DouglasBraga
commited on
Commit
•
6d026a2
1
Parent(s):
c51d1b2
End of training
Browse files- README.md +3 -3
- all_results.json +10 -10
- eval_results.json +6 -6
- train_results.json +5 -5
- trainer_state.json +309 -1905
README.md
CHANGED
@@ -23,7 +23,7 @@ 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 +33,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
33 |
|
34 |
This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on the imagefolder dataset.
|
35 |
It achieves the following results on the evaluation set:
|
36 |
-
- Loss: 0.
|
37 |
-
- Accuracy: 0.
|
38 |
|
39 |
## Model description
|
40 |
|
|
|
23 |
metrics:
|
24 |
- name: Accuracy
|
25 |
type: accuracy
|
26 |
+
value: 0.7025
|
27 |
---
|
28 |
|
29 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
33 |
|
34 |
This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on the imagefolder dataset.
|
35 |
It achieves the following results on the evaluation set:
|
36 |
+
- Loss: 0.8305
|
37 |
+
- Accuracy: 0.7025
|
38 |
|
39 |
## Model description
|
40 |
|
all_results.json
CHANGED
@@ -1,13 +1,13 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_loss": 0.
|
5 |
-
"eval_runtime":
|
6 |
-
"eval_samples_per_second": 25.
|
7 |
-
"eval_steps_per_second": 0.
|
8 |
-
"total_flos":
|
9 |
-
"train_loss": 0.
|
10 |
-
"train_runtime":
|
11 |
-
"train_samples_per_second": 2.
|
12 |
"train_steps_per_second": 0.023
|
13 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 2.9952,
|
3 |
+
"eval_accuracy": 0.7025,
|
4 |
+
"eval_loss": 0.830498456954956,
|
5 |
+
"eval_runtime": 156.0429,
|
6 |
+
"eval_samples_per_second": 25.634,
|
7 |
+
"eval_steps_per_second": 0.801,
|
8 |
+
"total_flos": 2.9779463284174356e+18,
|
9 |
+
"train_loss": 0.36125070786374247,
|
10 |
+
"train_runtime": 41238.7802,
|
11 |
+
"train_samples_per_second": 2.91,
|
12 |
"train_steps_per_second": 0.023
|
13 |
}
|
eval_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_loss": 0.
|
5 |
-
"eval_runtime":
|
6 |
-
"eval_samples_per_second": 25.
|
7 |
-
"eval_steps_per_second": 0.
|
8 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 2.9952,
|
3 |
+
"eval_accuracy": 0.7025,
|
4 |
+
"eval_loss": 0.830498456954956,
|
5 |
+
"eval_runtime": 156.0429,
|
6 |
+
"eval_samples_per_second": 25.634,
|
7 |
+
"eval_steps_per_second": 0.801
|
8 |
}
|
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": 2.
|
7 |
"train_steps_per_second": 0.023
|
8 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 2.9952,
|
3 |
+
"total_flos": 2.9779463284174356e+18,
|
4 |
+
"train_loss": 0.36125070786374247,
|
5 |
+
"train_runtime": 41238.7802,
|
6 |
+
"train_samples_per_second": 2.91,
|
7 |
"train_steps_per_second": 0.023
|
8 |
}
|
trainer_state.json
CHANGED
@@ -1,2301 +1,705 @@
|
|
1 |
{
|
2 |
-
"best_metric": 0.
|
3 |
-
"best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-leukemia-08-2024.v1.2\\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": 0.032,
|
13 |
-
"grad_norm": 5.
|
14 |
-
"learning_rate":
|
15 |
-
"loss": 0.
|
16 |
"step": 10
|
17 |
},
|
18 |
{
|
19 |
"epoch": 0.064,
|
20 |
-
"grad_norm":
|
21 |
-
"learning_rate":
|
22 |
-
"loss": 0.
|
23 |
"step": 20
|
24 |
},
|
25 |
{
|
26 |
"epoch": 0.096,
|
27 |
-
"grad_norm":
|
28 |
-
"learning_rate":
|
29 |
-
"loss": 0.
|
30 |
"step": 30
|
31 |
},
|
32 |
{
|
33 |
"epoch": 0.128,
|
34 |
-
"grad_norm":
|
35 |
-
"learning_rate":
|
36 |
-
"loss": 0.
|
37 |
"step": 40
|
38 |
},
|
39 |
{
|
40 |
"epoch": 0.16,
|
41 |
-
"grad_norm": 5.
|
42 |
-
"learning_rate":
|
43 |
-
"loss": 0.
|
44 |
"step": 50
|
45 |
},
|
46 |
{
|
47 |
"epoch": 0.192,
|
48 |
-
"grad_norm": 9.
|
49 |
-
"learning_rate":
|
50 |
-
"loss": 0.
|
51 |
"step": 60
|
52 |
},
|
53 |
{
|
54 |
"epoch": 0.224,
|
55 |
-
"grad_norm":
|
56 |
-
"learning_rate":
|
57 |
-
"loss": 0.
|
58 |
"step": 70
|
59 |
},
|
60 |
{
|
61 |
"epoch": 0.256,
|
62 |
-
"grad_norm":
|
63 |
-
"learning_rate":
|
64 |
-
"loss": 0.
|
65 |
"step": 80
|
66 |
},
|
67 |
{
|
68 |
"epoch": 0.288,
|
69 |
-
"grad_norm":
|
70 |
-
"learning_rate":
|
71 |
-
"loss": 0.
|
72 |
"step": 90
|
73 |
},
|
74 |
{
|
75 |
"epoch": 0.32,
|
76 |
-
"grad_norm":
|
77 |
-
"learning_rate":
|
78 |
-
"loss": 0.
|
79 |
"step": 100
|
80 |
},
|
81 |
{
|
82 |
"epoch": 0.352,
|
83 |
-
"grad_norm":
|
84 |
-
"learning_rate":
|
85 |
-
"loss": 0.
|
86 |
"step": 110
|
87 |
},
|
88 |
{
|
89 |
"epoch": 0.384,
|
90 |
-
"grad_norm":
|
91 |
-
"learning_rate":
|
92 |
-
"loss": 0.
|
93 |
"step": 120
|
94 |
},
|
95 |
{
|
96 |
"epoch": 0.416,
|
97 |
-
"grad_norm":
|
98 |
-
"learning_rate":
|
99 |
-
"loss": 0.
|
100 |
"step": 130
|
101 |
},
|
102 |
{
|
103 |
"epoch": 0.448,
|
104 |
-
"grad_norm":
|
105 |
-
"learning_rate":
|
106 |
-
"loss": 0.
|
107 |
"step": 140
|
108 |
},
|
109 |
{
|
110 |
"epoch": 0.48,
|
111 |
-
"grad_norm":
|
112 |
-
"learning_rate":
|
113 |
-
"loss": 0.
|
114 |
"step": 150
|
115 |
},
|
116 |
{
|
117 |
"epoch": 0.512,
|
118 |
-
"grad_norm":
|
119 |
-
"learning_rate":
|
120 |
-
"loss": 0.
|
121 |
"step": 160
|
122 |
},
|
123 |
{
|
124 |
"epoch": 0.544,
|
125 |
-
"grad_norm":
|
126 |
-
"learning_rate":
|
127 |
-
"loss": 0.
|
128 |
"step": 170
|
129 |
},
|
130 |
{
|
131 |
"epoch": 0.576,
|
132 |
-
"grad_norm":
|
133 |
-
"learning_rate":
|
134 |
-
"loss": 0.
|
135 |
"step": 180
|
136 |
},
|
137 |
{
|
138 |
"epoch": 0.608,
|
139 |
-
"grad_norm":
|
140 |
-
"learning_rate":
|
141 |
-
"loss": 0.
|
142 |
"step": 190
|
143 |
},
|
144 |
{
|
145 |
"epoch": 0.64,
|
146 |
-
"grad_norm":
|
147 |
-
"learning_rate":
|
148 |
-
"loss": 0.
|
149 |
"step": 200
|
150 |
},
|
151 |
{
|
152 |
"epoch": 0.672,
|
153 |
-
"grad_norm":
|
154 |
-
"learning_rate":
|
155 |
-
"loss": 0.
|
156 |
"step": 210
|
157 |
},
|
158 |
{
|
159 |
"epoch": 0.704,
|
160 |
-
"grad_norm":
|
161 |
-
"learning_rate":
|
162 |
-
"loss": 0.
|
163 |
"step": 220
|
164 |
},
|
165 |
{
|
166 |
"epoch": 0.736,
|
167 |
-
"grad_norm":
|
168 |
-
"learning_rate":
|
169 |
-
"loss": 0.
|
170 |
"step": 230
|
171 |
},
|
172 |
{
|
173 |
"epoch": 0.768,
|
174 |
-
"grad_norm":
|
175 |
-
"learning_rate":
|
176 |
-
"loss": 0.
|
177 |
"step": 240
|
178 |
},
|
179 |
{
|
180 |
"epoch": 0.8,
|
181 |
-
"grad_norm": 7.
|
182 |
-
"learning_rate":
|
183 |
-
"loss": 0.
|
184 |
"step": 250
|
185 |
},
|
186 |
{
|
187 |
"epoch": 0.832,
|
188 |
-
"grad_norm":
|
189 |
-
"learning_rate": 4.
|
190 |
-
"loss": 0.
|
191 |
"step": 260
|
192 |
},
|
193 |
{
|
194 |
"epoch": 0.864,
|
195 |
-
"grad_norm":
|
196 |
-
"learning_rate":
|
197 |
-
"loss": 0.
