
"learning_rate": 2.
|
311 |
-
"loss": 0.
|
312 |
"step": 420
|
313 |
},
|
314 |
{
|
315 |
-
"epoch":
|
316 |
-
"grad_norm":
|
317 |
-
"learning_rate": 2.
|
318 |
-
"loss": 0.
|
319 |
"step": 430
|
320 |
},
|
321 |
{
|
322 |
-
"epoch":
|
323 |
-
"grad_norm":
|
324 |
-
"learning_rate": 2.
|
325 |
-
"loss": 0.
|
326 |
"step": 440
|
327 |
},
|
328 |
{
|
329 |
-
"epoch":
|
330 |
-
"grad_norm":
|
331 |
-
"learning_rate": 2.
|
332 |
-
"loss": 0.
|
333 |
"step": 450
|
334 |
},
|
335 |
{
|
336 |
-
"epoch":
|
337 |
-
"grad_norm":
|
338 |
-
"learning_rate": 2.
|
339 |
-
"loss": 0.
|
340 |
"step": 460
|
341 |
},
|
342 |
{
|
343 |
-
"epoch":
|
344 |
-
"grad_norm":
|
345 |
-
"learning_rate": 2.
|
346 |
-
"loss": 0.
|
347 |
"step": 470
|
348 |
},
|
349 |
{
|
350 |
-
"epoch":
|
351 |
-
"grad_norm":
|
352 |
-
"learning_rate": 2.
|
353 |
-
"loss": 0.
|
354 |
"step": 480
|
355 |
},
|
356 |
{
|
357 |
-
"epoch":
|
358 |
-
"grad_norm":
|
359 |
-
"learning_rate": 2.
|
360 |
-
"loss": 0.
|
361 |
"step": 490
|
362 |
},
|
363 |
{
|
364 |
-
"epoch":
|
365 |
-
"grad_norm":
|
366 |
-
"learning_rate": 2.
|
367 |
-
"loss": 0.
|
368 |
"step": 500
|
369 |
},
|
370 |
{
|
371 |
-
"epoch":
|
372 |
-
"grad_norm":
|
373 |
-
"learning_rate": 2.
|
374 |
-
"loss": 0.
|
375 |
"step": 510
|
376 |
},
|
377 |
{
|
378 |
-
"epoch":
|
379 |
-
"grad_norm":
|
380 |
-
"learning_rate": 2.
|
381 |
-
"loss": 0.
|
382 |
"step": 520
|
383 |
},
|
384 |
{
|
385 |
-
"epoch":
|
386 |
-
"grad_norm": 4.
|
387 |
-
"learning_rate": 2.
|
388 |
-
"loss": 0.
|
389 |
"step": 530
|
390 |
},
|
391 |
{
|
392 |
-
"epoch":
|
393 |
-
"grad_norm":
|
394 |
-
"learning_rate": 2.
|
395 |
-
"loss": 0.
|
396 |
"step": 540
|
397 |
},
|
398 |
{
|
399 |
-
"epoch":
|
400 |
-
"grad_norm":
|
401 |
-
"learning_rate": 2.
|
402 |
-
"loss": 0.
|
403 |
"step": 550
|
404 |
},
|
405 |
{
|
406 |
-
"epoch":
|
407 |
-
"grad_norm":
|
408 |
-
"learning_rate": 2.
|
409 |
-
"loss": 0.
|
410 |
"step": 560
|
411 |
},
|
412 |
{
|
413 |
-
"epoch":
|
414 |
-
"grad_norm":
|
415 |
-
"learning_rate": 2.
|
416 |
-
"loss": 0.
|
417 |
"step": 570
|
418 |
},
|
419 |
{
|
420 |
-
"epoch":
|
421 |
-
"grad_norm":
|
422 |
-
"learning_rate": 2.
|
423 |
-
"loss": 0.
|
424 |
"step": 580
|
425 |
},
|
426 |
{
|
427 |
-
"epoch":
|
428 |
-
"grad_norm":
|
429 |
-
"learning_rate": 2.
|
430 |
-
"loss": 0.
|
431 |
"step": 590
|
432 |
},
|
433 |
{
|
434 |
-
"epoch":
|
435 |
-
"
|
436 |
-
"
|
437 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
438 |
"step": 600
|
439 |
},
|
440 |
{
|
441 |
-
"epoch":
|
442 |
-
"grad_norm":
|
443 |
-
"learning_rate": 2.
|
444 |
-
"loss": 0.
|
445 |
"step": 610
|
446 |
},
|
447 |
{
|
448 |
-
"epoch":
|
449 |
-
"grad_norm":
|
450 |
-
"learning_rate": 2.
|
451 |
-
"loss": 0.
|
452 |
"step": 620
|
453 |
},
|
454 |
{
|
455 |
-
"epoch":
|
456 |
-
"grad_norm":
|
457 |
-
"learning_rate": 2.
|
458 |
-
"loss": 0.
|
459 |
"step": 630
|
460 |
},
|
461 |
{
|
462 |
-
"epoch":
|
463 |
-
"grad_norm": 3.
|
464 |
-
"learning_rate":
|
465 |
-
"loss": 0.
|
466 |
"step": 640
|
467 |
},
|
468 |
{
|
469 |
-
"epoch":
|
470 |
-
"grad_norm": 2.
|
471 |
-
"learning_rate":
|
472 |
-
"loss": 0.
|
473 |
"step": 650
|
474 |
},
|
475 |
{
|
476 |
-
"epoch":
|
477 |
-
"grad_norm":
|
478 |
-
"learning_rate":
|
479 |
-
"loss": 0.
|
480 |
"step": 660
|
481 |
},
|
482 |
{
|
483 |
-
"epoch":
|
484 |
-
"grad_norm":
|
485 |
-
"learning_rate":
|
486 |
-
"loss": 0.
|
487 |
"step": 670
|
488 |
},
|
489 |
{
|
490 |
-
"epoch":
|
491 |
-
"grad_norm":
|
492 |
-
"learning_rate":
|
493 |
-
"loss": 0.
|
494 |
"step": 680
|
495 |
},
|
496 |
{
|
497 |
-
"epoch":
|
498 |
-
"grad_norm":
|
499 |
-
"learning_rate":
|
500 |
-
"loss": 0.
|
501 |
"step": 690
|
502 |
},
|
503 |
{
|
504 |
-
"epoch":
|
505 |
-
"grad_norm":
|
506 |
-
"learning_rate":
|
507 |
-
"loss": 0.
|
508 |
"step": 700
|
509 |
},
|
510 |
{
|
511 |
-
"epoch":
|
512 |
-
"grad_norm":
|
513 |
-
"learning_rate":
|
514 |
-
"loss": 0.
|
515 |
"step": 710
|
516 |
},
|
517 |
{
|
518 |
-
"epoch":
|
519 |
-
"grad_norm":
|
520 |
-
"learning_rate":
|
521 |
-
"loss": 0.
|
522 |
"step": 720
|
523 |
},
|
524 |
{
|
525 |
-
"epoch":
|
526 |
-
"grad_norm":
|
527 |
-
"learning_rate":
|
528 |
-
"loss": 0.
|
529 |
"step": 730
|
530 |
},
|
531 |
{
|
532 |
-
"epoch":
|
533 |
-
"grad_norm":
|
534 |
-
"learning_rate":
|
535 |
-
"loss": 0.
|
536 |
"step": 740
|
537 |
},
|
538 |
{
|
539 |
-
"epoch":
|
540 |
-
"grad_norm":
|
541 |
-
"learning_rate":
|
542 |
-
"loss": 0.
|
543 |
"step": 750
|
544 |
},
|
545 |
{
|
546 |
-
"epoch":
|
547 |
-
"grad_norm":
|
548 |
-
"learning_rate":
|
549 |
-
"loss": 0.
|
550 |
"step": 760
|
551 |
},
|
552 |
{
|
553 |
-
"epoch":
|
554 |
-
"grad_norm": 4.
|
555 |
-
"learning_rate":
|
556 |
-
"loss": 0.
|
557 |
"step": 770
|
558 |
},
|
559 |
{
|
560 |
-
"epoch":
|
561 |
-
"grad_norm":
|
562 |
-
"learning_rate":
|
563 |
-
"loss": 0.
|
564 |
"step": 780
|
565 |
},
|
566 |
{
|
567 |
-
"epoch":
|
568 |
-
"grad_norm":
|
569 |
-
"learning_rate":
|
570 |
-
"loss": 0.
|
571 |
"step": 790
|
572 |
},
|
573 |
{
|
574 |
-
"epoch":
|
575 |
-
"
|
576 |
-
"
|
577 |
-
"
|
578 |
-
"
|
|
|
|
|
579 |
},
|
580 |
{
|
581 |
-
"epoch":
|
582 |
-
"
|
583 |
-
"
|
584 |
-
"
|
585 |
-
"eval_samples_per_second": 1416.28,
|
586 |
-
"eval_steps_per_second": 44.376,
|
587 |
"step": 800
|
588 |
},
|
589 |
{
|
590 |
-
"epoch":
|
591 |
-
"grad_norm":
|
592 |
-
"learning_rate": 1.
|
593 |
-
"loss": 0.
|
594 |
"step": 810
|
595 |
},
|
596 |
{
|
597 |
-
"epoch":
|
598 |
-
"grad_norm": 4.
|
599 |
-
"learning_rate": 1.
|
600 |
-
"loss": 0.
|
601 |
"step": 820
|
602 |
},
|
603 |
{
|
604 |
-
"epoch":
|
605 |
-
"grad_norm":
|
606 |
-
"learning_rate": 1.
|
607 |
-
"loss": 0.
|
608 |
"step": 830
|
609 |
},
|
610 |
{
|
611 |
-
"epoch":
|
612 |
-
"grad_norm":
|
613 |
-
"learning_rate": 1.
|
614 |
-
"loss": 0.
|
615 |
"step": 840
|
616 |
},
|
617 |
{
|
618 |
-
"epoch":
|
619 |
-
"grad_norm":
|
620 |
-
"learning_rate": 1.
|
621 |
-
"loss": 0.
|
622 |
"step": 850
|
623 |
},
|
624 |
{
|
625 |
-
"epoch":
|
626 |
-
"grad_norm": 3.
|
627 |
-
"learning_rate": 1.
|
628 |
-
"loss": 0.
|
629 |
"step": 860
|
630 |
},
|
631 |
{
|
632 |
-
"epoch":
|
633 |
-
"grad_norm":
|
634 |
-
"learning_rate": 1.
|
635 |
-
"loss": 0.
|
636 |
"step": 870
|
637 |
},
|
638 |
{
|
639 |
-
"epoch":
|
640 |
-
"grad_norm":
|
641 |
-
"learning_rate": 1.
|
642 |
-
"loss": 0.
|
643 |
"step": 880
|
644 |
},
|
645 |
{
|
646 |
-
"epoch":
|
647 |
-
"grad_norm":
|
648 |
-
"learning_rate": 1.
|
649 |
-
"loss": 0.
|
650 |
"step": 890
|
651 |
},
|
652 |
{
|
653 |
-
"epoch":
|
654 |
-
"grad_norm":
|
655 |
-
"learning_rate": 1.
|
656 |
-
"loss": 0.
|
657 |
"step": 900
|
658 |
},
|
659 |
{
|
660 |
-
"epoch":
|
661 |
-
"grad_norm":
|
662 |
-
"learning_rate": 1.
|
663 |
-
"loss": 0.
|
664 |
"step": 910
|
665 |
},
|
666 |
{
|
667 |
-
"epoch":
|
668 |
-
"grad_norm":
|
669 |
-
"learning_rate": 1.
|
670 |
-
"loss": 0.
|
671 |
"step": 920
|
672 |
},
|
673 |
{
|
674 |
-
"epoch":
|
675 |
-
"grad_norm":
|
676 |
-
"learning_rate": 1.
|
677 |
-
"loss": 0.
|
678 |
"step": 930
|
679 |
},
|
680 |
{
|
681 |
-
"epoch":
|
682 |
-
"grad_norm": 3.
|
683 |
-
"learning_rate": 1.
|
684 |
-
"loss": 0.