|
198 |
"step": 270
|
199 |
},
|
200 |
{
|
201 |
"epoch": 0.896,
|
202 |
-
"grad_norm":
|
203 |
-
"learning_rate":
|
204 |
-
"loss": 0.
|
205 |
"step": 280
|
206 |
},
|
207 |
{
|
208 |
"epoch": 0.928,
|
209 |
-
"grad_norm":
|
210 |
-
"learning_rate":
|
211 |
-
"loss": 0.
|
212 |
"step": 290
|
213 |
},
|
214 |
{
|
215 |
"epoch": 0.96,
|
216 |
-
"grad_norm":
|
217 |
-
"learning_rate":
|
218 |
-
"loss": 0.
|
219 |
"step": 300
|
220 |
},
|
221 |
{
|
222 |
"epoch": 0.992,
|
223 |
-
"grad_norm":
|
224 |
-
"learning_rate":
|
225 |
-
"loss": 0.
|
226 |
"step": 310
|
227 |
},
|
228 |
{
|
229 |
"epoch": 0.9984,
|
230 |
-
"eval_accuracy": 0.
|
231 |
-
"eval_loss": 0.
|
232 |
-
"eval_runtime":
|
233 |
-
"eval_samples_per_second":
|
234 |
-
"eval_steps_per_second": 0.
|
235 |
"step": 312
|
236 |
},
|
237 |
{
|
238 |
"epoch": 1.024,
|
239 |
-
"grad_norm":
|
240 |
-
"learning_rate":
|
241 |
-
"loss": 0.
|
242 |
"step": 320
|
243 |
},
|
244 |
{
|
245 |
"epoch": 1.056,
|
246 |
-
"grad_norm":
|
247 |
-
"learning_rate":
|
248 |
-
"loss": 0.
|
249 |
"step": 330
|
250 |
},
|
251 |
{
|
252 |
"epoch": 1.088,
|
253 |
-
"grad_norm":
|
254 |
-
"learning_rate":
|
255 |
-
"loss": 0.
|
256 |
"step": 340
|
257 |
},
|
258 |
{
|
259 |
"epoch": 1.12,
|
260 |
-
"grad_norm":
|
261 |
-
"learning_rate":
|
262 |
-
"loss": 0.
|
263 |
"step": 350
|
264 |
},
|
265 |
{
|
266 |
"epoch": 1.152,
|
267 |
-
"grad_norm":
|
268 |
-
"learning_rate":
|
269 |
-
"loss": 0.
|
270 |
"step": 360
|
271 |
},
|
272 |
{
|
273 |
"epoch": 1.184,
|
274 |
-
"grad_norm":
|
275 |
-
"learning_rate":
|
276 |
-
"loss": 0.
|
277 |
"step": 370
|
278 |
},
|
279 |
{
|
280 |
"epoch": 1.216,
|
281 |
-
"grad_norm":
|
282 |
-
"learning_rate":
|
283 |
-
"loss": 0.
|
284 |
"step": 380
|
285 |
},
|
286 |
{
|
287 |
"epoch": 1.248,
|
288 |
-
"grad_norm":
|
289 |
-
"learning_rate":
|
290 |
-
"loss": 0.
|
291 |
"step": 390
|
292 |
},
|
293 |
{
|
294 |
"epoch": 1.28,
|
295 |
-
"grad_norm":
|
296 |
-
"learning_rate":
|
297 |
-
"loss": 0.
|
298 |
"step": 400
|
299 |
},
|
300 |
{
|
301 |
"epoch": 1.312,
|
302 |
-
"grad_norm":
|
303 |
-
"learning_rate":
|
304 |
-
"loss": 0.
|
305 |
"step": 410
|
306 |
},
|
307 |
{
|
308 |
"epoch": 1.3439999999999999,
|
309 |
-
"grad_norm":
|
310 |
-
"learning_rate":
|
311 |
-
"loss": 0.
|
312 |
"step": 420
|
313 |
},
|
314 |
{
|
315 |
"epoch": 1.376,
|
316 |
-
"grad_norm":
|
317 |
-
"learning_rate":
|
318 |
-
"loss": 0.
|
319 |
"step": 430
|
320 |
},
|
321 |
{
|
322 |
"epoch": 1.408,
|
323 |
-
"grad_norm": 9.
|
324 |
-
"learning_rate":
|
325 |
-
"loss": 0.
|
326 |
"step": 440
|
327 |
},
|
328 |
{
|
329 |
"epoch": 1.44,
|
330 |
-
"grad_norm":
|
331 |
-
"learning_rate":
|
332 |
-
"loss": 0.
|
333 |
"step": 450
|
334 |
},
|
335 |
{
|
336 |
"epoch": 1.472,
|
337 |
-
"grad_norm":
|
338 |
-
"learning_rate":
|
339 |
-
"loss": 0.
|
340 |
"step": 460
|
341 |
},
|
342 |
{
|
343 |
"epoch": 1.504,
|
344 |
-
"grad_norm":
|
345 |
-
"learning_rate":
|
346 |
-
"loss": 0.
|
347 |
"step": 470
|
348 |
},
|
349 |
{
|
350 |
"epoch": 1.536,
|
351 |
-
"grad_norm":
|
352 |
-
"learning_rate":
|
353 |
-
"loss": 0.
|
354 |
"step": 480
|
355 |
},
|
356 |
{
|
357 |
"epoch": 1.568,
|
358 |
-
"grad_norm":
|
359 |
-
"learning_rate":
|
360 |
-
"loss": 0.
|
361 |
"step": 490
|
362 |
},
|
363 |
{
|
364 |
"epoch": 1.6,
|
365 |
-
"grad_norm":
|
366 |
-
"learning_rate":
|
367 |
-
"loss": 0.
|
368 |
"step": 500
|
369 |
},
|
370 |
{
|
371 |
"epoch": 1.6320000000000001,
|
372 |
-
"grad_norm":
|
373 |
-
"learning_rate":
|
374 |
-
"loss": 0.
|
375 |
"step": 510
|
376 |
},
|
377 |
{
|
378 |
"epoch": 1.6640000000000001,
|
379 |
-
"grad_norm":
|
380 |
-
"learning_rate":
|
381 |
-
"loss": 0.
|
382 |
"step": 520
|
383 |
},
|
384 |
{
|
385 |
"epoch": 1.696,
|
386 |
-
"grad_norm":
|
387 |
-
"learning_rate":
|
388 |
-
"loss": 0.
|
389 |
"step": 530
|
390 |
},
|
391 |
{
|
392 |
"epoch": 1.728,
|
393 |
-
"grad_norm":
|
394 |
-
"learning_rate":
|
395 |
-
"loss": 0.
|
396 |
"step": 540
|
397 |
},
|
398 |
{
|
399 |
"epoch": 1.76,
|
400 |
-
"grad_norm":
|
401 |
-
"learning_rate":
|
402 |
-
"loss": 0.
|
403 |
"step": 550
|
404 |
},
|
405 |
{
|
406 |
"epoch": 1.792,
|
407 |
-
"grad_norm":
|
408 |
-
"learning_rate":
|
409 |
-
"loss": 0.
|
410 |
"step": 560
|
411 |
},
|
412 |
{
|
413 |
"epoch": 1.8239999999999998,
|
414 |
-
"grad_norm":
|
415 |
-
"learning_rate":
|
416 |
-
"loss": 0.
|
417 |
"step": 570
|
418 |
},
|
419 |
{
|
420 |
"epoch": 1.8559999999999999,
|
421 |
-
"grad_norm":
|
422 |
-
"learning_rate":
|
423 |
-
"loss": 0.
|
424 |
"step": 580
|
425 |
},
|
426 |
{
|
427 |
"epoch": 1.888,
|
428 |
-
"grad_norm": 9.
|
429 |
-
"learning_rate":
|
430 |
-
"loss": 0.
|
431 |
"step": 590
|
432 |
},
|
433 |
{
|
434 |
"epoch": 1.92,
|
435 |
-
"grad_norm":
|
436 |
-
"learning_rate":
|
437 |
-
"loss": 0.
|
438 |
"step": 600
|
439 |
},
|
440 |
{
|
441 |
"epoch": 1.952,
|
442 |
-
"grad_norm":
|
443 |
-
"learning_rate":
|
444 |
-
"loss": 0.
|
445 |
"step": 610
|
446 |
},
|
447 |
{
|
448 |
"epoch": 1.984,
|
449 |
-
"grad_norm":
|
450 |
-
"learning_rate":
|
451 |
-
"loss": 0.
|
452 |
"step": 620
|
453 |
},
|
454 |
{
|
455 |
"epoch": 2.0,
|
456 |
-
"eval_accuracy": 0.
|
457 |
-
"eval_loss": 0.
|
458 |
-
"eval_runtime":
|
459 |
-
"eval_samples_per_second": 25.
|
460 |
-
"eval_steps_per_second": 0.
|
461 |
"step": 625
|
462 |
},
|
463 |
{
|
464 |
"epoch": 2.016,
|
465 |
-
"grad_norm":
|
466 |
-
"learning_rate":
|
467 |
-
"loss": 0.
|
468 |
"step": 630
|
469 |
},
|
470 |
{
|
471 |
"epoch": 2.048,
|
472 |
-
"grad_norm":
|
473 |
-
"learning_rate":
|
474 |
-
"loss": 0.
|
475 |
"step": 640
|
476 |
},
|
477 |
{
|
478 |
"epoch": 2.08,
|
479 |
-
"grad_norm":
|
480 |
-
"learning_rate":
|
481 |
-
"loss": 0.