|
685 |
"step": 940
|
686 |
},
|
687 |
{
|
688 |
-
"epoch":
|
689 |
-
"grad_norm":
|
690 |
-
"learning_rate": 1.
|
691 |
-
"loss": 0.
|
692 |
"step": 950
|
693 |
},
|
694 |
{
|
695 |
-
"epoch":
|
696 |
-
"grad_norm":
|
697 |
-
"learning_rate": 1.
|
698 |
-
"loss": 0.
|
699 |
"step": 960
|
700 |
},
|
701 |
{
|
702 |
-
"epoch":
|
703 |
-
"grad_norm": 3.
|
704 |
-
"learning_rate": 1.
|
705 |
-
"loss": 0.
|
706 |
"step": 970
|
707 |
},
|
708 |
{
|
709 |
-
"epoch":
|
710 |
-
"grad_norm":
|
711 |
-
"learning_rate": 1.
|
712 |
-
"loss": 0.
|
713 |
"step": 980
|
714 |
},
|
715 |
{
|
716 |
-
"epoch":
|
717 |
-
"grad_norm":
|
718 |
-
"learning_rate": 1.
|
719 |
-
"loss": 0.
|
720 |
"step": 990
|
721 |
},
|
722 |
{
|
723 |
-
"epoch":
|
724 |
-
"
|
725 |
-
"
|
726 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
727 |
"step": 1000
|
728 |
},
|
729 |
{
|
730 |
-
"epoch":
|
731 |
-
"grad_norm":
|
732 |
-
"learning_rate": 1.
|
733 |
-
"loss": 0.
|
734 |
"step": 1010
|
735 |
},
|
736 |
{
|
737 |
-
"epoch":
|
738 |
-
"grad_norm": 3.
|
739 |
-
"learning_rate": 1.
|
740 |
-
"loss": 0.
|
741 |
"step": 1020
|
742 |
},
|
743 |
{
|
744 |
-
"epoch":
|
745 |
-
"grad_norm":
|
746 |
-
"learning_rate": 1.
|
747 |
-
"loss": 0.
|
748 |
"step": 1030
|
749 |
},
|
750 |
{
|
751 |
-
"epoch":
|
752 |
-
"grad_norm":
|
753 |
-
"learning_rate": 1.
|
754 |
-
"loss": 0.
|
755 |
"step": 1040
|
756 |
},
|
757 |
{
|
758 |
-
"epoch":
|
759 |
-
"grad_norm": 2.
|
760 |
-
"learning_rate": 1.
|
761 |
-
"loss": 0.
|
762 |
"step": 1050
|
763 |
},
|
764 |
{
|
765 |
-
"epoch":
|
766 |
-
"grad_norm":
|
767 |
-
"learning_rate": 1.
|
768 |
-
"loss": 0.
|
769 |
"step": 1060
|
770 |
},
|
771 |
{
|
772 |
-
"epoch":
|
773 |
-
"grad_norm":
|
774 |
-
"learning_rate": 1.
|
775 |
-
"loss": 0.
|
776 |
"step": 1070
|
777 |
},
|
778 |
{
|
779 |
-
"epoch":
|
780 |
-
"grad_norm": 3.
|
781 |
-
"learning_rate": 1.
|
782 |
-
"loss": 0.
|
783 |
"step": 1080
|
784 |
},
|
785 |
{
|
786 |
-
"epoch":
|
787 |
-
"grad_norm":
|
788 |
-
"learning_rate": 1.
|
789 |
-
"loss": 0.
|
790 |
"step": 1090
|
791 |
},
|
792 |
{
|
793 |
-
"epoch":
|
794 |
-
"grad_norm": 3.
|
795 |
-
"learning_rate": 1.
|
796 |
-
"loss": 0.
|
797 |
"step": 1100
|
798 |
},
|
799 |
{
|
800 |
-
"epoch":
|
801 |
-
"grad_norm":
|
802 |
-
"learning_rate": 1.
|
803 |
-
"loss": 0.
|
804 |
"step": 1110
|
805 |
},
|
806 |
{
|
807 |
-
"epoch":
|
808 |
-
"grad_norm": 3.
|
809 |
-
"learning_rate":
|
810 |
-
"loss": 0.
|
811 |
"step": 1120
|
812 |
},
|
813 |
{
|
814 |
-
"epoch":
|
815 |
-
"grad_norm":
|
816 |
-
"learning_rate":
|
817 |
-
"loss": 0.
|
818 |
"step": 1130
|
819 |
},
|
820 |
{
|
821 |
-
"epoch":
|
822 |
-
"grad_norm":
|
823 |
-
"learning_rate":
|
824 |
-
"loss": 0.
|
825 |
"step": 1140
|
826 |
},
|
827 |
{
|
828 |
-
"epoch":
|
829 |
-
"grad_norm":
|
830 |
-
"learning_rate":
|
831 |
-
"loss": 0.
|
832 |
"step": 1150
|
833 |
},
|
834 |
{
|
835 |
-
"epoch":
|
836 |
-
"grad_norm":
|
837 |
-
"learning_rate":
|
838 |
-
"loss": 0.
|
839 |
"step": 1160
|
840 |
},
|
841 |
{
|
842 |
-
"epoch":
|
843 |
-
"grad_norm":
|
844 |
-
"learning_rate":
|
845 |
-
"loss": 0.
|
846 |
"step": 1170
|
847 |
},
|
848 |
{
|
849 |
-
"epoch":
|
850 |
-
"grad_norm":
|
851 |
-
"learning_rate":
|
852 |
-
"loss": 0.
|
853 |
"step": 1180
|
854 |
},
|
855 |
{
|
856 |
-
"epoch":
|
857 |
-
"grad_norm":
|
858 |
-
"learning_rate":
|
859 |
-
"loss": 0.
|
860 |
"step": 1190
|
861 |
},
|
862 |
{
|
863 |
-
"epoch":
|
864 |
-
"
|
865 |
-
"
|
866 |
-
"
|
867 |
-
"
|
|
|
|
|
868 |
},
|
869 |
{
|
870 |
-
"epoch":
|
871 |
-
"
|
872 |
-
"
|
873 |
-
"
|
874 |
-
"eval_samples_per_second": 1430.516,
|
875 |
-
"eval_steps_per_second": 44.822,
|
876 |
"step": 1200
|
877 |
},
|
878 |
{
|
879 |
-
"epoch":
|
880 |
-
"grad_norm":
|
881 |
-
"learning_rate":
|
882 |
-
"loss": 0.
|
883 |
"step": 1210
|
884 |
},
|
885 |
{
|
886 |
-
"epoch":
|
887 |
-
"grad_norm":
|
888 |
-
"learning_rate":
|
889 |
-
"loss": 0.
|
890 |
"step": 1220
|
891 |
},
|
892 |
{
|
893 |
-
"epoch":
|
894 |
-
"grad_norm":
|
895 |
-
"learning_rate":
|
896 |
-
"loss": 0.
|
897 |
"step": 1230
|
898 |
},
|
899 |
{
|
900 |
-
"epoch":
|
901 |
-
"grad_norm":
|
902 |
-
"learning_rate":
|
903 |
-
"loss": 0.
|
904 |
"step": 1240
|
905 |
},
|
906 |
{
|
907 |
-
"epoch":
|
908 |
-
"grad_norm":
|
909 |
-
"learning_rate":
|
910 |
-
"loss": 0.
|
911 |
"step": 1250
|
912 |
},
|
913 |
{
|
914 |
-
"epoch":
|
915 |
-
"grad_norm": 4.
|
916 |
-
"learning_rate":
|
917 |
-
"loss": 0.
|
918 |
"step": 1260
|
919 |
},
|
920 |
{
|
921 |
-
"epoch":
|
922 |
-
"grad_norm":
|
923 |
-
"learning_rate":
|
924 |
-
"loss": 0.
|
925 |
"step": 1270
|
926 |
},
|
927 |
{
|
928 |
-
"epoch":
|
929 |
-
"grad_norm":
|
930 |
-
"learning_rate":
|
931 |
-
"loss": 0.
|
932 |
"step": 1280
|
933 |
},
|
934 |
{
|
935 |
-
"epoch":
|
936 |
-
"grad_norm":
|
937 |
-
"learning_rate":
|
938 |
-
"loss": 0.
|
939 |
"step": 1290
|
940 |
},
|
941 |
{
|
942 |
-
"epoch":
|
943 |
-
"grad_norm":
|
944 |
-
"learning_rate":
|
945 |
-
"loss": 0.
|
946 |
"step": 1300
|
947 |
},
|
948 |
{
|
949 |
-
"epoch":
|
950 |
-
"grad_norm":
|
951 |
-
"learning_rate":
|
952 |
-
"loss": 0.
|
953 |
"step": 1310
|
954 |
},
|
955 |
{
|
956 |
-
"epoch":
|
957 |
-
"grad_norm":
|
958 |
-
"learning_rate":
|
959 |
-
"loss": 0.
|
960 |
"step": 1320
|
961 |
},
|
962 |
{
|
963 |
-
"epoch":
|
964 |
-
"grad_norm": 2.
|
965 |
-
"learning_rate":
|
966 |
-
"loss": 0.
|
967 |
"step": 1330
|
968 |
},
|
969 |
{
|
970 |
-
"epoch":
|
971 |
-
"grad_norm":
|
972 |
-
"learning_rate":
|
973 |
-
"loss": 0.
|
974 |
"step": 1340
|
975 |
},
|
976 |
{
|
977 |
-
"epoch":
|
978 |
-
"grad_norm":
|
979 |
-
"learning_rate":
|
980 |
-
"loss": 0.
|
981 |
"step": 1350
|
982 |
},
|
983 |
{
|
984 |
-
"epoch":
|
985 |
-
"grad_norm":
|
986 |
-
"learning_rate":
|
987 |
-
"loss": 0.
|
988 |
"step": 1360
|
989 |
},
|
990 |
{
|
991 |
-
"epoch":
|
992 |
-
"grad_norm":
|
993 |
-
"learning_rate":
|
994 |
-
"loss": 0.
|
995 |
"step": 1370
|
996 |
},
|
997 |
{
|
998 |
-
"epoch":
|
999 |
-
"grad_norm":
|
1000 |
-
"learning_rate":
|
1001 |
-
"loss": 0.
|
1002 |
"step": 1380
|
1003 |
},
|
1004 |
{
|
1005 |
-
"epoch":
|
1006 |
-
"grad_norm": 2.
|
1007 |
-
"learning_rate":
|
1008 |
-
"loss": 0.
|
1009 |
"step": 1390
|
1010 |
},
|
1011 |
{
|
1012 |
-
"epoch":
|
1013 |
-
"
|
1014 |
-
"
|
1015 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1016 |
"step": 1400
|
1017 |
},
|
1018 |
{
|
1019 |
-
"epoch":
|
1020 |
-
"grad_norm":
|
1021 |
-
"learning_rate":
|
1022 |
-
"loss": 0.
|
1023 |
"step": 1410
|
1024 |
},
|
1025 |
{
|
1026 |
-
"epoch":
|
1027 |
-
"grad_norm":
|
1028 |
-
"learning_rate":
|
1029 |
-
"loss": 0.
|
1030 |
"step": 1420
|
1031 |
},
|
1032 |
{
|
1033 |
-
"epoch":
|
1034 |
-
"grad_norm":
|
1035 |
-
"learning_rate":
|
1036 |
-
"loss": 0.
|
1037 |
"step": 1430
|
1038 |
},
|
1039 |
{
|
1040 |
-
"epoch":
|
1041 |
-
"grad_norm":
|
1042 |
-
"learning_rate":
|
1043 |
-
"loss": 0.
|
1044 |
"step": 1440
|
1045 |
},
|
1046 |
{
|
1047 |
-
"epoch":
|
1048 |
-
"grad_norm":
|
1049 |
-
"learning_rate":
|
1050 |
-
"loss": 0.
|
1051 |
"step": 1450
|
1052 |
},
|
1053 |
{
|
1054 |
-
"epoch":
|
1055 |
-
"grad_norm": 3.
|
1056 |
-
"learning_rate":
|
1057 |
-
"loss": 0.
|
1058 |
"step": 1460
|
1059 |
},
|
1060 |
{
|
1061 |
-
"epoch":
|
1062 |
-
"grad_norm":
|
1063 |
-
"learning_rate":
|
1064 |
-
"loss": 0.
|
1065 |
"step": 1470
|
1066 |
},
|
1067 |
{
|
1068 |
-
"epoch":
|
1069 |
-
"grad_norm":
|
1070 |
-
"learning_rate":
|
1071 |
-
"loss": 0.
|
1072 |
"step": 1480
|
1073 |
},
|
1074 |
{
|
1075 |
-
"epoch":
|
1076 |
-
"grad_norm":
|
1077 |
-
"learning_rate":
|
1078 |
-
"loss": 0.
|
1079 |
"step": 1490
|
1080 |
},
|
1081 |
{
|
1082 |
-
"epoch":
|
1083 |
-
"grad_norm":
|
1084 |
-
"learning_rate":
|
1085 |
-
"loss": 0.
|
1086 |
"step": 1500
|
1087 |
},
|
1088 |
{
|
1089 |
-
"epoch":
|
1090 |
-
"grad_norm":
|
1091 |
-
"learning_rate":
|
1092 |
-
"loss": 0.
|
1093 |
"step": 1510
|
1094 |
},
|
1095 |
{
|
1096 |
-
"epoch":
|
1097 |
-
"grad_norm":
|
1098 |
-
"learning_rate":
|
1099 |
-
"loss": 0.
|
1100 |
"step": 1520
|
1101 |
},
|
1102 |
{
|
1103 |
-
"epoch":
|
1104 |
-
"grad_norm":
|
1105 |
-
"learning_rate":
|
1106 |
-
"loss": 0.
|
1107 |
"step": 1530
|
1108 |
},
|
1109 |
{
|
1110 |
-
"epoch":
|
1111 |
-
"grad_norm":
|
1112 |
-
"learning_rate":
|
1113 |
-
"loss": 0.
|
1114 |
"step": 1540
|
1115 |
},
|
1116 |
{
|
1117 |
-
"epoch":
|
1118 |
-
"grad_norm":
|
1119 |
-
"learning_rate":
|
1120 |
-
"loss": 0.
|
1121 |
"step": 1550
|
1122 |
},
|
1123 |
{
|
1124 |
-
"epoch":
|
1125 |
-
"grad_norm":
|
1126 |
-
"learning_rate":
|
1127 |
-
"loss": 0.
|
1128 |
"step": 1560
|
1129 |
},
|
1130 |
{
|
1131 |
-
"epoch":
|
1132 |
-
"grad_norm":
|
1133 |
-
"learning_rate":
|
1134 |
-
"loss": 0.
|
1135 |
"step": 1570
|
1136 |
},
|
1137 |
{
|
1138 |
-
"epoch":
|
1139 |
-
"grad_norm":
|
1140 |
-
"learning_rate":
|
1141 |
-
"loss": 0.
|
1142 |
"step": 1580
|
1143 |
},
|
1144 |
{
|
1145 |
-
"epoch":
|
1146 |
-
"grad_norm":
|
1147 |
-
"learning_rate":
|
1148 |
-
"loss": 0.
|
1149 |
"step": 1590
|
1150 |
},
|
1151 |
{
|
1152 |
-
"epoch":
|
1153 |
-
"
|
1154 |
-
"
|
1155 |
-
"
|
1156 |
-
"
|
1157 |
-
|
1158 |
-
|
1159 |
-
"epoch": 4.0,
|
1160 |
-
"eval_accuracy": 0.9805825242718447,
|
1161 |
-
"eval_loss": 0.10449180752038956,
|
1162 |
-
"eval_runtime": 5.032,
|
1163 |
-
"eval_samples_per_second": 1350.958,
|
1164 |
-
"eval_steps_per_second": 42.329,
|
1165 |
-
"step": 1600
|
1166 |
-
},
|
1167 |
-
{
|
1168 |
-
"epoch": 4.025046963055729,
|
1169 |
-
"grad_norm": 3.8492510318756104,
|
1170 |
-
"learning_rate": 6.4512534818941505e-06,
|
1171 |
-
"loss": 0.1797,
|
1172 |
-
"step": 1610
|
1173 |
-
},
|
1174 |
-
{
|
1175 |
-
"epoch": 4.050093926111459,
|
1176 |
-
"grad_norm": 4.300637245178223,
|
1177 |
-
"learning_rate": 6.284122562674095e-06,
|
1178 |
-
"loss": 0.1598,
|
1179 |
-
"step": 1620
|
1180 |
-
},
|
1181 |
-
{
|
1182 |
-
"epoch": 4.075140889167188,
|
1183 |
-
"grad_norm": 1.9370712041854858,
|
1184 |
-
"learning_rate": 6.116991643454039e-06,
|
1185 |
-
"loss": 0.1746,
|
1186 |
-
"step": 1630
|
1187 |
-
},
|
1188 |
-
{
|
1189 |
-
"epoch": 4.100187852222918,
|
1190 |
-
"grad_norm": 2.7867672443389893,
|
1191 |
-
"learning_rate": 5.949860724233983e-06,
|
1192 |
-
"loss": 0.1913,
|
1193 |
-
"step": 1640
|
1194 |
-
},
|
1195 |
-
{
|
1196 |
-
"epoch": 4.125234815278647,
|
1197 |
-
"grad_norm": 3.536440372467041,
|
1198 |
-
"learning_rate": 5.782729805013928e-06,
|
1199 |
-
"loss": 0.1731,
|
1200 |
-
"step": 1650
|
1201 |
-
},
|
1202 |
-
{
|
1203 |
-
"epoch": 4.150281778334377,
|
1204 |
-
"grad_norm": 3.2297356128692627,
|
1205 |
-
"learning_rate": 5.615598885793872e-06,
|
1206 |
-
"loss": 0.2019,
|
1207 |
-
"step": 1660
|
1208 |
-
},
|
1209 |
-
{
|
1210 |
-
"epoch": 4.175328741390106,
|
1211 |
-
"grad_norm": 5.139032363891602,
|
1212 |
-
"learning_rate": 5.448467966573816e-06,
|
1213 |
-
"loss": 0.1793,
|
1214 |
-
"step": 1670
|
1215 |
-
},
|
1216 |
-
{
|
1217 |
-
"epoch": 4.200375704445836,
|
1218 |
-
"grad_norm": 3.031764030456543,
|
1219 |
-
"learning_rate": 5.281337047353761e-06,
|
1220 |
-
"loss": 0.1836,
|
1221 |
-
"step": 1680
|
1222 |
-
},
|
1223 |
-
{
|
1224 |
-
"epoch": 4.225422667501565,
|
1225 |
-
"grad_norm": 2.6612586975097656,
|
1226 |
-
"learning_rate": 5.114206128133705e-06,
|
1227 |
-
"loss": 0.1629,
|
1228 |
-
"step": 1690
|
1229 |
-
},
|
1230 |
-
{
|
1231 |
-
"epoch": 4.250469630557295,
|
1232 |
-
"grad_norm": 2.8296072483062744,
|
1233 |
-
"learning_rate": 4.947075208913649e-06,
|
1234 |
-
"loss": 0.1691,
|
1235 |
-
"step": 1700
|
1236 |
-
},
|
1237 |
-
{
|
1238 |
-
"epoch": 4.275516593613024,
|
1239 |
-
"grad_norm": 4.683578968048096,
|
1240 |
-
"learning_rate": 4.7799442896935936e-06,
|
1241 |
-
"loss": 0.1489,
|
1242 |
-
"step": 1710
|
1243 |
-
},
|
1244 |
-
{
|
1245 |
-
"epoch": 4.300563556668754,
|
1246 |
-
"grad_norm": 3.7361319065093994,
|
1247 |
-
"learning_rate": 4.612813370473538e-06,
|
1248 |
-
"loss": 0.1453,
|
1249 |
-
"step": 1720
|
1250 |
-
},
|
1251 |
-
{
|
1252 |
-
"epoch": 4.325610519724483,
|
1253 |
-
"grad_norm": 4.293661117553711,
|
1254 |
-
"learning_rate": 4.445682451253482e-06,
|
1255 |
-
"loss": 0.1445,
|
1256 |
-
"step": 1730
|
1257 |
-
},
|
1258 |
-
{
|
1259 |
-
"epoch": 4.350657482780213,
|
1260 |
-
"grad_norm": 3.1756207942962646,
|
1261 |
-
"learning_rate": 4.278551532033426e-06,
|
1262 |
-
"loss": 0.1523,
|
1263 |
-
"step": 1740
|
1264 |
-
},
|
1265 |
-
{
|
1266 |
-
"epoch": 4.375704445835942,
|
1267 |
-
"grad_norm": 3.921405792236328,
|
1268 |
-
"learning_rate": 4.111420612813371e-06,
|
1269 |
-
"loss": 0.1603,
|
1270 |
-
"step": 1750
|
1271 |
-
},
|
1272 |
-
{
|
1273 |
-
"epoch": 4.400751408891672,
|
1274 |
-
"grad_norm": 2.8336334228515625,
|
1275 |
-
"learning_rate": 3.944289693593315e-06,
|
1276 |
-
"loss": 0.1764,
|
1277 |
-
"step": 1760
|
1278 |
-
},
|
1279 |
-
{
|
1280 |
-
"epoch": 4.425798371947401,
|
1281 |
-
"grad_norm": 3.7013275623321533,
|
1282 |
-
"learning_rate": 3.7771587743732592e-06,
|
1283 |
-
"loss": 0.1481,
|
1284 |
-
"step": 1770
|
1285 |
-
},
|
1286 |
-
{
|
1287 |
-
"epoch": 4.450845335003131,
|
1288 |
-
"grad_norm": 1.9302759170532227,
|
1289 |
-
"learning_rate": 3.6100278551532034e-06,
|
1290 |
-
"loss": 0.1526,
|
1291 |
-
"step": 1780
|
1292 |
-
},
|
1293 |
-
{
|
1294 |
-
"epoch": 4.47589229805886,
|
1295 |
-
"grad_norm": 4.772688865661621,
|
1296 |
-
"learning_rate": 3.4428969359331475e-06,
|
1297 |
-
"loss": 0.1595,
|
1298 |
-
"step": 1790
|
1299 |
-
},
|
1300 |
-
{
|
1301 |
-
"epoch": 4.50093926111459,
|
1302 |
-
"grad_norm": 3.1509013175964355,
|
1303 |
-
"learning_rate": 3.275766016713092e-06,
|
1304 |
-
"loss": 0.1384,
|
1305 |
-
"step": 1800
|
1306 |
-
},
|
1307 |
-
{
|
1308 |
-
"epoch": 4.525986224170319,
|
1309 |
-
"grad_norm": 2.2213551998138428,
|
1310 |
-
"learning_rate": 3.108635097493036e-06,
|
1311 |
-
"loss": 0.1542,
|
1312 |
-
"step": 1810
|
1313 |
-
},
|
1314 |
-
{
|
1315 |
-
"epoch": 4.551033187226049,
|
1316 |
-
"grad_norm": 2.7349748611450195,
|
1317 |
-
"learning_rate": 2.9415041782729803e-06,
|
1318 |
-
"loss": 0.1739,
|
1319 |
-
"step": 1820
|
1320 |
-
},
|
1321 |
-
{
|
1322 |
-
"epoch": 4.576080150281778,
|
1323 |
-
"grad_norm": 5.009521961212158,
|
1324 |
-
"learning_rate": 2.774373259052925e-06,
|
1325 |
-
"loss": 0.1355,
|
1326 |
-
"step": 1830
|
1327 |
-
},
|
1328 |
-
{
|
1329 |
-
"epoch": 4.601127113337508,
|
1330 |
-
"grad_norm": 3.5362050533294678,
|
1331 |
-
"learning_rate": 2.607242339832869e-06,
|
1332 |
-
"loss": 0.1335,
|
1333 |
-
"step": 1840
|
1334 |
-
},
|
1335 |
-
{
|
1336 |
-
"epoch": 4.626174076393237,
|
1337 |
-
"grad_norm": 4.31157922744751,
|
1338 |
-
"learning_rate": 2.4401114206128136e-06,
|
1339 |
-
"loss": 0.1848,
|
1340 |
-
"step": 1850
|
1341 |
-
},
|
1342 |
-
{
|
1343 |
-
"epoch": 4.651221039448966,
|
1344 |
-
"grad_norm": 3.6075448989868164,
|
1345 |
-
"learning_rate": 2.2729805013927577e-06,
|
1346 |
-
"loss": 0.1763,
|
1347 |
-
"step": 1860
|
1348 |
-
},
|