|
482 |
"step": 650
|
483 |
},
|
484 |
{
|
485 |
"epoch": 2.112,
|
486 |
-
"grad_norm":
|
487 |
-
"learning_rate":
|
488 |
-
"loss": 0.
|
489 |
"step": 660
|
490 |
},
|
491 |
{
|
492 |
"epoch": 2.144,
|
493 |
-
"grad_norm":
|
494 |
-
"learning_rate":
|
495 |
-
"loss": 0.
|
496 |
"step": 670
|
497 |
},
|
498 |
{
|
499 |
"epoch": 2.176,
|
500 |
-
"grad_norm":
|
501 |
-
"learning_rate":
|
502 |
-
"loss": 0.
|
503 |
"step": 680
|
504 |
},
|
505 |
{
|
506 |
"epoch": 2.208,
|
507 |
-
"grad_norm":
|
508 |
-
"learning_rate":
|
509 |
-
"loss": 0.
|
510 |
"step": 690
|
511 |
},
|
512 |
{
|
513 |
"epoch": 2.24,
|
514 |
-
"grad_norm":
|
515 |
-
"learning_rate":
|
516 |
-
"loss": 0.
|
517 |
"step": 700
|
518 |
},
|
519 |
{
|
520 |
"epoch": 2.2720000000000002,
|
521 |
-
"grad_norm":
|
522 |
-
"learning_rate":
|
523 |
-
"loss": 0.
|
524 |
"step": 710
|
525 |
},
|
526 |
{
|
527 |
"epoch": 2.304,
|
528 |
-
"grad_norm":
|
529 |
-
"learning_rate":
|
530 |
-
"loss": 0.
|
531 |
"step": 720
|
532 |
},
|
533 |
{
|
534 |
"epoch": 2.336,
|
535 |
-
"grad_norm":
|
536 |
-
"learning_rate":
|
537 |
-
"loss": 0.
|
538 |
"step": 730
|
539 |
},
|
540 |
{
|
541 |
"epoch": 2.368,
|
542 |
-
"grad_norm":
|
543 |
-
"learning_rate":
|
544 |
-
"loss": 0.
|
545 |
"step": 740
|
546 |
},
|
547 |
{
|
548 |
"epoch": 2.4,
|
549 |
-
"grad_norm":
|
550 |
-
"learning_rate":
|
551 |
-
"loss": 0.
|
552 |
"step": 750
|
553 |
},
|
554 |
{
|
555 |
"epoch": 2.432,
|
556 |
-
"grad_norm":
|
557 |
-
"learning_rate":
|
558 |
-
"loss": 0.
|
559 |
"step": 760
|
560 |
},
|
561 |
{
|
562 |
"epoch": 2.464,
|
563 |
-
"grad_norm":
|
564 |
-
"learning_rate":
|
565 |
-
"loss": 0.
|
566 |
"step": 770
|
567 |
},
|
568 |
{
|
569 |
"epoch": 2.496,
|
570 |
-
"grad_norm":
|
571 |
-
"learning_rate":
|
572 |
-
"loss": 0.
|
573 |
"step": 780
|
574 |
},
|
575 |
{
|
576 |
"epoch": 2.528,
|
577 |
-
"grad_norm":
|
578 |
-
"learning_rate":
|
579 |
-
"loss": 0.
|
580 |
"step": 790
|
581 |
},
|
582 |
{
|
583 |
"epoch": 2.56,
|
584 |
-
"grad_norm": 7.
|
585 |
-
"learning_rate":
|
586 |
-
"loss": 0.
|
587 |
"step": 800
|
588 |
},
|
589 |
{
|
590 |
"epoch": 2.592,
|
591 |
-
"grad_norm":
|
592 |
-
"learning_rate":
|
593 |
-
"loss": 0.
|
594 |
"step": 810
|
595 |
},
|
596 |
{
|
597 |
"epoch": 2.624,
|
598 |
-
"grad_norm":
|
599 |
-
"learning_rate":
|
600 |
-
"loss": 0.
|
601 |
"step": 820
|
602 |
},
|
603 |
{
|
604 |
"epoch": 2.656,
|
605 |
-
"grad_norm":
|
606 |
-
"learning_rate":
|
607 |
-
"loss": 0.
|
608 |
"step": 830
|
609 |
},
|
610 |
{
|
611 |
"epoch": 2.6879999999999997,
|
612 |
-
"grad_norm":
|
613 |
-
"learning_rate":
|
614 |
-
"loss": 0.
|
615 |
"step": 840
|
616 |
},
|
617 |
{
|
618 |
"epoch": 2.7199999999999998,
|
619 |
-
"grad_norm":
|
620 |
-
"learning_rate":
|
621 |
-
"loss": 0.
|
622 |
"step": 850
|
623 |
},
|
624 |
{
|
625 |
"epoch": 2.752,
|
626 |
-
"grad_norm":
|
627 |
-
"learning_rate": 4.
|
628 |
-
"loss": 0.
|
629 |
"step": 860
|
630 |
},
|
631 |
{
|
632 |
"epoch": 2.784,
|
633 |
-
"grad_norm":
|
634 |
-
"learning_rate": 4.
|
635 |
-
"loss": 0.
|
636 |
"step": 870
|
637 |
},
|
638 |
{
|
639 |
"epoch": 2.816,
|
640 |
-
"grad_norm": 12.
|
641 |
-
"learning_rate": 3.
|
642 |
-
"loss": 0.
|
643 |
"step": 880
|
644 |
},
|
645 |
{
|
646 |
"epoch": 2.848,
|
647 |
-
"grad_norm":
|
648 |
-
"learning_rate":
|
649 |
-
"loss": 0.
|
650 |
"step": 890
|
651 |
},
|
652 |
{
|
653 |
"epoch": 2.88,
|
654 |
-
"grad_norm":
|
655 |
-
"learning_rate":
|
656 |
-
"loss": 0.
|
657 |
"step": 900
|
658 |
},
|
659 |
{
|
660 |
"epoch": 2.912,
|
661 |
-
"grad_norm":
|
662 |
-
"learning_rate":
|
663 |
-
"loss": 0.
|
664 |
"step": 910
|
665 |
},
|
666 |
{
|
667 |
"epoch": 2.944,
|
668 |
-
"grad_norm":
|
669 |
-
"learning_rate":
|
670 |
-
"loss": 0.
|
671 |
"step": 920
|
672 |
},
|
673 |
{
|
674 |
"epoch": 2.976,
|
675 |
-
"grad_norm":
|
676 |
-
"learning_rate":
|
677 |
-
"loss": 0.
|
678 |
"step": 930
|
679 |
},
|
680 |
{
|
681 |
-
"epoch": 2.
|
682 |
-
"eval_accuracy": 0.
|
683 |
-
"eval_loss":
|
684 |
-
"eval_runtime": 155.
|
685 |
-
"eval_samples_per_second": 25.
|
686 |
-
"eval_steps_per_second": 0.
|
687 |
-
"step":
|
688 |
-
},
|
689 |
-
{
|
690 |
-
"epoch": 3.008,
|
691 |
-
"grad_norm": 51.52549743652344,
|
692 |
-
"learning_rate": 3.8871082621082625e-05,
|
693 |
-
"loss": 0.3457,
|
694 |
-
"step": 940
|
695 |
-
},
|
696 |
-
{
|
697 |
-
"epoch": 3.04,
|
698 |
-
"grad_norm": 22.345596313476562,
|
699 |
-
"learning_rate": 3.869301994301994e-05,
|
700 |
-
"loss": 0.3204,
|
701 |
-
"step": 950
|
702 |
-
},
|
703 |
-
{
|
704 |
-
"epoch": 3.072,
|
705 |
-
"grad_norm": 16.317461013793945,
|
706 |
-
"learning_rate": 3.851495726495727e-05,
|
707 |
-
"loss": 0.3101,
|
708 |
-
"step": 960
|
709 |
-
},
|
710 |
-
{
|
711 |
-
"epoch": 3.104,
|
712 |
-
"grad_norm": 27.974485397338867,
|
713 |
-
"learning_rate": 3.8336894586894586e-05,
|
714 |
-
"loss": 0.2918,
|
715 |
-
"step": 970
|
716 |
-
},
|
717 |
-
{
|
718 |
-
"epoch": 3.136,
|
719 |
-
"grad_norm": 23.675912857055664,
|
720 |
-
"learning_rate": 3.815883190883191e-05,
|
721 |
-
"loss": 0.2817,
|
722 |
-
"step": 980
|
723 |
-
},
|
724 |
-
{
|
725 |
-
"epoch": 3.168,
|
726 |
-
"grad_norm": 30.348739624023438,
|
727 |
-
"learning_rate": 3.798076923076923e-05,
|
728 |
-
"loss": 0.2431,
|
729 |
-
"step": 990
|
730 |
-
},
|
731 |
-
{
|
732 |
-
"epoch": 3.2,
|
733 |
-
"grad_norm": 11.50307846069336,
|
734 |
-
"learning_rate": 3.7802706552706555e-05,
|
735 |
-
"loss": 0.2584,
|
736 |
-
"step": 1000
|
737 |
-
},
|
738 |
-
{
|
739 |
-
"epoch": 3.232,
|
740 |
-
"grad_norm": 19.67528533935547,
|
741 |
-
"learning_rate": 3.762464387464388e-05,
|
742 |
-
"loss": 0.2714,
|
743 |
-
"step": 1010
|
744 |
-
},
|
745 |
-
{
|
746 |
-
"epoch": 3.2640000000000002,
|
747 |
-
"grad_norm": 22.57136344909668,
|
748 |
-
"learning_rate": 3.74465811965812e-05,
|
749 |
-
"loss": 0.2544,
|
750 |
-
"step": 1020
|
751 |
-
},
|
752 |
-
{
|
753 |
-
"epoch": 3.296,
|
754 |
-
"grad_norm": 16.600223541259766,
|
755 |
-
"learning_rate": 3.726851851851852e-05,
|
756 |
-
"loss": 0.2469,
|
757 |
-
"step": 1030
|
758 |
-
},
|
759 |
-
{
|
760 |
-
"epoch": 3.328,
|
761 |
-
"grad_norm": 13.466314315795898,
|
762 |
-
"learning_rate": 3.709045584045584e-05,
|
763 |
-
"loss": 0.3036,
|
764 |
-
"step": 1040
|
765 |
-
},
|
766 |
-
{
|
767 |
-
"epoch": 3.36,
|
768 |
-
"grad_norm": 21.5009765625,
|
769 |
-
"learning_rate": 3.691239316239317e-05,
|
770 |
-
"loss": 0.2864,
|
771 |
-
"step": 1050
|
772 |
-
},
|
773 |
-
{
|
774 |
-
"epoch": 3.392,
|
775 |
-
"grad_norm": 7.8777008056640625,
|
776 |
-
"learning_rate": 3.6734330484330485e-05,
|
777 |
-
"loss": 0.2907,
|
778 |
-
"step": 1060
|
779 |
-
},
|
780 |
-
{
|
781 |
-
"epoch": 3.424,
|
782 |
-
"grad_norm": 14.131542205810547,
|
783 |
-
"learning_rate": 3.655626780626781e-05,
|
784 |
-
"loss": 0.2849,
|
785 |
-
"step": 1070
|
786 |
-
},
|
787 |
-
{
|
788 |
-
"epoch": 3.456,
|
789 |
-
"grad_norm": 11.86939525604248,
|
790 |
-
"learning_rate": 3.637820512820513e-05,
|
791 |
-
"loss": 0.2659,
|
792 |
-
"step": 1080
|
793 |
-
},
|
794 |
-
{
|
795 |
-
"epoch": 3.488,
|
796 |
-
"grad_norm": 8.989656448364258,
|
797 |
-
"learning_rate": 3.6200142450142454e-05,
|
798 |
-
"loss": 0.2714,
|
799 |
-
"step": 1090
|
800 |
-
},
|
801 |
-
{
|
802 |
-
"epoch": 3.52,
|
803 |
-
"grad_norm": 22.53978729248047,
|
804 |
-
"learning_rate": 3.602207977207977e-05,
|
805 |
-
"loss": 0.2592,
|
806 |
-
"step": 1100
|
807 |
-
},
|
808 |
-
{
|
809 |
-
"epoch": 3.552,
|
810 |
-
"grad_norm": 11.309256553649902,
|
811 |
-
"learning_rate": 3.58440170940171e-05,
|
812 |
-
"loss": 0.248,
|
813 |
-
"step": 1110
|
814 |
-
},
|
815 |
-
{
|
816 |
-
"epoch": 3.584,
|
817 |
-
"grad_norm": 8.17591381072998,
|
818 |
-
"learning_rate": 3.5665954415954415e-05,
|
819 |
-
"loss": 0.2481,
|
820 |
-
"step": 1120
|
821 |
-
},
|
822 |
-
{
|
823 |
-
"epoch": 3.616,
|
824 |
-
"grad_norm": 44.116905212402344,
|
825 |
-
"learning_rate": 3.548789173789174e-05,
|
826 |
-
"loss": 0.2424,
|
827 |
-
"step": 1130
|
828 |
-
},
|
829 |
-
{
|
830 |
-
"epoch": 3.648,
|
831 |
-
"grad_norm": 15.891951560974121,
|
832 |
-
"learning_rate": 3.5309829059829066e-05,
|
833 |
-
"loss": 0.2467,
|
834 |
-
"step": 1140
|
835 |
-
},
|
836 |
-
{
|
837 |
-
"epoch": 3.68,
|
838 |
-
"grad_norm": 9.102974891662598,
|
839 |
-
"learning_rate": 3.5131766381766384e-05,
|
840 |
-
"loss": 0.2456,
|
841 |
-
"step": 1150
|
842 |
-
},
|
843 |
-
{
|
844 |
-
"epoch": 3.7119999999999997,
|
845 |
-
"grad_norm": 14.138261795043945,
|
846 |
-
"learning_rate": 3.49537037037037e-05,
|
847 |
-
"loss": 0.2479,
|
848 |
-
"step": 1160
|
849 |
-
},
|
850 |
-
{
|
851 |
-
"epoch": 3.7439999999999998,
|
852 |
-
"grad_norm": 37.55530548095703,
|
853 |
-
"learning_rate": 3.477564102564103e-05,
|
854 |
-
"loss": 0.2604,
|
855 |
-
"step": 1170
|
856 |
-
},
|
857 |
-
{
|
858 |
-
"epoch": 3.776,
|
859 |
-
"grad_norm": 20.72405242919922,
|
860 |
-
"learning_rate": 3.4597578347578346e-05,
|
861 |
-
"loss": 0.2392,
|
862 |
-
"step": 1180
|
863 |
-
},
|
864 |
-
{
|
865 |
-
"epoch": 3.808,
|
866 |
-
"grad_norm": 9.120800018310547,
|
867 |
-
"learning_rate": 3.441951566951567e-05,
|
868 |
-
"loss": 0.2243,
|
869 |
-
"step": 1190
|
870 |
-
},
|
871 |
-
{
|
872 |
-
"epoch": 3.84,
|
873 |
-
"grad_norm": 5.410057067871094,
|
874 |
-
"learning_rate": 3.4241452991452996e-05,
|
875 |
-
"loss": 0.2796,
|
876 |
-
"step": 1200
|
877 |
-
},
|
878 |
-
{
|
879 |
-
"epoch": 3.872,
|
880 |
-
"grad_norm": 13.668625831604004,
|
881 |
-
"learning_rate": 3.4063390313390314e-05,
|
882 |
-
"loss": 0.2386,
|
883 |
-
"step": 1210
|
884 |
-
},
|
885 |
-
{
|
886 |
-
"epoch": 3.904,
|
887 |
-
"grad_norm": 15.398672103881836,
|
888 |
-
"learning_rate": 3.388532763532763e-05,
|
889 |
-
"loss": 0.2461,
|
890 |
-
"step": 1220
|
891 |
-
},
|
892 |
-
{
|
893 |
-
"epoch": 3.936,
|
894 |
-
"grad_norm": 21.709699630737305,
|
895 |
-
"learning_rate": 3.3707264957264964e-05,
|
896 |
-
"loss": 0.2225,
|
897 |
-
"step": 1230
|
898 |
-
},
|
899 |
-
{
|
900 |
-
"epoch": 3.968,
|
901 |
-
"grad_norm": 6.977370262145996,
|
902 |
-
"learning_rate": 3.352920227920228e-05,
|
903 |
-
"loss": 0.223,
|
904 |
-
"step": 1240
|
905 |
-
},
|
906 |
-
{
|
907 |
-
"epoch": 4.0,
|
908 |
-
"grad_norm": 8.210868835449219,
|
909 |
-
"learning_rate": 3.33511396011396e-05,