1349 |
-
{
|
1350 |
-
"epoch": 4.676268002504696,
|
1351 |
-
"grad_norm": 3.1636250019073486,
|
1352 |
-
"learning_rate": 2.1058495821727023e-06,
|
1353 |
-
"loss": 0.1672,
|
1354 |
-
"step": 1870
|
1355 |
-
},
|
1356 |
-
{
|
1357 |
-
"epoch": 4.701314965560426,
|
1358 |
-
"grad_norm": 0.7837923169136047,
|
1359 |
-
"learning_rate": 1.9387186629526464e-06,
|
1360 |
-
"loss": 0.1223,
|
1361 |
-
"step": 1880
|
1362 |
-
},
|
1363 |
-
{
|
1364 |
-
"epoch": 4.726361928616155,
|
1365 |
-
"grad_norm": 3.2040278911590576,
|
1366 |
-
"learning_rate": 1.7715877437325906e-06,
|
1367 |
-
"loss": 0.1676,
|
1368 |
-
"step": 1890
|
1369 |
-
},
|
1370 |
-
{
|
1371 |
-
"epoch": 4.751408891671884,
|
1372 |
-
"grad_norm": 2.6808199882507324,
|
1373 |
-
"learning_rate": 1.604456824512535e-06,
|
1374 |
-
"loss": 0.1606,
|
1375 |
-
"step": 1900
|
1376 |
-
},
|
1377 |
-
{
|
1378 |
-
"epoch": 4.776455854727614,
|
1379 |
-
"grad_norm": 2.356374979019165,
|
1380 |
-
"learning_rate": 1.437325905292479e-06,
|
1381 |
-
"loss": 0.1655,
|
1382 |
-
"step": 1910
|
1383 |
-
},
|
1384 |
-
{
|
1385 |
-
"epoch": 4.801502817783343,
|
1386 |
-
"grad_norm": 3.077162504196167,
|
1387 |
-
"learning_rate": 1.2701949860724234e-06,
|
1388 |
-
"loss": 0.1332,
|
1389 |
-
"step": 1920
|
1390 |
-
},
|
1391 |
-
{
|
1392 |
-
"epoch": 4.826549780839073,
|
1393 |
-
"grad_norm": 3.580504894256592,
|
1394 |
-
"learning_rate": 1.1030640668523677e-06,
|
1395 |
-
"loss": 0.1687,
|
1396 |
-
"step": 1930
|
1397 |
-
},
|
1398 |
-
{
|
1399 |
-
"epoch": 4.851596743894802,
|
1400 |
-
"grad_norm": 4.193363189697266,
|
1401 |
-
"learning_rate": 9.35933147632312e-07,
|
1402 |
-
"loss": 0.163,
|
1403 |
-
"step": 1940
|
1404 |
-
},
|
1405 |
-
{
|
1406 |
-
"epoch": 4.876643706950532,
|
1407 |
-
"grad_norm": 3.2785427570343018,
|
1408 |
-
"learning_rate": 7.688022284122563e-07,
|
1409 |
-
"loss": 0.1546,
|
1410 |
-
"step": 1950
|
1411 |
-
},
|
1412 |
-
{
|
1413 |
-
"epoch": 4.9016906700062615,
|
1414 |
-
"grad_norm": 2.1263206005096436,
|
1415 |
-
"learning_rate": 6.016713091922006e-07,
|
1416 |
-
"loss": 0.1445,
|
1417 |
-
"step": 1960
|
1418 |
-
},
|
1419 |
-
{
|
1420 |
-
"epoch": 4.926737633061991,
|
1421 |
-
"grad_norm": 1.8054914474487305,
|
1422 |
-
"learning_rate": 4.3454038997214486e-07,
|
1423 |
-
"loss": 0.1489,
|
1424 |
-
"step": 1970
|
1425 |
-
},
|
1426 |
-
{
|
1427 |
-
"epoch": 4.9517845961177205,
|
1428 |
-
"grad_norm": 2.179539680480957,
|
1429 |
-
"learning_rate": 2.6740947075208915e-07,
|
1430 |
-
"loss": 0.1632,
|
1431 |
-
"step": 1980
|
1432 |
-
},
|
1433 |
-
{
|
1434 |
-
"epoch": 4.97683155917345,
|
1435 |
-
"grad_norm": 2.8142571449279785,
|
1436 |
-
"learning_rate": 1.0027855153203343e-07,
|
1437 |
-
"loss": 0.15,
|
1438 |
-
"step": 1990
|
1439 |
-
},
|
1440 |
-
{
|
1441 |
-
"epoch": 4.989355040701315,
|
1442 |
-
"eval_accuracy": 0.9814651368049426,
|
1443 |
-
"eval_loss": 0.09822726994752884,
|
1444 |
-
"eval_runtime": 5.1844,
|
1445 |
-
"eval_samples_per_second": 1311.237,
|
1446 |
-
"eval_steps_per_second": 41.085,
|
1447 |
-
"step": 1995
|
1448 |
},
|
1449 |
{
|
1450 |
-
"epoch":
|
1451 |
-
"step":
|
1452 |
-
"total_flos":
|
1453 |
-
"train_loss": 0.
|
1454 |
-
"train_runtime":
|
1455 |
-
"train_samples_per_second":
|
1456 |
-
"train_steps_per_second":
|
1457 |
}
|
1458 |
],
|
1459 |
"logging_steps": 10,
|
1460 |
-
"max_steps":
|
1461 |
"num_input_tokens_seen": 0,
|
1462 |
-
"num_train_epochs":
|
1463 |
"save_steps": 500,
|
1464 |
"stateful_callbacks": {
|
1465 |
"TrainerControl": {
|
@@ -1473,8 +1220,8 @@
|
|
1473 |
"attributes": {}
|
1474 |
}
|
1475 |
},
|
1476 |
-
"total_flos":
|
1477 |
-
"train_batch_size":
|
1478 |
"trial_name": null,
|
1479 |
"trial_params": null
|
1480 |
}
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.9830832597822889,
|
3 |
+
"best_model_checkpoint": "wav2vec2-base-ft-keyword-spotting/checkpoint-1393",
|
4 |
+
"epoch": 7.996245306633291,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 1592,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
+
"epoch": 0.05006257822277847,
|
13 |
+
"grad_norm": 2.299422264099121,
|
14 |
+
"learning_rate": 1.875e-06,
|
15 |
+
"loss": 4.1412,
|
16 |
"step": 10
|
17 |
},
|
18 |
{
|
19 |
+
"epoch": 0.10012515644555695,
|
20 |
+
"grad_norm": 3.199314832687378,
|
21 |
+
"learning_rate": 3.75e-06,
|
22 |
+
"loss": 4.1637,
|
23 |
"step": 20
|
24 |
},
|
25 |
{
|
26 |
+
"epoch": 0.15018773466833543,
|
27 |
+
"grad_norm": 3.3083832263946533,
|
28 |
+
"learning_rate": 5.625e-06,
|
29 |
+
"loss": 4.0438,
|
30 |
"step": 30
|
31 |
},
|
32 |
{
|
33 |
+
"epoch": 0.2002503128911139,
|
34 |
+
"grad_norm": 4.229264736175537,
|
35 |
+
"learning_rate": 7.5e-06,
|
36 |
+
"loss": 3.8012,
|
37 |
"step": 40
|
38 |
},
|
39 |
{
|
40 |
+
"epoch": 0.2503128911138924,
|
41 |
+
"grad_norm": 5.718367576599121,
|
42 |
+
"learning_rate": 9.375000000000001e-06,
|
43 |
+
"loss": 3.3779,
|
44 |
"step": 50
|
45 |
},
|
46 |
{
|
47 |
+
"epoch": 0.30037546933667086,
|
48 |
+
"grad_norm": 6.0788254737854,
|
49 |
+
"learning_rate": 1.125e-05,
|
50 |
+
"loss": 2.8533,
|
51 |
"step": 60
|
52 |
},
|
53 |
{
|
54 |
+
"epoch": 0.3504380475594493,
|
55 |
+
"grad_norm": 5.600748538970947,
|
56 |
+
"learning_rate": 1.3125e-05,
|
57 |
+
"loss": 2.4796,
|
58 |
"step": 70
|
59 |
},
|
60 |
{
|
61 |
+
"epoch": 0.4005006257822278,
|
62 |
+
"grad_norm": 5.035912990570068,
|
63 |
+
"learning_rate": 1.5e-05,
|
64 |
+
"loss": 2.253,
|
65 |
"step": 80
|
66 |
},
|
67 |
{
|
68 |
+
"epoch": 0.45056320400500627,
|
69 |
+
"grad_norm": 4.351953983306885,
|
70 |
+
"learning_rate": 1.6875e-05,
|
71 |
+
"loss": 2.0939,
|
72 |
"step": 90
|
73 |
},
|
74 |
{
|
75 |
+
"epoch": 0.5006257822277848,
|
76 |
+
"grad_norm": 3.4278855323791504,
|
77 |
+
"learning_rate": 1.8750000000000002e-05,
|
78 |
+
"loss": 1.9864,
|
79 |
"step": 100
|
80 |
},
|
81 |
{
|
82 |
+
"epoch": 0.5506883604505632,
|
83 |
+
"grad_norm": 2.862748384475708,
|
84 |
+
"learning_rate": 2.0625e-05,
|
85 |
+
"loss": 1.8611,
|
86 |
"step": 110
|
87 |
},
|
88 |
{
|
89 |
+
"epoch": 0.6007509386733417,
|
90 |
+
"grad_norm": 1.2488276958465576,
|
91 |
+
"learning_rate": 2.25e-05,
|
92 |
+
"loss": 1.8302,
|
93 |
"step": 120
|
94 |
},
|
95 |
{
|
96 |
+
"epoch": 0.6508135168961201,
|
97 |
+
"grad_norm": 0.6365911960601807,
|
98 |
+
"learning_rate": 2.4375e-05,
|
99 |
+
"loss": 1.7982,
|
100 |
"step": 130
|
101 |
},
|
102 |
{
|
103 |
+
"epoch": 0.7008760951188986,
|
104 |
+
"grad_norm": 0.5073445439338684,
|
105 |
+
"learning_rate": 2.625e-05,
|
106 |
+
"loss": 1.6792,
|
107 |
"step": 140
|
108 |
},
|
109 |
{
|
110 |
+
"epoch": 0.7509386733416771,
|
111 |
+
"grad_norm": 0.9256235361099243,
|
112 |
+
"learning_rate": 2.8125e-05,
|
113 |
+
"loss": 1.7528,
|
114 |
"step": 150
|
115 |
},
|
116 |
{
|
117 |
+
"epoch": 0.8010012515644556,
|
118 |
+
"grad_norm": 5.669793128967285,
|
119 |
+
"learning_rate": 3e-05,
|
120 |
+
"loss": 1.7688,
|
121 |
"step": 160
|
122 |
},
|
123 |
{
|
124 |
+
"epoch": 0.851063829787234,
|
125 |
+
"grad_norm": 10.126972198486328,
|
126 |
+
"learning_rate": 2.979050279329609e-05,
|
127 |
+
"loss": 1.6876,
|
128 |
"step": 170
|
129 |
},
|
130 |
{
|
131 |
+
"epoch": 0.9011264080100125,
|
132 |
+
"grad_norm": 1.5617619752883911,
|
133 |
+
"learning_rate": 2.958100558659218e-05,
|
134 |
+
"loss": 1.6298,
|
135 |
"step": 180
|
136 |
},
|
137 |
{
|
138 |
+
"epoch": 0.951188986232791,
|
139 |
+
"grad_norm": 1.5987392663955688,
|
140 |
+
"learning_rate": 2.937150837988827e-05,
|
141 |
+
"loss": 1.6106,
|
142 |
"step": 190
|
143 |
},
|
144 |
{
|
145 |
+
"epoch": 0.9962453066332916,
|
146 |
+
"eval_accuracy": 0.6209179170344219,
|
147 |
+
"eval_loss": 1.4251551628112793,
|
148 |
+
"eval_runtime": 4.8605,
|
149 |
+
"eval_samples_per_second": 1398.631,
|
150 |
+