|
910 |
-
"loss": 0.2529,
|
911 |
-
"step": 1250
|
912 |
-
},
|
913 |
-
{
|
914 |
-
"epoch": 4.0,
|
915 |
-
"eval_accuracy": 0.713,
|
916 |
-
"eval_loss": 0.8170278668403625,
|
917 |
-
"eval_runtime": 160.9705,
|
918 |
-
"eval_samples_per_second": 24.849,
|
919 |
-
"eval_steps_per_second": 0.777,
|
920 |
-
"step": 1250
|
921 |
-
},
|
922 |
-
{
|
923 |
-
"epoch": 4.032,
|
924 |
-
"grad_norm": 10.20702075958252,
|
925 |
-
"learning_rate": 3.3173076923076926e-05,
|
926 |
-
"loss": 0.2091,
|
927 |
-
"step": 1260
|
928 |
-
},
|
929 |
-
{
|
930 |
-
"epoch": 4.064,
|
931 |
-
"grad_norm": 10.824769020080566,
|
932 |
-
"learning_rate": 3.2995014245014244e-05,
|
933 |
-
"loss": 0.2298,
|
934 |
-
"step": 1270
|
935 |
-
},
|
936 |
-
{
|
937 |
-
"epoch": 4.096,
|
938 |
-
"grad_norm": 10.660799026489258,
|
939 |
-
"learning_rate": 3.281695156695157e-05,
|
940 |
-
"loss": 0.2172,
|
941 |
-
"step": 1280
|
942 |
-
},
|
943 |
-
{
|
944 |
-
"epoch": 4.128,
|
945 |
-
"grad_norm": 12.774826049804688,
|
946 |
-
"learning_rate": 3.263888888888889e-05,
|
947 |
-
"loss": 0.2204,
|
948 |
-
"step": 1290
|
949 |
-
},
|
950 |
-
{
|
951 |
-
"epoch": 4.16,
|
952 |
-
"grad_norm": 9.919928550720215,
|
953 |
-
"learning_rate": 3.246082621082621e-05,
|
954 |
-
"loss": 0.2355,
|
955 |
-
"step": 1300
|
956 |
-
},
|
957 |
-
{
|
958 |
-
"epoch": 4.192,
|
959 |
-
"grad_norm": 10.62849235534668,
|
960 |
-
"learning_rate": 3.228276353276353e-05,
|
961 |
-
"loss": 0.2046,
|
962 |
-
"step": 1310
|
963 |
-
},
|
964 |
-
{
|
965 |
-
"epoch": 4.224,
|
966 |
-
"grad_norm": 11.318085670471191,
|
967 |
-
"learning_rate": 3.2104700854700856e-05,
|
968 |
-
"loss": 0.2058,
|
969 |
-
"step": 1320
|
970 |
-
},
|
971 |
-
{
|
972 |
-
"epoch": 4.256,
|
973 |
-
"grad_norm": 23.96332359313965,
|
974 |
-
"learning_rate": 3.192663817663818e-05,
|
975 |
-
"loss": 0.2158,
|
976 |
-
"step": 1330
|
977 |
-
},
|
978 |
-
{
|
979 |
-
"epoch": 4.288,
|
980 |
-
"grad_norm": 12.120290756225586,
|
981 |
-
"learning_rate": 3.17485754985755e-05,
|
982 |
-
"loss": 0.1955,
|
983 |
-
"step": 1340
|
984 |
-
},
|
985 |
-
{
|
986 |
-
"epoch": 4.32,
|
987 |
-
"grad_norm": 22.462135314941406,
|
988 |
-
"learning_rate": 3.157051282051282e-05,
|
989 |
-
"loss": 0.2213,
|
990 |
-
"step": 1350
|
991 |
-
},
|
992 |
-
{
|
993 |
-
"epoch": 4.352,
|
994 |
-
"grad_norm": 10.982671737670898,
|
995 |
-
"learning_rate": 3.139245014245014e-05,
|
996 |
-
"loss": 0.204,
|
997 |
-
"step": 1360
|
998 |
-
},
|
999 |
-
{
|
1000 |
-
"epoch": 4.384,
|
1001 |
-
"grad_norm": 9.690351486206055,
|
1002 |
-
"learning_rate": 3.121438746438747e-05,
|
1003 |
-
"loss": 0.2072,
|
1004 |
-
"step": 1370
|
1005 |
-
},
|
1006 |
-
{
|
1007 |
-
"epoch": 4.416,
|
1008 |
-
"grad_norm": 21.02212142944336,
|
1009 |
-
"learning_rate": 3.103632478632479e-05,
|
1010 |
-
"loss": 0.2084,
|
1011 |
-
"step": 1380
|
1012 |
-
},
|
1013 |
-
{
|
1014 |
-
"epoch": 4.448,
|
1015 |
-
"grad_norm": 11.301972389221191,
|
1016 |
-
"learning_rate": 3.085826210826211e-05,
|
1017 |
-
"loss": 0.2046,
|
1018 |
-
"step": 1390
|
1019 |
-
},
|
1020 |
-
{
|
1021 |
-
"epoch": 4.48,
|
1022 |
-
"grad_norm": 13.271730422973633,
|
1023 |
-
"learning_rate": 3.068019943019943e-05,
|
1024 |
-
"loss": 0.2341,
|
1025 |
-
"step": 1400
|
1026 |
-
},
|
1027 |
-
{
|
1028 |
-
"epoch": 4.5120000000000005,
|
1029 |
-
"grad_norm": 12.261125564575195,
|
1030 |
-
"learning_rate": 3.0502136752136755e-05,
|
1031 |
-
"loss": 0.2151,
|
1032 |
-
"step": 1410
|
1033 |
-
},
|
1034 |
-
{
|
1035 |
-
"epoch": 4.5440000000000005,
|
1036 |
-
"grad_norm": 5.64617919921875,
|
1037 |
-
"learning_rate": 3.0324074074074077e-05,
|
1038 |
-
"loss": 0.2046,
|
1039 |
-
"step": 1420
|
1040 |
-
},
|
1041 |
-
{
|
1042 |
-
"epoch": 4.576,
|
1043 |
-
"grad_norm": 17.69122314453125,
|
1044 |
-
"learning_rate": 3.01460113960114e-05,
|
1045 |
-
"loss": 0.2058,
|
1046 |
-
"step": 1430
|
1047 |
-
},
|
1048 |
-
{
|
1049 |
-
"epoch": 4.608,
|
1050 |
-
"grad_norm": 10.206043243408203,
|
1051 |
-
"learning_rate": 2.9967948717948717e-05,
|
1052 |
-
"loss": 0.2235,
|
1053 |
-
"step": 1440
|
1054 |
-
},
|
1055 |
-
{
|
1056 |
-
"epoch": 4.64,
|
1057 |
-
"grad_norm": 13.48324203491211,
|
1058 |
-
"learning_rate": 2.978988603988604e-05,
|
1059 |
-
"loss": 0.1907,
|
1060 |
-
"step": 1450
|
1061 |
-
},
|
1062 |
-
{
|
1063 |
-
"epoch": 4.672,
|
1064 |
-
"grad_norm": 7.78882360458374,
|
1065 |
-
"learning_rate": 2.9611823361823364e-05,
|
1066 |
-
"loss": 0.2012,
|
1067 |
-
"step": 1460
|
1068 |
-
},
|
1069 |
-
{
|
1070 |
-
"epoch": 4.704,
|
1071 |
-
"grad_norm": 10.230093002319336,
|
1072 |
-
"learning_rate": 2.9433760683760685e-05,
|
1073 |
-
"loss": 0.1803,
|
1074 |
-
"step": 1470
|
1075 |
-
},
|
1076 |
-
{
|
1077 |
-
"epoch": 4.736,
|
1078 |
-
"grad_norm": 37.21369934082031,
|
1079 |
-
"learning_rate": 2.9255698005698007e-05,
|
1080 |
-
"loss": 0.1995,
|
1081 |
-
"step": 1480
|
1082 |
-
},
|
1083 |
-
{
|
1084 |
-
"epoch": 4.768,
|
1085 |
-
"grad_norm": 10.13237190246582,
|
1086 |
-
"learning_rate": 2.907763532763533e-05,
|
1087 |
-
"loss": 0.2096,
|
1088 |
-
"step": 1490
|
1089 |
-
},
|
1090 |
-
{
|
1091 |
-
"epoch": 4.8,
|
1092 |
-
"grad_norm": 6.978107929229736,
|
1093 |
-
"learning_rate": 2.8899572649572647e-05,
|
1094 |
-
"loss": 0.1905,
|
1095 |
-
"step": 1500
|
1096 |
-
},
|
1097 |
-
{
|
1098 |
-
"epoch": 4.832,
|
1099 |
-
"grad_norm": 17.814905166625977,
|
1100 |
-
"learning_rate": 2.8721509971509976e-05,
|
1101 |
-
"loss": 0.1818,
|
1102 |
-
"step": 1510
|
1103 |
-
},
|
1104 |
-
{
|
1105 |
-
"epoch": 4.864,
|
1106 |
-
"grad_norm": 10.511373519897461,
|
1107 |
-
"learning_rate": 2.8543447293447294e-05,
|
1108 |
-
"loss": 0.2339,
|
1109 |
-
"step": 1520
|
1110 |
-
},
|
1111 |
-
{
|
1112 |
-
"epoch": 4.896,
|
1113 |
-
"grad_norm": 16.811450958251953,
|
1114 |
-
"learning_rate": 2.8365384615384616e-05,
|
1115 |
-
"loss": 0.1969,
|
1116 |
-
"step": 1530
|
1117 |
-
},
|
1118 |
-
{
|
1119 |
-
"epoch": 4.928,
|
1120 |
-
"grad_norm": 8.571528434753418,
|
1121 |
-
"learning_rate": 2.8187321937321937e-05,
|
1122 |
-
"loss": 0.1951,
|
1123 |
-
"step": 1540
|
1124 |
-
},
|
1125 |
-
{
|
1126 |
-
"epoch": 4.96,
|
1127 |
-
"grad_norm": 15.971531867980957,
|
1128 |
-
"learning_rate": 2.8009259259259263e-05,
|
1129 |
-
"loss": 0.2012,
|
1130 |
-
"step": 1550
|
1131 |
-
},
|
1132 |
-
{
|
1133 |
-
"epoch": 4.992,
|
1134 |
-
"grad_norm": 22.862905502319336,
|
1135 |
-
"learning_rate": 2.7831196581196584e-05,
|
1136 |
-
"loss": 0.1925,
|
1137 |
-
"step": 1560
|
1138 |
-
},
|
1139 |
-
{
|
1140 |
-
"epoch": 4.9984,
|
1141 |
-
"eval_accuracy": 0.69075,
|
1142 |
-
"eval_loss": 1.0642794370651245,
|
1143 |
-
"eval_runtime": 164.1188,
|
1144 |
-
"eval_samples_per_second": 24.373,
|
1145 |
-
"eval_steps_per_second": 0.762,
|
1146 |
-
"step": 1562
|
1147 |
-
},
|
1148 |
-
{
|
1149 |
-
"epoch": 5.024,
|
1150 |
-
"grad_norm": 14.107147216796875,
|
1151 |
-
"learning_rate": 2.7653133903133903e-05,
|
1152 |
-
"loss": 0.2248,
|
1153 |
-
"step": 1570
|
1154 |
-
},
|
1155 |
-
{
|
1156 |
-
"epoch": 5.056,
|
1157 |
-
"grad_norm": 16.5964412689209,
|
1158 |
-
"learning_rate": 2.7475071225071224e-05,
|
1159 |
-
"loss": 0.1826,
|
1160 |
-
"step": 1580
|
1161 |
-
},
|
1162 |
-
{
|
1163 |
-
"epoch": 5.088,
|
1164 |
-
"grad_norm": 9.056426048278809,
|
1165 |
-
"learning_rate": 2.7297008547008546e-05,
|
1166 |
-
"loss": 0.1975,
|
1167 |
-
"step": 1590
|
1168 |
-
},
|
1169 |
-
{
|
1170 |
-
"epoch": 5.12,
|
1171 |
-
"grad_norm": 11.496504783630371,
|
1172 |
-
"learning_rate": 2.711894586894587e-05,
|
1173 |
-
"loss": 0.1818,
|
1174 |
-
"step": 1600
|
1175 |
-
},
|
1176 |
-
{
|
1177 |
-
"epoch": 5.152,
|
1178 |
-
"grad_norm": 14.043261528015137,
|
1179 |
-
"learning_rate": 2.6940883190883193e-05,
|
1180 |
-
"loss": 0.1744,
|
1181 |
-
"step": 1610
|
1182 |
-
},
|
1183 |
-
{
|
1184 |
-
"epoch": 5.184,
|
1185 |
-
"grad_norm": 8.685189247131348,
|
1186 |
-
"learning_rate": 2.6762820512820515e-05,
|
1187 |
-
"loss": 0.2042,
|
1188 |
-
"step": 1620
|
1189 |
-
},
|
1190 |
-
{
|
1191 |
-
"epoch": 5.216,
|
1192 |
-
"grad_norm": 7.994570732116699,
|
1193 |
-
"learning_rate": 2.6584757834757833e-05,
|
1194 |
-
"loss": 0.1703,
|
1195 |
-
"step": 1630
|
1196 |
-
},
|
1197 |
-
{
|
1198 |
-
"epoch": 5.248,
|
1199 |
-
"grad_norm": 14.493515968322754,
|
1200 |
-
"learning_rate": 2.640669515669516e-05,
|
1201 |
-
"loss": 0.1736,
|
1202 |
-
"step": 1640
|
1203 |
-
},
|
1204 |
-
{
|
1205 |
-
"epoch": 5.28,
|
1206 |
-
"grad_norm": 7.479248523712158,
|
1207 |
-
"learning_rate": 2.622863247863248e-05,
|
1208 |
-
"loss": 0.1868,
|
1209 |
-
"step": 1650
|
1210 |
-
},
|
1211 |
-
{
|
1212 |
-
"epoch": 5.312,
|
1213 |
-
"grad_norm": 12.55150318145752,
|
1214 |
-
"learning_rate": 2.60505698005698e-05,
|
1215 |
-
"loss": 0.2088,
|
1216 |
-
"step": 1660
|
1217 |
-
},
|
1218 |
-
{
|
1219 |
-
"epoch": 5.344,
|
1220 |
-
"grad_norm": 8.034307479858398,
|
1221 |
-
"learning_rate": 2.5872507122507123e-05,
|
1222 |
-
"loss": 0.18,
|
1223 |
-
"step": 1670
|
1224 |
-
},
|
1225 |
-
{
|
1226 |
-
"epoch": 5.376,
|
1227 |
-
"grad_norm": 7.336935997009277,
|
1228 |
-
"learning_rate": 2.5694444444444445e-05,
|
1229 |
-
"loss": 0.173,
|
1230 |
-
"step": 1680
|
1231 |
-
},
|
1232 |
-
{
|
1233 |
-
"epoch": 5.408,
|
1234 |
-
"grad_norm": 7.364587306976318,
|
1235 |
-
"learning_rate": 2.551638176638177e-05,
|
1236 |
-
"loss": 0.1739,
|
1237 |
-
"step": 1690
|
1238 |
-
},
|
1239 |
-
{
|
1240 |
-
"epoch": 5.44,
|
1241 |
-
"grad_norm": 12.251140594482422,
|
1242 |
-
"learning_rate": 2.533831908831909e-05,
|
1243 |
-
"loss": 0.1766,
|
1244 |
-
"step": 1700
|
1245 |
-
},
|
1246 |
-
{
|
1247 |
-
"epoch": 5.4719999999999995,
|
1248 |
-
"grad_norm": 8.115775108337402,
|
1249 |
-
"learning_rate": 2.516025641025641e-05,
|
1250 |
-
"loss": 0.1862,
|
1251 |
-
"step": 1710
|
1252 |
-
},
|
1253 |
-
{
|
1254 |
-
"epoch": 5.504,
|
1255 |
-
"grad_norm": 12.475907325744629,
|
1256 |
-
"learning_rate": 2.4982193732193735e-05,
|
1257 |
-
"loss": 0.1988,
|
1258 |
-
"step": 1720
|
1259 |
-
},
|
1260 |
-
{
|
1261 |
-
"epoch": 5.536,
|
1262 |
-
"grad_norm": 7.327911376953125,
|
1263 |
-
"learning_rate": 2.4804131054131057e-05,
|
1264 |
-
"loss": 0.1828,
|
1265 |
-
"step": 1730
|
1266 |
-
},
|
1267 |
-
{
|
1268 |
-
"epoch": 5.568,
|
1269 |
-
"grad_norm": 8.060572624206543,
|
1270 |
-
"learning_rate": 2.462606837606838e-05,
|
1271 |
-
"loss": 0.1692,
|
1272 |
-
"step": 1740
|
1273 |
-
},
|
1274 |
-
{
|
1275 |
-
"epoch": 5.6,
|
1276 |
-
"grad_norm": 28.06664276123047,
|
1277 |
-
"learning_rate": 2.44480056980057e-05,
|
1278 |
-
"loss": 0.1345,
|
1279 |
-
"step": 1750
|
1280 |
-
},
|
1281 |
-
{
|
1282 |
-
"epoch": 5.632,
|
1283 |
-
"grad_norm": 12.124792098999023,
|
1284 |
-
"learning_rate": 2.426994301994302e-05,
|
1285 |
-
"loss": 0.2026,
|
1286 |
-
"step": 1760
|
1287 |
-
},
|
1288 |
-
{
|
1289 |
-
"epoch": 5.664,
|
1290 |
-
"grad_norm": 12.650993347167969,
|
1291 |
-
"learning_rate": 2.4091880341880344e-05,
|
1292 |
-
"loss": 0.1861,
|
1293 |
-
"step": 1770
|
1294 |
-
},
|
1295 |
-
{
|
1296 |
-
"epoch": 5.696,
|
1297 |
-
"grad_norm": 15.481111526489258,
|
1298 |
-
"learning_rate": 2.3913817663817665e-05,
|
1299 |
-
"loss": 0.1963,
|
1300 |
-
"step": 1780
|
1301 |
-
},
|
1302 |
-
{
|
1303 |
-
"epoch": 5.728,
|
1304 |
-
"grad_norm": 12.246734619140625,
|
1305 |
-
"learning_rate": 2.3735754985754987e-05,
|
1306 |
-
"loss": 0.162,
|
1307 |
-
"step": 1790
|
1308 |
-
},
|
1309 |
-
{
|
1310 |
-
"epoch": 5.76,
|
1311 |
-
"grad_norm": 11.539756774902344,
|
1312 |
-
"learning_rate": 2.355769230769231e-05,
|
1313 |
-
"loss": 0.2141,
|
1314 |
-
"step": 1800
|
1315 |
-
},
|
1316 |
-
{
|
1317 |
-
"epoch": 5.792,
|
1318 |
-
"grad_norm": 7.406561374664307,
|
1319 |
-
"learning_rate": 2.337962962962963e-05,
|
1320 |
-
"loss": 0.1877,
|
1321 |
-
"step": 1810
|
1322 |
-
},
|
1323 |
-
{
|
1324 |
-
"epoch": 5.824,
|
1325 |
-
"grad_norm": 14.728952407836914,
|
1326 |
-
"learning_rate": 2.3201566951566952e-05,
|
1327 |
-
"loss": 0.1796,
|
1328 |
-
"step": 1820
|
1329 |
-
},
|
1330 |
-
{
|
1331 |
-
"epoch": 5.856,
|
1332 |
-
"grad_norm": 7.532665729522705,
|
1333 |
-
"learning_rate": 2.3023504273504274e-05,
|
1334 |
-
"loss": 0.157,
|
1335 |
-
"step": 1830
|
1336 |
-
},
|
1337 |
-
{
|
1338 |
-
"epoch": 5.888,
|
1339 |
-
"grad_norm": 16.279264450073242,
|
1340 |
-
"learning_rate": 2.2845441595441596e-05,
|
1341 |