"eval_steps_per_second": 43.823,
|
151 |
+
"step": 199
|
152 |
+
},
|
153 |
+
{
|
154 |
+
"epoch": 1.0050062578222778,
|
155 |
+
"grad_norm": 3.5913760662078857,
|
156 |
+
"learning_rate": 2.9162011173184356e-05,
|
157 |
+
"loss": 1.6815,
|
158 |
"step": 200
|
159 |
},
|
160 |
{
|
161 |
+
"epoch": 1.0550688360450564,
|
162 |
+
"grad_norm": 2.8698642253875732,
|
163 |
+
"learning_rate": 2.895251396648045e-05,
|
164 |
+
"loss": 1.4457,
|
165 |
"step": 210
|
166 |
},
|
167 |
{
|
168 |
+
"epoch": 1.1051314142678348,
|
169 |
+
"grad_norm": 2.3613011837005615,
|
170 |
+
"learning_rate": 2.8743016759776535e-05,
|
171 |
+
"loss": 1.4503,
|
172 |
"step": 220
|
173 |
},
|
174 |
{
|
175 |
+
"epoch": 1.1551939924906134,
|
176 |
+
"grad_norm": 3.4527103900909424,
|
177 |
+
"learning_rate": 2.8533519553072625e-05,
|
178 |
+
"loss": 1.2686,
|
179 |
"step": 230
|
180 |
},
|
181 |
{
|
182 |
+
"epoch": 1.2052565707133918,
|
183 |
+
"grad_norm": 4.879206657409668,
|
184 |
+
"learning_rate": 2.8324022346368715e-05,
|
185 |
+
"loss": 1.2226,
|
186 |
"step": 240
|
187 |
},
|
188 |
{
|
189 |
+
"epoch": 1.2553191489361701,
|
190 |
+
"grad_norm": 3.2351438999176025,
|
191 |
+
"learning_rate": 2.8114525139664805e-05,
|
192 |
+
"loss": 1.1545,
|
193 |
"step": 250
|
194 |
},
|
195 |
{
|
196 |
+
"epoch": 1.3053817271589487,
|
197 |
+
"grad_norm": 5.1034464836120605,
|
198 |
+
"learning_rate": 2.7905027932960894e-05,
|
199 |
+
"loss": 1.1284,
|
200 |
"step": 260
|
201 |
},
|
202 |
{
|
203 |
+
"epoch": 1.355444305381727,
|
204 |
+
"grad_norm": 2.128084421157837,
|
205 |
+
"learning_rate": 2.7695530726256984e-05,
|
206 |
+
"loss": 1.0926,
|
207 |
"step": 270
|
208 |
},
|
209 |
{
|
210 |
+
"epoch": 1.4055068836045057,
|
211 |
+
"grad_norm": 5.853870391845703,
|
212 |
+
"learning_rate": 2.7486033519553074e-05,
|
213 |
+
"loss": 1.075,
|
214 |
"step": 280
|
215 |
},
|
216 |
{
|
217 |
+
"epoch": 1.455569461827284,
|
218 |
+
"grad_norm": 2.4751949310302734,
|
219 |
+
"learning_rate": 2.7276536312849163e-05,
|
220 |
+
"loss": 0.9992,
|
221 |
"step": 290
|
222 |
},
|
223 |
{
|
224 |
+
"epoch": 1.5056320400500627,
|
225 |
+
"grad_norm": 3.3400278091430664,
|
226 |
+
"learning_rate": 2.7067039106145253e-05,
|
227 |
+
"loss": 0.9649,
|
228 |
"step": 300
|
229 |
},
|
230 |
{
|
231 |
+
"epoch": 1.555694618272841,
|
232 |
+
"grad_norm": 2.893463611602783,
|
233 |
+
"learning_rate": 2.685754189944134e-05,
|
234 |
+
"loss": 1.0066,
|
235 |
"step": 310
|
236 |
},
|
237 |
{
|
238 |
+
"epoch": 1.6057571964956194,
|
239 |
+
"grad_norm": 2.179349660873413,
|
240 |
+
"learning_rate": 2.6648044692737432e-05,
|
241 |
+
"loss": 0.9203,
|
242 |
"step": 320
|
243 |
},
|
244 |
{
|
245 |
+
"epoch": 1.655819774718398,
|
246 |
+
"grad_norm": 4.882504463195801,
|
247 |
+
"learning_rate": 2.643854748603352e-05,
|
248 |
+
"loss": 0.9159,
|
249 |
"step": 330
|
250 |
},
|
251 |
{
|
252 |
+
"epoch": 1.7058823529411766,
|
253 |
+
"grad_norm": 4.354543685913086,
|
254 |
+
"learning_rate": 2.622905027932961e-05,
|
255 |
+
"loss": 0.8608,
|
256 |
"step": 340
|
257 |
},
|
258 |
{
|
259 |
+
"epoch": 1.7559449311639548,
|
260 |
+
"grad_norm": 3.870502233505249,
|
261 |
+
"learning_rate": 2.60195530726257e-05,
|
262 |
+
"loss": 0.8231,
|
263 |
"step": 350
|
264 |
},
|
265 |
{
|
266 |
+
"epoch": 1.8060075093867334,
|
267 |
+
"grad_norm": 3.579007148742676,
|
268 |
+
"learning_rate": 2.5810055865921788e-05,
|
269 |
+
"loss": 0.7965,
|
270 |
"step": 360
|
271 |
},
|
272 |
{
|
273 |
+
"epoch": 1.856070087609512,
|
274 |
+
"grad_norm": 4.881648540496826,
|
275 |
+
"learning_rate": 2.5600558659217877e-05,
|
276 |
+
"loss": 0.7647,
|
277 |
"step": 370
|
278 |
},
|
279 |
{
|
280 |
+
"epoch": 1.9061326658322904,
|
281 |
+
"grad_norm": 2.9336421489715576,
|
282 |
+
"learning_rate": 2.5391061452513967e-05,
|
283 |
+
"loss": 0.7005,
|
284 |
"step": 380
|
285 |
},
|
286 |
{
|
287 |
+
"epoch": 1.9561952440550687,
|
288 |
+
"grad_norm": 2.542874813079834,
|
289 |
+
"learning_rate": 2.5181564245810057e-05,
|
290 |
+
"loss": 0.6495,
|
291 |
"step": 390
|
292 |
},
|
293 |
{
|
294 |
+
"epoch": 1.9962453066332917,
|
295 |
+
"eval_accuracy": 0.9682259488084731,
|
296 |
+
"eval_loss": 0.5032486915588379,
|
297 |
+
"eval_runtime": 4.9258,
|
298 |
+
"eval_samples_per_second": 1380.074,
|
299 |
+
"eval_steps_per_second": 43.242,
|
300 |
+
"step": 398
|
301 |
},
|
302 |
{
|
303 |
+
"epoch": 2.0100125156445556,
|
304 |
+
"grad_norm": 1.9056552648544312,
|
305 |
+
"learning_rate": 2.4972067039106143e-05,
|
306 |
+
"loss": 0.6495,
|
|
|
|
|
307 |
"step": 400
|
308 |
},
|
309 |
{
|
310 |
+
"epoch": 2.0600750938673342,
|
311 |
+
"grad_norm": 3.567265033721924,
|
312 |
+
"learning_rate": 2.4762569832402236e-05,
|
313 |
+
"loss": 0.5869,
|
314 |
"step": 410
|
315 |
},
|
316 |
{
|
317 |
+
"epoch": 2.110137672090113,
|
318 |
+
"grad_norm": 2.240018844604492,
|
319 |
+
"learning_rate": 2.4553072625698326e-05,
|
320 |
+
"loss": 0.5728,
|
321 |
"step": 420
|
322 |
},
|
323 |
{
|
324 |
+
"epoch": 2.160200250312891,
|
325 |
+
"grad_norm": 2.6313724517822266,
|
326 |
+
"learning_rate": 2.4343575418994412e-05,
|
327 |
+
"loss": 0.5028,
|
328 |
"step": 430
|
329 |
},
|
330 |
{
|
331 |
+
"epoch": 2.2102628285356696,
|
332 |
+
"grad_norm": 3.360229015350342,
|
333 |
+
"learning_rate": 2.4134078212290505e-05,
|
334 |
+
"loss": 0.4928,
|
335 |
"step": 440
|
336 |
},
|
337 |
{
|
338 |
+
"epoch": 2.260325406758448,
|
339 |
+
"grad_norm": 5.249541282653809,
|
340 |
+
"learning_rate": 2.392458100558659e-05,
|
341 |
+
"loss": 0.4773,
|
342 |
"step": 450
|
343 |
},
|
344 |
{
|
345 |
+
"epoch": 2.3103879849812268,
|
346 |
+
"grad_norm": 3.6117191314697266,
|
347 |
+
"learning_rate": 2.3715083798882685e-05,
|
348 |
+
"loss": 0.4852,
|
349 |
"step": 460
|
350 |
},
|
351 |
{
|
352 |
+
"epoch": 2.360450563204005,
|
353 |
+
"grad_norm": 4.820945739746094,
|
354 |
+
"learning_rate": 2.350558659217877e-05,
|
355 |
+
"loss": 0.4331,
|
356 |
"step": 470
|
357 |
},
|
358 |
{
|
359 |
+
"epoch": 2.4105131414267835,
|
360 |
+
"grad_norm": 4.089610576629639,
|
361 |
+
"learning_rate": 2.329608938547486e-05,
|
362 |
+
"loss": 0.4246,
|
363 |
"step": 480
|
364 |
},
|
365 |
{
|
366 |
+
"epoch": 2.460575719649562,
|
367 |
+
"grad_norm": 4.083464622497559,
|
368 |
+
"learning_rate": 2.308659217877095e-05,
|
369 |
+
"loss": 0.3752,
|
370 |
"step": 490
|
371 |
},
|
372 |
{
|
373 |
+
"epoch": 2.5106382978723403,
|
374 |
+
"grad_norm": 4.422226428985596,
|
375 |
+
"learning_rate": 2.287709497206704e-05,
|
376 |
+
"loss": 0.3916,
|
377 |
"step": 500
|
378 |
},
|
379 |
{
|
380 |
+
"epoch": 2.560700876095119,
|
381 |
+
"grad_norm": 2.952890634536743,
|
382 |
+
"learning_rate": 2.266759776536313e-05,
|
383 |
+
"loss": 0.3973,
|
384 |
"step": 510
|
385 |
},
|
386 |
{
|
387 |
+
"epoch": 2.6107634543178975,
|
388 |
+
"grad_norm": 3.720259428024292,
|
389 |
+
"learning_rate": 2.245810055865922e-05,
|
390 |
+
"loss": 0.3432,
|
391 |
"step": 520
|
392 |
},
|
393 |
{
|
394 |
+
"epoch": 2.660826032540676,
|
395 |
+
"grad_norm": 4.10168981552124,
|
396 |
+
"learning_rate": 2.224860335195531e-05,
|
397 |
+
"loss": 0.3479,
|
398 |
"step": 530
|
399 |
},
|
400 |
{
|
401 |
+
"epoch": 2.710888610763454,
|
402 |
+
"grad_norm": 4.39931058883667,
|
403 |
+
"learning_rate": 2.2039106145251395e-05,
|
404 |
+
"loss": 0.3418,
|
405 |
"step": 540
|
406 |
},
|
407 |
{
|
408 |
+
"epoch": 2.760951188986233,
|
409 |
+
"grad_norm": 2.6174728870391846,
|
410 |
+
"learning_rate": 2.182960893854749e-05,
|
411 |
+
"loss": 0.3153,
|
412 |
"step": 550
|
413 |
},
|
414 |
{
|
415 |
+
"epoch": 2.8110137672090114,
|
416 |
+
"grad_norm": 3.489020347595215,
|
417 |
+
"learning_rate": 2.1620111731843575e-05,
|
418 |
+
"loss": 0.3242,
|
419 |
"step": 560
|
420 |
},
|
421 |
{
|
422 |
+