-
"loss": 0.1836,
|
1342 |
-
"step": 1840
|
1343 |
-
},
|
1344 |
-
{
|
1345 |
-
"epoch": 5.92,
|
1346 |
-
"grad_norm": 7.029083728790283,
|
1347 |
-
"learning_rate": 2.2667378917378917e-05,
|
1348 |
-
"loss": 0.1788,
|
1349 |
-
"step": 1850
|
1350 |
-
},
|
1351 |
-
{
|
1352 |
-
"epoch": 5.952,
|
1353 |
-
"grad_norm": 7.0097174644470215,
|
1354 |
-
"learning_rate": 2.2489316239316242e-05,
|
1355 |
-
"loss": 0.1658,
|
1356 |
-
"step": 1860
|
1357 |
-
},
|
1358 |
-
{
|
1359 |
-
"epoch": 5.984,
|
1360 |
-
"grad_norm": 8.254197120666504,
|
1361 |
-
"learning_rate": 2.231125356125356e-05,
|
1362 |
-
"loss": 0.177,
|
1363 |
-
"step": 1870
|
1364 |
-
},
|
1365 |
-
{
|
1366 |
-
"epoch": 6.0,
|
1367 |
-
"eval_accuracy": 0.68425,
|
1368 |
-
"eval_loss": 1.2558131217956543,
|
1369 |
-
"eval_runtime": 148.6431,
|
1370 |
-
"eval_samples_per_second": 26.91,
|
1371 |
-
"eval_steps_per_second": 0.841,
|
1372 |
-
"step": 1875
|
1373 |
-
},
|
1374 |
-
{
|
1375 |
-
"epoch": 6.016,
|
1376 |
-
"grad_norm": 10.65648078918457,
|
1377 |
-
"learning_rate": 2.2133190883190886e-05,
|
1378 |
-
"loss": 0.197,
|
1379 |
-
"step": 1880
|
1380 |
-
},
|
1381 |
-
{
|
1382 |
-
"epoch": 6.048,
|
1383 |
-
"grad_norm": 12.431746482849121,
|
1384 |
-
"learning_rate": 2.1955128205128208e-05,
|
1385 |
-
"loss": 0.1876,
|
1386 |
-
"step": 1890
|
1387 |
-
},
|
1388 |
-
{
|
1389 |
-
"epoch": 6.08,
|
1390 |
-
"grad_norm": 20.30584716796875,
|
1391 |
-
"learning_rate": 2.177706552706553e-05,
|
1392 |
-
"loss": 0.1651,
|
1393 |
-
"step": 1900
|
1394 |
-
},
|
1395 |
-
{
|
1396 |
-
"epoch": 6.112,
|
1397 |
-
"grad_norm": 13.690505981445312,
|
1398 |
-
"learning_rate": 2.159900284900285e-05,
|
1399 |
-
"loss": 0.1646,
|
1400 |
-
"step": 1910
|
1401 |
-
},
|
1402 |
-
{
|
1403 |
-
"epoch": 6.144,
|
1404 |
-
"grad_norm": 23.424381256103516,
|
1405 |
-
"learning_rate": 2.142094017094017e-05,
|
1406 |
-
"loss": 0.1732,
|
1407 |
-
"step": 1920
|
1408 |
-
},
|
1409 |
-
{
|
1410 |
-
"epoch": 6.176,
|
1411 |
-
"grad_norm": 8.711691856384277,
|
1412 |
-
"learning_rate": 2.1242877492877494e-05,
|
1413 |
-
"loss": 0.1537,
|
1414 |
-
"step": 1930
|
1415 |
-
},
|
1416 |
-
{
|
1417 |
-
"epoch": 6.208,
|
1418 |
-
"grad_norm": 10.551396369934082,
|
1419 |
-
"learning_rate": 2.1064814814814816e-05,
|
1420 |
-
"loss": 0.1512,
|
1421 |
-
"step": 1940
|
1422 |
-
},
|
1423 |
-
{
|
1424 |
-
"epoch": 6.24,
|
1425 |
-
"grad_norm": 31.970632553100586,
|
1426 |
-
"learning_rate": 2.0886752136752138e-05,
|
1427 |
-
"loss": 0.1607,
|
1428 |
-
"step": 1950
|
1429 |
-
},
|
1430 |
-
{
|
1431 |
-
"epoch": 6.272,
|
1432 |
-
"grad_norm": 14.962709426879883,
|
1433 |
-
"learning_rate": 2.070868945868946e-05,
|
1434 |
-
"loss": 0.1476,
|
1435 |
-
"step": 1960
|
1436 |
-
},
|
1437 |
-
{
|
1438 |
-
"epoch": 6.304,
|
1439 |
-
"grad_norm": 11.659952163696289,
|
1440 |
-
"learning_rate": 2.053062678062678e-05,
|
1441 |
-
"loss": 0.1627,
|
1442 |
-
"step": 1970
|
1443 |
-
},
|
1444 |
-
{
|
1445 |
-
"epoch": 6.336,
|
1446 |
-
"grad_norm": 8.625643730163574,
|
1447 |
-
"learning_rate": 2.0352564102564103e-05,
|
1448 |
-
"loss": 0.1644,
|
1449 |
-
"step": 1980
|
1450 |
-
},
|
1451 |
-
{
|
1452 |
-
"epoch": 6.368,
|
1453 |
-
"grad_norm": 20.484344482421875,
|
1454 |
-
"learning_rate": 2.0174501424501425e-05,
|
1455 |
-
"loss": 0.1605,
|
1456 |
-
"step": 1990
|
1457 |
-
},
|
1458 |
-
{
|
1459 |
-
"epoch": 6.4,
|
1460 |
-
"grad_norm": 8.137100219726562,
|
1461 |
-
"learning_rate": 1.9996438746438746e-05,
|
1462 |
-
"loss": 0.1482,
|
1463 |
-
"step": 2000
|
1464 |
-
},
|
1465 |
-
{
|
1466 |
-
"epoch": 6.432,
|
1467 |
-
"grad_norm": 9.601692199707031,
|
1468 |
-
"learning_rate": 1.9818376068376068e-05,
|
1469 |
-
"loss": 0.1654,
|
1470 |
-
"step": 2010
|
1471 |
-
},
|
1472 |
-
{
|
1473 |
-
"epoch": 6.464,
|
1474 |
-
"grad_norm": 10.012831687927246,
|
1475 |
-
"learning_rate": 1.9640313390313393e-05,
|
1476 |
-
"loss": 0.1925,
|
1477 |
-
"step": 2020
|
1478 |
-
},
|
1479 |
-
{
|
1480 |
-
"epoch": 6.496,
|
1481 |
-
"grad_norm": 7.507844924926758,
|
1482 |
-
"learning_rate": 1.946225071225071e-05,
|
1483 |
-
"loss": 0.1735,
|
1484 |
-
"step": 2030
|
1485 |
-
},
|
1486 |
-
{
|
1487 |
-
"epoch": 6.5280000000000005,
|
1488 |
-
"grad_norm": 19.662033081054688,
|
1489 |
-
"learning_rate": 1.9284188034188037e-05,
|
1490 |
-
"loss": 0.1547,
|
1491 |
-
"step": 2040
|
1492 |
-
},
|
1493 |
-
{
|
1494 |
-
"epoch": 6.5600000000000005,
|
1495 |
-
"grad_norm": 7.522573947906494,
|
1496 |
-
"learning_rate": 1.910612535612536e-05,
|
1497 |
-
"loss": 0.1602,
|
1498 |
-
"step": 2050
|
1499 |
-
},
|
1500 |
-
{
|
1501 |
-
"epoch": 6.592,
|
1502 |
-
"grad_norm": 9.965826034545898,
|
1503 |
-
"learning_rate": 1.892806267806268e-05,
|
1504 |
-
"loss": 0.1488,
|
1505 |
-
"step": 2060
|
1506 |
-
},
|
1507 |
-
{
|
1508 |
-
"epoch": 6.624,
|
1509 |
-
"grad_norm": 6.560938358306885,
|
1510 |
-
"learning_rate": 1.8750000000000002e-05,
|
1511 |
-
"loss": 0.1439,
|
1512 |
-
"step": 2070
|
1513 |
-
},
|
1514 |
-
{
|
1515 |
-
"epoch": 6.656,
|
1516 |
-
"grad_norm": 16.683101654052734,
|
1517 |
-
"learning_rate": 1.857193732193732e-05,
|
1518 |
-
"loss": 0.1384,
|
1519 |
-
"step": 2080
|
1520 |
-
},
|
1521 |
-
{
|
1522 |
-
"epoch": 6.688,
|
1523 |
-
"grad_norm": 7.641848564147949,
|
1524 |
-
"learning_rate": 1.8393874643874645e-05,
|
1525 |
-
"loss": 0.1429,
|
1526 |
-
"step": 2090
|
1527 |
-
},
|
1528 |
-
{
|
1529 |
-
"epoch": 6.72,
|
1530 |
-
"grad_norm": 11.998226165771484,
|
1531 |
-
"learning_rate": 1.8215811965811967e-05,
|
1532 |
-
"loss": 0.1485,
|
1533 |
-
"step": 2100
|
1534 |
-
},
|
1535 |
-
{
|
1536 |
-
"epoch": 6.752,
|
1537 |
-
"grad_norm": 7.238443374633789,
|
1538 |
-
"learning_rate": 1.803774928774929e-05,
|
1539 |
-
"loss": 0.1544,
|
1540 |
-
"step": 2110
|
1541 |
-
},
|
1542 |
-
{
|
1543 |
-
"epoch": 6.784,
|
1544 |
-
"grad_norm": 7.6015777587890625,
|
1545 |
-
"learning_rate": 1.785968660968661e-05,
|
1546 |
-
"loss": 0.1864,
|
1547 |
-
"step": 2120
|
1548 |
-
},
|
1549 |
-
{
|
1550 |
-
"epoch": 6.816,
|
1551 |
-
"grad_norm": 12.038315773010254,
|
1552 |
-
"learning_rate": 1.7681623931623932e-05,
|
1553 |
-
"loss": 0.1479,
|
1554 |
-
"step": 2130
|
1555 |