"epoch": 2.8610763454317896,
|
423 |
+
"grad_norm": 3.2841830253601074,
|
424 |
+
"learning_rate": 2.1410614525139664e-05,
|
425 |
+
"loss": 0.3016,
|
426 |
"step": 570
|
427 |
},
|
428 |
{
|
429 |
+
"epoch": 2.911138923654568,
|
430 |
+
"grad_norm": 4.06994104385376,
|
431 |
+
"learning_rate": 2.1201117318435754e-05,
|
432 |
+
"loss": 0.3237,
|
433 |
"step": 580
|
434 |
},
|
435 |
{
|
436 |
+
"epoch": 2.9612015018773468,
|
437 |
+
"grad_norm": 2.4799962043762207,
|
438 |
+
"learning_rate": 2.0991620111731844e-05,
|
439 |
+
"loss": 0.2978,
|
440 |
"step": 590
|
441 |
},
|
442 |
{
|
443 |
+
"epoch": 2.9962453066332917,
|
444 |
+
"eval_accuracy": 0.9782288908502501,
|
445 |
+
"eval_loss": 0.19027507305145264,
|
446 |
+
"eval_runtime": 4.8737,
|
447 |
+
"eval_samples_per_second": 1394.823,
|
448 |
+
"eval_steps_per_second": 43.704,
|
449 |
+
"step": 597
|
450 |
+
},
|
451 |
+
{
|
452 |
+
"epoch": 3.0150187734668337,
|
453 |
+
"grad_norm": 2.1213083267211914,
|
454 |
+
"learning_rate": 2.0782122905027933e-05,
|
455 |
+
"loss": 0.3231,
|
456 |
"step": 600
|
457 |
},
|
458 |
{
|
459 |
+
"epoch": 3.065081351689612,
|
460 |
+
"grad_norm": 4.8361945152282715,
|
461 |
+
"learning_rate": 2.0572625698324023e-05,
|
462 |
+
"loss": 0.302,
|
463 |
"step": 610
|
464 |
},
|
465 |
{
|
466 |
+
"epoch": 3.1151439299123904,
|
467 |
+
"grad_norm": 3.293104887008667,
|
468 |
+
"learning_rate": 2.0363128491620113e-05,
|
469 |
+
"loss": 0.2883,
|
470 |
"step": 620
|
471 |
},
|
472 |
{
|
473 |
+
"epoch": 3.165206508135169,
|
474 |
+
"grad_norm": 3.274291515350342,
|
475 |
+
"learning_rate": 2.01536312849162e-05,
|
476 |
+
"loss": 0.2961,
|
477 |
"step": 630
|
478 |
},
|
479 |
{
|
480 |
+
"epoch": 3.2152690863579476,
|
481 |
+
"grad_norm": 3.7976105213165283,
|
482 |
+
"learning_rate": 1.9944134078212292e-05,
|
483 |
+
"loss": 0.2688,
|
484 |
"step": 640
|
485 |
},
|
486 |
{
|
487 |
+
"epoch": 3.2653316645807258,
|
488 |
+
"grad_norm": 2.9893229007720947,
|
489 |
+
"learning_rate": 1.973463687150838e-05,
|
490 |
+
"loss": 0.2446,
|
491 |
"step": 650
|
492 |
},
|
493 |
{
|
494 |
+
"epoch": 3.3153942428035044,
|
495 |
+
"grad_norm": 2.2266604900360107,
|
496 |
+
"learning_rate": 1.952513966480447e-05,
|
497 |
+
"loss": 0.2613,
|
498 |
"step": 660
|
499 |
},
|
500 |
{
|
501 |
+
"epoch": 3.365456821026283,
|
502 |
+
"grad_norm": 3.621093511581421,
|
503 |
+
"learning_rate": 1.9315642458100558e-05,
|
504 |
+
"loss": 0.2512,
|
505 |
"step": 670
|
506 |
},
|
507 |
{
|
508 |
+
"epoch": 3.415519399249061,
|
509 |
+
"grad_norm": 2.952971935272217,
|
510 |
+
"learning_rate": 1.9106145251396648e-05,
|
511 |
+
"loss": 0.2536,
|
512 |
"step": 680
|
513 |
},
|
514 |
{
|
515 |
+
"epoch": 3.4655819774718397,
|
516 |
+
"grad_norm": 2.7361905574798584,
|
517 |
+
"learning_rate": 1.889664804469274e-05,
|
518 |
+
"loss": 0.2396,
|
519 |
"step": 690
|
520 |
},
|
521 |
{
|
522 |
+
"epoch": 3.5156445556946183,
|
523 |
+
"grad_norm": 2.3844313621520996,
|
524 |
+
"learning_rate": 1.8687150837988827e-05,
|
525 |
+
"loss": 0.2518,
|
526 |
"step": 700
|
527 |
},
|
528 |
{
|
529 |
+
"epoch": 3.565707133917397,
|
530 |
+
"grad_norm": 3.0508193969726562,
|
531 |
+
"learning_rate": 1.8477653631284917e-05,
|
532 |
+
"loss": 0.2325,
|
533 |
"step": 710
|
534 |
},
|
535 |
{
|
536 |
+
"epoch": 3.615769712140175,
|
537 |
+
"grad_norm": 3.923941135406494,
|
538 |
+
"learning_rate": 1.8268156424581006e-05,
|
539 |
+
"loss": 0.2277,
|
540 |
"step": 720
|
541 |
},
|
542 |
{
|
543 |
+
"epoch": 3.6658322903629537,
|
544 |
+
"grad_norm": 2.638787031173706,
|
545 |
+
"learning_rate": 1.8058659217877096e-05,
|
546 |
+
"loss": 0.2292,
|
547 |
"step": 730
|
548 |
},
|
549 |
{
|
550 |
+
"epoch": 3.7158948685857323,
|
551 |
+
"grad_norm": 2.75313138961792,
|
552 |
+
"learning_rate": 1.7849162011173182e-05,
|
553 |
+
"loss": 0.2364,
|
554 |
"step": 740
|
555 |
},
|
556 |
{
|
557 |
+
"epoch": 3.7659574468085104,
|
558 |
+
"grad_norm": 3.686354398727417,
|
559 |
+
"learning_rate": 1.7639664804469275e-05,
|
560 |
+
"loss": 0.2409,
|
561 |
"step": 750
|
562 |
},
|
563 |
{
|
564 |
+
"epoch": 3.816020025031289,
|
565 |
+
"grad_norm": 4.230103969573975,
|
566 |
+
"learning_rate": 1.7430167597765365e-05,
|
567 |
+
"loss": 0.2293,
|
568 |
"step": 760
|
569 |
},
|
570 |
{
|
571 |
+
"epoch": 3.8660826032540676,
|
572 |
+
"grad_norm": 4.4972100257873535,
|
573 |
+
"learning_rate": 1.722067039106145e-05,
|
574 |
+
"loss": 0.2431,
|
575 |
"step": 770
|
576 |
},
|
577 |
{
|
578 |
+
"epoch": 3.916145181476846,
|
579 |
+
"grad_norm": 3.6224372386932373,
|
580 |
+
"learning_rate": 1.7011173184357544e-05,
|
581 |
+
"loss": 0.2099,
|
582 |
"step": 780
|
583 |
},
|
584 |
{
|
585 |
+
"epoch": 3.966207759699625,
|
586 |
+
"grad_norm": 3.072998285293579,
|
587 |
+
"learning_rate": 1.680167597765363e-05,
|
588 |
+
"loss": 0.2273,
|
589 |
"step": 790
|
590 |
},
|
591 |
{
|
592 |
+
"epoch": 3.9962453066332917,
|
593 |
+
"eval_accuracy": 0.9771991762283024,
|
594 |
+
"eval_loss": 0.14364813268184662,
|
595 |
+
"eval_runtime": 4.9858,
|
596 |
+
"eval_samples_per_second": 1363.468,
|
597 |
+
"eval_steps_per_second": 42.721,
|
598 |
+
"step": 796
|
599 |
},
|
600 |
{
|
601 |
+
"epoch": 4.020025031289111,
|
602 |
+
"grad_norm": 3.3532874584198,
|
603 |
+
"learning_rate": 1.659217877094972e-05,
|
604 |
+
"loss": 0.23,
|
|
|
|
|
605 |
"step": 800
|
606 |
},
|
607 |
{
|
608 |
+
"epoch": 4.07008760951189,
|
609 |
+
"grad_norm": 2.384181261062622,
|
610 |
+
"learning_rate": 1.638268156424581e-05,
|
611 |
+
"loss": 0.2157,
|
612 |
"step": 810
|
613 |
},
|
614 |
{
|
615 |
+
"epoch": 4.1201501877346685,
|
616 |
+
"grad_norm": 4.237877368927002,
|
617 |
+
"learning_rate": 1.61731843575419e-05,
|
618 |
+
"loss": 0.2141,
|
619 |
"step": 820
|
620 |
},
|
621 |
{
|
622 |
+
"epoch": 4.170212765957447,
|
623 |
+
"grad_norm": 3.8752825260162354,
|
624 |
+
"learning_rate": 1.5963687150837986e-05,
|
625 |
+
"loss": 0.214,
|
626 |
"step": 830
|
627 |
},
|
628 |
{
|
629 |
+
"epoch": 4.220275344180226,
|
630 |
+
"grad_norm": 2.571617364883423,
|
631 |
+
"learning_rate": 1.575418994413408e-05,
|
632 |
+
"loss": 0.2124,
|
633 |
"step": 840
|
634 |
},
|
635 |
{
|
636 |
+
"epoch": 4.270337922403003,
|
637 |
+
"grad_norm": 1.8986117839813232,
|
638 |
+
"learning_rate": 1.554469273743017e-05,
|
639 |
+
"loss": 0.2142,
|
640 |
"step": 850
|
641 |
},
|
642 |
{
|
643 |
+
"epoch": 4.320400500625782,
|
644 |
+
"grad_norm": 3.71712589263916,
|
645 |
+
"learning_rate": 1.533519553072626e-05,
|
646 |
+
"loss": 0.1939,
|
647 |
"step": 860
|
648 |
},
|
649 |
{
|
650 |
+
"epoch": 4.370463078848561,
|
651 |
+
"grad_norm": 2.1387696266174316,
|
652 |
+
"learning_rate": 1.5125698324022348e-05,
|
653 |
+
"loss": 0.1905,
|
654 |
"step": 870
|
655 |
},
|
656 |
{
|
657 |
+
"epoch": 4.420525657071339,
|
658 |
+
"grad_norm": 3.3053841590881348,
|
659 |
+
"learning_rate": 1.4916201117318435e-05,
|
660 |
+
"loss": 0.2156,
|
661 |
"step": 880
|
662 |
},
|
663 |
{
|
664 |
+
"epoch": 4.470588235294118,
|
665 |
+
"grad_norm": 2.574657440185547,
|
666 |
+
"learning_rate": 1.4706703910614526e-05,
|
667 |
+
"loss": 0.198,
|
668 |
"step": 890
|
669 |
},
|
670 |
{
|
671 |
+
"epoch": 4.520650813516896,
|
672 |
+
"grad_norm": 2.290309429168701,
|
673 |
+
"learning_rate": 1.4497206703910616e-05,
|
674 |
+
"loss": 0.1754,
|
675 |
"step": 900
|
676 |
},
|
677 |
{
|
678 |
+
"epoch": 4.570713391739675,
|
679 |
+
"grad_norm": 2.4253950119018555,
|
680 |
+
"learning_rate": 1.4287709497206705e-05,
|
681 |
+
"loss": 0.1971,
|
682 |
"step": 910
|
683 |
},
|
684 |
{
|
685 |
+
"epoch": 4.6207759699624535,
|
686 |
+
"grad_norm": 3.070322275161743,
|
687 |
+
"learning_rate": 1.4078212290502793e-05,
|
688 |
+
"loss": 0.1928,
|
689 |
"step": 920
|
690 |
},
|
691 |
{
|
692 |
+
"epoch": 4.670838548185231,
|
693 |
+
"grad_norm": 1.5523446798324585,
|
694 |
+
"learning_rate": 1.3868715083798883e-05,
|
695 |
+
"loss": 0.1948,
|
696 |
"step": 930
|
697 |
},
|
698 |
{
|
699 |
+
"epoch": 4.72090112640801,
|
700 |
+
"grad_norm": 3.076679229736328,
|
701 |
+
"learning_rate": 1.3659217877094973e-05,
|
702 |
+
"loss": 0.1822,
|
703 |
"step": 940
|
704 |
},
|
705 |
{
|
706 |
+
"epoch": 4.7709637046307884,
|
707 |
+
"grad_norm": 3.6825084686279297,
|
708 |
+
"learning_rate": 1.344972067039106e-05,
|
709 |
+
"loss": 0.1954,
|
710 |
"step": 950
|
711 |
},
|
712 |
{
|
713 |
+
"epoch": 4.821026282853567,
|
714 |
+
"grad_norm": 4.037261009216309,
|
715 |
+
"learning_rate": 1.324022346368715e-05,
|
716 |
+
"loss": 0.1946,
|
717 |
"step": 960
|
718 |
},
|
719 |
{
|
720 |
+
"epoch": 4.871088861076346,
|
721 |
+
"grad_norm": 3.026543378829956,
|
722 |
+
"learning_rate": 1.3030726256983242e-05,
|
723 |
+
"loss": 0.2155,
|
724 |
"step": 970
|
725 |
},
|
726 |
{
|
727 |
+
"epoch": 4.921151439299124,
|
728 |
+
"grad_norm": 2.0362164974212646,
|
729 |
+
"learning_rate": 1.282122905027933e-05,
|
730 |
+
"loss": 0.1851,
|
731 |
"step": 980
|
732 |
},
|
733 |
{
|
734 |
+
"epoch": 4.971214017521902,
|
735 |
+
"grad_norm": 1.8788719177246094,
|
736 |
+
"learning_rate": 1.261173184357542e-05,
|
737 |
+
"loss": 0.1866,
|
738 |
"step": 990
|
739 |
},
|
740 |
{
|
741 |
+
"epoch": 4.996245306633291,
|
742 |
+
"eval_accuracy": 0.981759340982642,
|
743 |
+
"eval_loss": 0.11026876419782639,
|
744 |
+
"eval_runtime": 4.7749,
|
745 |
+
"eval_samples_per_second": 1423.702,
|
746 |
+
"eval_steps_per_second": 44.608,
|
747 |
+
"step": 995
|
748 |
+
},
|
749 |
+
{
|
750 |
+
"epoch": 5.025031289111389,
|
751 |
+
"grad_norm": 3.612476110458374,
|
752 |
+
"learning_rate": 1.2402234636871509e-05,
|
753 |
+
"loss": 0.1911,
|
754 |
"step": 1000
|
755 |
},
|
756 |
{
|
757 |
+
"epoch": 5.075093867334168,
|
758 |
+
"grad_norm": 1.3886641263961792,
|
759 |
+
"learning_rate": 1.2192737430167599e-05,
|
760 |
+
"loss": 0.1768,
|
761 |
"step": 1010
|
762 |
},
|
763 |
{
|
764 |
+
"epoch": 5.1251564455569465,
|
765 |
+
"grad_norm": 3.217656135559082,
|
766 |
+
"learning_rate": 1.1983240223463687e-05,
|
767 |
+
"loss": 0.1835,
|
768 |
"step": 1020
|
769 |
},
|
770 |
{
|
771 |
+
"epoch": 5.175219023779725,
|
772 |
+
"grad_norm": 2.281695604324341,
|
773 |
+
"learning_rate": 1.1773743016759776e-05,
|
774 |
+
"loss": 0.1858,
|
775 |
"step": 1030
|
776 |
},
|
777 |
{
|
778 |
+
"epoch": 5.225281602002503,
|
779 |
+
"grad_norm": 2.7055630683898926,
|
780 |
+
"learning_rate": 1.1564245810055866e-05,
|
781 |
+
"loss": 0.1718,
|
782 |
"step": 1040
|
783 |
},
|
784 |
{
|
785 |
+
"epoch": 5.275344180225281,
|
786 |
+
"grad_norm": 2.00937819480896,
|
787 |
+
"learning_rate": 1.1354748603351954e-05,
|
788 |
+
"loss": 0.1479,
|
789 |
"step": 1050
|
790 |
},
|
791 |
{
|
792 |
+
"epoch": 5.32540675844806,
|
793 |
+
"grad_norm": 2.65446138381958,
|
794 |
+
"learning_rate": 1.1145251396648046e-05,
|
795 |
+
"loss": 0.1664,
|
796 |
"step": 1060
|
797 |
},
|
798 |
{
|
799 |
+
"epoch": 5.375469336670839,
|
800 |
+
"grad_norm": 2.499176502227783,
|
801 |
+
"learning_rate": 1.0935754189944135e-05,
|
802 |
+
"loss": 0.1882,
|
803 |
"step": 1070
|
804 |
},
|
805 |
{
|
806 |
+
"epoch": 5.425531914893617,
|
807 |
+
"grad_norm": 3.318516492843628,
|
808 |
+
"learning_rate": 1.0726256983240223e-05,
|
809 |
+
"loss": 0.1823,
|
810 |
"step": 1080
|
811 |
},
|
812 |
{
|
813 |
+
"epoch": 5.475594493116396,
|
814 |
+
"grad_norm": 2.1236233711242676,
|
815 |
+
"learning_rate": 1.0516759776536313e-05,
|
816 |
+
"loss": 0.1657,
|
817 |
"step": 1090
|
818 |
},
|
819 |
{
|
820 |
+
"epoch": 5.5256570713391735,
|
821 |
+
"grad_norm": 3.342689037322998,
|
822 |
+
"learning_rate": 1.0307262569832403e-05,
|
823 |
+
"loss": 0.1823,
|
824 |
"step": 1100
|
825 |
},
|
826 |
{
|
827 |
+
"epoch": 5.575719649561952,
|
828 |
+
"grad_norm": 2.687920331954956,
|
829 |
+
"learning_rate": 1.0097765363128492e-05,
|
830 |
+
"loss": 0.1888,
|
831 |
"step": 1110
|
832 |
},
|
833 |
{
|
834 |
+
"epoch": 5.625782227784731,
|
835 |
+
"grad_norm": 3.692422866821289,
|
836 |
+
"learning_rate": 9.88826815642458e-06,
|
837 |
+
"loss": 0.2053,
|
838 |
"step": 1120
|
839 |
},
|
840 |
{
|
841 |
+
"epoch": 5.675844806007509,
|
842 |
+
"grad_norm": 3.453005790710449,
|
843 |
+
"learning_rate": 9.67877094972067e-06,
|
844 |
+
"loss": 0.179,
|
845 |
"step": 1130
|
846 |
},
|
847 |
{
|
848 |
+
"epoch": 5.725907384230288,
|
849 |
+
"grad_norm": 4.005608081817627,
|
850 |
+
"learning_rate": 9.46927374301676e-06,
|
851 |
+
"loss": 0.1748,
|
852 |
"step": 1140
|
853 |
},
|
854 |
{
|
855 |
+
"epoch": 5.7759699624530665,
|
856 |
+
"grad_norm": 2.1113505363464355,
|
857 |
+
"learning_rate": 9.25977653631285e-06,
|
858 |
+
"loss": 0.1574,
|
859 |
"step": 1150
|
860 |
},
|
861 |
{
|
862 |
+
"epoch": 5.826032540675845,
|
863 |
+
"grad_norm": 4.529311180114746,
|
864 |
+
"learning_rate": 9.050279329608939e-06,
|
865 |
+
"loss": 0.1599,
|
866 |
"step": 1160
|
867 |
},
|
868 |
{
|
869 |
+
"epoch": 5.876095118898624,
|
870 |
+
"grad_norm": 1.885956048965454,
|
871 |
+
"learning_rate": 8.840782122905029e-06,
|
872 |
+
"loss": 0.1909,
|
873 |
"step": 1170
|
874 |
},
|
875 |
{
|
876 |
+
"epoch": 5.926157697121401,
|
877 |
+
"grad_norm": 2.369316816329956,
|
878 |
+
"learning_rate": 8.631284916201118e-06,
|
879 |
+
"loss": 0.1603,
|
880 |
"step": 1180
|
881 |
},
|
882 |
{
|
883 |
+
"epoch": 5.97622027534418,
|
884 |
+
"grad_norm": 1.402648687362671,
|
885 |
+
"learning_rate": 8.421787709497206e-06,
|
886 |
+
"loss": 0.1616,
|
887 |
"step": 1190
|
888 |
},
|
889 |
{
|
890 |
+
"epoch": 5.996245306633291,
|
891 |
+
"eval_accuracy": 0.9819064430714917,
|
892 |
+
"eval_loss": 0.0981006771326065,
|
893 |
+
"eval_runtime": 4.775,
|
894 |
+
"eval_samples_per_second": 1423.653,
|
895 |
+
"eval_steps_per_second": 44.607,
|
896 |
+
"step": 1194
|
897 |
},
|
898 |
{
|
899 |
+
"epoch": 6.030037546933667,
|
900 |
+
"grad_norm": 1.6693238019943237,
|
901 |
+
"learning_rate": 8.212290502793296e-06,
|
902 |
+
"loss": 0.1931,
|
|
|
|
|
903 |
"step": 1200
|
904 |
},
|
905 |
{
|
906 |
+
"epoch": 6.080100125156446,
|
907 |
+
"grad_norm": 2.3462257385253906,
|
908 |
+
"learning_rate": 8.002793296089386e-06,
|
909 |
+
"loss": 0.1506,
|
910 |
"step": 1210
|
911 |
},
|
912 |
{
|
913 |
+
"epoch": 6.130162703379224,
|
914 |
+
"grad_norm": 1.6939945220947266,
|
915 |
+
"learning_rate": 7.793296089385474e-06,
|
916 |
+
"loss": 0.1677,
|
917 |
"step": 1220
|
918 |
},
|
919 |
{
|
920 |
+
"epoch": 6.180225281602002,
|
921 |
+
"grad_norm": 1.728092908859253,
|
922 |
+
"learning_rate": 7.583798882681565e-06,
|
923 |
+
"loss": 0.1569,
|
924 |
"step": 1230
|
925 |
},
|
926 |
{
|
927 |
+
"epoch": 6.230287859824781,
|
928 |
+
"grad_norm": 1.6664111614227295,
|
929 |
+
"learning_rate": 7.374301675977653e-06,
|
930 |
+
"loss": 0.1564,
|
931 |
"step": 1240
|
932 |
},
|
933 |
{
|
934 |
+
"epoch": 6.280350438047559,
|
935 |
+
"grad_norm": 2.0160274505615234,
|
936 |
+
"learning_rate": 7.164804469273744e-06,
|
937 |
+
"loss": 0.1513,
|
938 |
"step": 1250
|
939 |
},
|
940 |
{
|
941 |
+
"epoch": 6.330413016270338,
|
942 |
+
"grad_norm": 4.013051986694336,
|
943 |
+
"learning_rate": 6.9553072625698325e-06,
|
944 |
+
"loss": 0.1594,
|
945 |
"step": 1260
|
946 |
},
|
947 |
{
|
948 |
+
"epoch": 6.380475594493117,
|
949 |
+
"grad_norm": 3.11110258102417,
|
950 |
+
"learning_rate": 6.745810055865922e-06,
|
951 |
+
"loss": 0.1445,
|
952 |
"step": 1270
|
953 |
},
|
954 |
{
|
955 |
+
"epoch": 6.430538172715895,
|
956 |
+
"grad_norm": 3.418999433517456,
|
957 |
+
"learning_rate": 6.536312849162011e-06,
|
958 |
+
"loss": 0.1679,
|
959 |
"step": 1280
|
960 |
},
|
961 |
{
|
962 |
+
"epoch": 6.480600750938673,
|
963 |
+
"grad_norm": 2.4514362812042236,
|
964 |
+
"learning_rate": 6.326815642458101e-06,
|
965 |
+
"loss": 0.152,
|
966 |
"step": 1290
|
967 |
},
|
968 |
{
|
969 |
+
"epoch": 6.5306633291614515,
|
970 |
+
"grad_norm": 3.2242462635040283,
|
971 |
+
"learning_rate": 6.1173184357541904e-06,
|
972 |
+
"loss": 0.1676,
|
973 |
"step": 1300
|
974 |
},
|
975 |
{
|
976 |
+
"epoch": 6.58072590738423,
|
977 |
+
"grad_norm": 4.046393871307373,
|
978 |
+
"learning_rate": 5.907821229050279e-06,
|
979 |
+
"loss": 0.1615,
|
980 |
"step": 1310
|
981 |
},
|
982 |
{
|
983 |
+
"epoch": 6.630788485607009,
|
984 |
+
"grad_norm": 1.9088122844696045,
|
985 |
+
"learning_rate": 5.698324022346369e-06,
|
986 |
+
"loss": 0.1465,
|
987 |
"step": 1320
|
988 |
},
|
989 |
{
|
990 |
+
"epoch": 6.680851063829787,
|
991 |
+
"grad_norm": 2.5699033737182617,
|
992 |
+
"learning_rate": 5.488826815642458e-06,
|
993 |
+
"loss": 0.1472,
|
994 |
"step": 1330
|
995 |
},
|
996 |
{
|
997 |
+
"epoch": 6.730913642052566,
|
998 |
+
"grad_norm": 1.9140872955322266,
|
999 |
+
"learning_rate": 5.2793296089385475e-06,
|
1000 |
+
"loss": 0.1585,
|
1001 |
"step": 1340
|
1002 |
},
|
1003 |
{
|
1004 |
+
"epoch": 6.7809762202753445,
|
1005 |
+
"grad_norm": 2.022095203399658,
|
1006 |
+
"learning_rate": 5.069832402234637e-06,
|
1007 |
+
"loss": 0.1455,
|
1008 |
"step": 1350
|
1009 |
},
|
1010 |
{
|
1011 |
+
"epoch": 6.831038798498122,
|
1012 |
+
"grad_norm": 2.5366971492767334,
|
1013 |
+
"learning_rate": 4.860335195530726e-06,
|
1014 |
+
"loss": 0.1542,
|
1015 |
"step": 1360
|
1016 |
},
|
1017 |
{
|
1018 |
+
"epoch": 6.881101376720901,
|
1019 |
+
"grad_norm": 1.6112697124481201,
|
1020 |
+
"learning_rate": 4.650837988826816e-06,
|
1021 |
+
"loss": 0.1569,
|
1022 |
"step": 1370
|
1023 |
},
|
1024 |
{
|
1025 |
+
"epoch": 6.931163954943679,
|
1026 |
+
"grad_norm": 2.8735201358795166,
|
1027 |
+
"learning_rate": 4.441340782122905e-06,
|
1028 |
+
"loss": 0.1628,
|
1029 |
"step": 1380
|
1030 |
},
|
1031 |
{
|
1032 |
+
"epoch": 6.981226533166458,
|
1033 |
+
"grad_norm": 2.044304132461548,
|
1034 |
+
"learning_rate": 4.231843575418994e-06,
|
1035 |
+
"loss": 0.1385,
|
1036 |
"step": 1390
|
1037 |
},
|
1038 |
{
|
1039 |
+
"epoch": 6.996245306633291,
|
1040 |
+
"eval_accuracy": 0.9830832597822889,
|
1041 |
+
"eval_loss": 0.0956372618675232,
|
1042 |
+
"eval_runtime": 4.8597,
|
1043 |
+
"eval_samples_per_second": 1398.844,
|
1044 |
+
"eval_steps_per_second": 43.83,
|
1045 |
+
"step": 1393
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 7.035043804755945,
|
1049 |
+
"grad_norm": 1.4923343658447266,
|
1050 |
+
"learning_rate": 4.022346368715084e-06,
|
1051 |
+
"loss": 0.1558,
|
1052 |
"step": 1400
|
1053 |
},
|
1054 |
{
|
1055 |
+
"epoch": 7.085106382978723,
|
1056 |
+
"grad_norm": 3.339181423187256,
|
1057 |
+
"learning_rate": 3.812849162011173e-06,
|
1058 |
+
"loss": 0.1391,
|
1059 |
"step": 1410
|
1060 |
},
|
1061 |
{
|
1062 |
+
"epoch": 7.135168961201502,
|
1063 |
+
"grad_norm": 2.091777801513672,
|
1064 |
+
"learning_rate": 3.6033519553072625e-06,
|
1065 |
+
"loss": 0.1541,
|
1066 |
"step": 1420
|
1067 |
},
|
1068 |
{
|
1069 |
+
"epoch": 7.18523153942428,
|
1070 |
+
"grad_norm": 2.6580100059509277,
|
1071 |
+
"learning_rate": 3.393854748603352e-06,
|
1072 |
+
"loss": 0.1379,
|
1073 |
"step": 1430
|
1074 |
},
|
1075 |
{
|
1076 |
+
"epoch": 7.235294117647059,
|
1077 |
+
"grad_norm": 2.4065537452697754,
|
1078 |
+
"learning_rate": 3.1843575418994414e-06,
|
1079 |
+
"loss": 0.1475,
|
1080 |
"step": 1440
|
1081 |
},
|
1082 |
{
|
1083 |
+
"epoch": 7.2853566958698375,
|
1084 |
+
"grad_norm": 2.618218183517456,
|
1085 |
+
"learning_rate": 2.974860335195531e-06,
|
1086 |
+
"loss": 0.1828,
|
1087 |
"step": 1450
|
1088 |
},
|
1089 |
{
|
1090 |
+
"epoch": 7.335419274092616,
|
1091 |
+
"grad_norm": 3.5904743671417236,
|
1092 |
+
"learning_rate": 2.7653631284916204e-06,
|
1093 |
+
"loss": 0.1365,
|
1094 |
"step": 1460
|
1095 |
},
|
1096 |
{
|
1097 |
+
"epoch": 7.385481852315394,
|
1098 |
+
"grad_norm": 2.245260000228882,
|
1099 |
+
"learning_rate": 2.555865921787709e-06,
|
1100 |
+
"loss": 0.1394,
|
1101 |
"step": 1470
|
1102 |
},
|
1103 |
{
|
1104 |
+
"epoch": 7.435544430538172,
|
1105 |
+
"grad_norm": 2.558086395263672,
|
1106 |
+
"learning_rate": 2.346368715083799e-06,
|
1107 |
+
"loss": 0.1533,
|
1108 |
"step": 1480
|
1109 |
},
|
1110 |
{
|
1111 |
+
"epoch": 7.485607008760951,
|
1112 |
+
"grad_norm": 2.851020097732544,
|
1113 |
+
"learning_rate": 2.136871508379888e-06,
|
1114 |
+
"loss": 0.1313,
|
1115 |
"step": 1490
|
1116 |
},
|
1117 |
{
|
1118 |
+
"epoch": 7.53566958698373,
|
1119 |
+
"grad_norm": 1.7011760473251343,
|
1120 |
+
"learning_rate": 1.927374301675978e-06,
|
1121 |
+
"loss": 0.1509,
|
1122 |
"step": 1500
|
1123 |
},
|
1124 |
{
|
1125 |
+
"epoch": 7.585732165206508,
|
1126 |
+
"grad_norm": 2.6264467239379883,
|
1127 |
+
"learning_rate": 1.717877094972067e-06,
|
1128 |
+
"loss": 0.1515,
|
1129 |
"step": 1510
|
1130 |
},
|
1131 |
{
|
1132 |
+
"epoch": 7.635794743429287,
|
1133 |
+
"grad_norm": 1.6332521438598633,
|
1134 |
+
"learning_rate": 1.5083798882681566e-06,
|
1135 |
+
"loss": 0.1489,
|
1136 |
"step": 1520
|
1137 |
},
|
1138 |
{
|
1139 |
+
"epoch": 7.685857321652065,
|
1140 |
+
"grad_norm": 2.0622401237487793,
|
1141 |
+
"learning_rate": 1.2988826815642458e-06,
|
1142 |
+
"loss": 0.1594,
|
1143 |
"step": 1530
|
1144 |
},
|
1145 |
{
|
1146 |
+
"epoch": 7.735919899874844,
|
1147 |
+
"grad_norm": 2.3861618041992188,
|
1148 |
+
"learning_rate": 1.0893854748603353e-06,
|
1149 |
+
"loss": 0.1669,
|
1150 |
"step": 1540
|
1151 |
},
|
1152 |
{
|
1153 |
+
"epoch": 7.785982478097622,
|
1154 |
+
"grad_norm": 4.30822229385376,
|
1155 |
+
"learning_rate": 8.798882681564246e-07,
|
1156 |
+
"loss": 0.1759,
|
1157 |
"step": 1550
|
1158 |
},
|
1159 |
{
|
1160 |
+
"epoch": 7.8360450563204,
|
1161 |
+
"grad_norm": 1.4631046056747437,
|
1162 |
+
"learning_rate": 6.70391061452514e-07,
|
1163 |
+
"loss": 0.1645,
|
1164 |
"step": 1560
|
1165 |
},
|
1166 |
{
|
1167 |
+
"epoch": 7.886107634543179,
|
1168 |
+
"grad_norm": 2.452613115310669,
|
1169 |
+
"learning_rate": 4.608938547486033e-07,
|
1170 |
+
"loss": 0.1575,
|
1171 |
"step": 1570
|
1172 |
},
|
1173 |
{
|
1174 |
+
"epoch": 7.9361702127659575,
|
1175 |
+
"grad_norm": 1.279895305633545,
|
1176 |
+
"learning_rate": 2.5139664804469275e-07,
|
1177 |
+
"loss": 0.149,
|
1178 |
"step": 1580
|
1179 |
},
|
1180 |
{
|
1181 |
+
"epoch": 7.986232790988736,
|
1182 |
+
"grad_norm": 2.141481399536133,
|
1183 |
+
"learning_rate": 4.189944134078212e-08,
|
1184 |
+
"loss": 0.1524,
|
1185 |
"step": 1590
|
1186 |
},
|
1187 |
{
|
1188 |
+
"epoch": 7.996245306633291,
|
1189 |
+
"eval_accuracy": 0.9824948514268903,
|
1190 |
+
"eval_loss": 0.09257339686155319,
|
1191 |
+
"eval_runtime": 5.6151,
|
1192 |
+
"eval_samples_per_second": 1210.67,
|
1193 |
+
"eval_steps_per_second": 37.934,
|
1194 |
+
"step": 1592
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1195 |
},
|
1196 |
{
|
1197 |
+
"epoch": 7.996245306633291,
|
1198 |
+
"step": 1592,
|
1199 |
+
"total_flos": 3.777723239743488e+18,
|
1200 |
+
"train_loss": 0.596273283347787,
|
1201 |
+
"train_runtime": 640.7753,
|
1202 |
+
"train_samples_per_second": 637.902,
|
1203 |
+
"train_steps_per_second": 2.484
|
1204 |
}
|
1205 |
],
|
1206 |
"logging_steps": 10,
|
1207 |
+
"max_steps": 1592,
|
1208 |
"num_input_tokens_seen": 0,
|
1209 |
+
"num_train_epochs": 8,
|
1210 |
"save_steps": 500,
|
1211 |
"stateful_callbacks": {
|
1212 |
"TrainerControl": {
|
|
|
1220 |
"attributes": {}
|
1221 |
}
|
1222 |
},
|
1223 |
+
"total_flos": 3.777723239743488e+18,
|
1224 |
+
"train_batch_size": 64,
|
1225 |
"trial_name": null,
|
1226 |
"trial_params": null
|
1227 |
}
|