-
},
|
1556 |
-
{
|
1557 |
-
"epoch": 6.848,
|
1558 |
-
"grad_norm": 25.752042770385742,
|
1559 |
-
"learning_rate": 1.7503561253561254e-05,
|
1560 |
-
"loss": 0.1479,
|
1561 |
-
"step": 2140
|
1562 |
-
},
|
1563 |
-
{
|
1564 |
-
"epoch": 6.88,
|
1565 |
-
"grad_norm": 12.058791160583496,
|
1566 |
-
"learning_rate": 1.732549857549858e-05,
|
1567 |
-
"loss": 0.1342,
|
1568 |
-
"step": 2150
|
1569 |
-
},
|
1570 |
-
{
|
1571 |
-
"epoch": 6.912,
|
1572 |
-
"grad_norm": 18.634387969970703,
|
1573 |
-
"learning_rate": 1.7147435897435897e-05,
|
1574 |
-
"loss": 0.165,
|
1575 |
-
"step": 2160
|
1576 |
-
},
|
1577 |
-
{
|
1578 |
-
"epoch": 6.944,
|
1579 |
-
"grad_norm": 8.238666534423828,
|
1580 |
-
"learning_rate": 1.696937321937322e-05,
|
1581 |
-
"loss": 0.1452,
|
1582 |
-
"step": 2170
|
1583 |
-
},
|
1584 |
-
{
|
1585 |
-
"epoch": 6.976,
|
1586 |
-
"grad_norm": 8.537055015563965,
|
1587 |
-
"learning_rate": 1.6791310541310544e-05,
|
1588 |
-
"loss": 0.1563,
|
1589 |
-
"step": 2180
|
1590 |
-
},
|
1591 |
-
{
|
1592 |
-
"epoch": 6.9984,
|
1593 |
-
"eval_accuracy": 0.7445,
|
1594 |
-
"eval_loss": 0.9205208420753479,
|
1595 |
-
"eval_runtime": 156.8331,
|
1596 |
-
"eval_samples_per_second": 25.505,
|
1597 |
-
"eval_steps_per_second": 0.797,
|
1598 |
-
"step": 2187
|
1599 |
-
},
|
1600 |
-
{
|
1601 |
-
"epoch": 7.008,
|
1602 |
-
"grad_norm": 21.315475463867188,
|
1603 |
-
"learning_rate": 1.6613247863247862e-05,
|
1604 |
-
"loss": 0.1621,
|
1605 |
-
"step": 2190
|
1606 |
-
},
|
1607 |
-
{
|
1608 |
-
"epoch": 7.04,
|
1609 |
-
"grad_norm": 4.930037975311279,
|
1610 |
-
"learning_rate": 1.6435185185185187e-05,
|
1611 |
-
"loss": 0.143,
|
1612 |
-
"step": 2200
|
1613 |
-
},
|
1614 |
-
{
|
1615 |
-
"epoch": 7.072,
|
1616 |
-
"grad_norm": 20.74496078491211,
|
1617 |
-
"learning_rate": 1.625712250712251e-05,
|
1618 |
-
"loss": 0.1334,
|
1619 |
-
"step": 2210
|
1620 |
-
},
|
1621 |
-
{
|
1622 |
-
"epoch": 7.104,
|
1623 |
-
"grad_norm": 32.969154357910156,
|
1624 |
-
"learning_rate": 1.607905982905983e-05,
|
1625 |
-
"loss": 0.1502,
|
1626 |
-
"step": 2220
|
1627 |
-
},
|
1628 |
-
{
|
1629 |
-
"epoch": 7.136,
|
1630 |
-
"grad_norm": 10.03094482421875,
|
1631 |
-
"learning_rate": 1.5900997150997153e-05,
|
1632 |
-
"loss": 0.1611,
|
1633 |
-
"step": 2230
|
1634 |
-
},
|
1635 |
-
{
|
1636 |
-
"epoch": 7.168,
|
1637 |
-
"grad_norm": 10.585143089294434,
|
1638 |
-
"learning_rate": 1.572293447293447e-05,
|
1639 |
-
"loss": 0.1583,
|
1640 |
-
"step": 2240
|
1641 |
-
},
|
1642 |
-
{
|
1643 |
-
"epoch": 7.2,
|
1644 |
-
"grad_norm": 11.96883487701416,
|
1645 |
-
"learning_rate": 1.5544871794871796e-05,
|
1646 |
-
"loss": 0.1433,
|
1647 |
-
"step": 2250
|
1648 |
-
},
|
1649 |
-
{
|
1650 |
-
"epoch": 7.232,
|
1651 |
-
"grad_norm": 17.068143844604492,
|
1652 |
-
"learning_rate": 1.5366809116809118e-05,
|
1653 |
-
"loss": 0.1641,
|
1654 |
-
"step": 2260
|
1655 |
-
},
|
1656 |
-
{
|
1657 |
-
"epoch": 7.264,
|
1658 |
-
"grad_norm": 9.417765617370605,
|
1659 |
-
"learning_rate": 1.518874643874644e-05,
|
1660 |
-
"loss": 0.1266,
|
1661 |
-
"step": 2270
|
1662 |
-
},
|
1663 |
-
{
|
1664 |
-
"epoch": 7.296,
|
1665 |
-
"grad_norm": 12.315200805664062,
|
1666 |
-
"learning_rate": 1.5010683760683761e-05,
|
1667 |
-
"loss": 0.1337,
|
1668 |
-
"step": 2280
|
1669 |
-
},
|
1670 |
-
{
|
1671 |
-
"epoch": 7.328,
|
1672 |
-
"grad_norm": 20.546875,
|
1673 |
-
"learning_rate": 1.4832621082621084e-05,
|
1674 |
-
"loss": 0.1312,
|
1675 |
-
"step": 2290
|
1676 |
-
},
|
1677 |
-
{
|
1678 |
-
"epoch": 7.36,
|
1679 |
-
"grad_norm": 7.2887468338012695,
|
1680 |
-
"learning_rate": 1.4654558404558405e-05,
|
1681 |
-
"loss": 0.1534,
|
1682 |
-
"step": 2300
|
1683 |
-
},
|
1684 |
-
{
|
1685 |
-
"epoch": 7.392,
|
1686 |
-
"grad_norm": 8.704584121704102,
|
1687 |
-
"learning_rate": 1.4476495726495728e-05,
|
1688 |
-
"loss": 0.1317,
|
1689 |
-
"step": 2310
|
1690 |
-
},
|
1691 |
-
{
|
1692 |
-
"epoch": 7.424,
|
1693 |
-
"grad_norm": 9.408025741577148,
|
1694 |
-
"learning_rate": 1.429843304843305e-05,
|
1695 |
-
"loss": 0.1336,
|
1696 |
-
"step": 2320
|
1697 |
-
},
|
1698 |
-
{
|
1699 |
-
"epoch": 7.456,
|
1700 |
-
"grad_norm": 16.08376121520996,
|
1701 |
-
"learning_rate": 1.412037037037037e-05,
|
1702 |
-
"loss": 0.1361,
|
1703 |
-
"step": 2330
|
1704 |
-
},
|
1705 |
-
{
|
1706 |
-
"epoch": 7.4879999999999995,
|
1707 |
-
"grad_norm": 13.386305809020996,
|
1708 |
-
"learning_rate": 1.3942307692307693e-05,
|
1709 |
-
"loss": 0.1254,
|
1710 |
-
"step": 2340
|
1711 |
-
},
|
1712 |
-
{
|
1713 |
-
"epoch": 7.52,
|
1714 |
-
"grad_norm": 8.729687690734863,
|
1715 |
-
"learning_rate": 1.3764245014245015e-05,
|
1716 |
-
"loss": 0.1477,
|
1717 |
-
"step": 2350
|
1718 |
-
},
|
1719 |
-
{
|
1720 |
-
"epoch": 7.552,
|
1721 |
-
"grad_norm": 7.066646099090576,
|
1722 |
-
"learning_rate": 1.3586182336182338e-05,
|
1723 |
-
"loss": 0.1266,
|
1724 |
-
"step": 2360
|
1725 |
-
},
|
1726 |
-
{
|
1727 |
-
"epoch": 7.584,
|
1728 |
-
"grad_norm": 11.6298828125,
|
1729 |
-
"learning_rate": 1.3408119658119658e-05,
|
1730 |
-
"loss": 0.1303,
|
1731 |
-
"step": 2370
|
1732 |
-
},
|
1733 |
-
{
|
1734 |
-
"epoch": 7.616,
|
1735 |
-
"grad_norm": 5.477276802062988,
|
1736 |
-
"learning_rate": 1.3230056980056982e-05,
|
1737 |
-
"loss": 0.126,
|
1738 |
-
"step": 2380
|
1739 |
-
},
|
1740 |
-
{
|
1741 |
-
"epoch": 7.648,
|
1742 |
-
"grad_norm": 9.086950302124023,
|
1743 |
-
"learning_rate": 1.3051994301994303e-05,
|
1744 |
-
"loss": 0.1383,
|
1745 |
-
"step": 2390
|
1746 |
-
},
|
1747 |
-
{
|
1748 |
-
"epoch": 7.68,
|
1749 |
-
"grad_norm": 7.193227291107178,
|
1750 |
-
"learning_rate": 1.2873931623931623e-05,
|
1751 |
-
"loss": 0.1371,
|
1752 |
-
"step": 2400
|
1753 |
-
},
|
1754 |
-
{
|
1755 |
-
"epoch": 7.712,
|
1756 |
-
"grad_norm": 11.82983684539795,
|
1757 |
-
"learning_rate": 1.2695868945868947e-05,
|
1758 |
-
"loss": 0.1406,
|
1759 |
-
"step": 2410
|
1760 |
-
},
|
1761 |
-
{
|
1762 |
-
"epoch": 7.744,
|
1763 |
-
"grad_norm": 9.807685852050781,
|
1764 |
-
"learning_rate": 1.2517806267806267e-05,
|
1765 |
-
"loss": 0.1349,
|
1766 |
-
"step": 2420
|
1767 |
-
},
|
1768 |
-
{
|
1769 |
-
"epoch": 7.776,
|
1770 |
-
"grad_norm": 13.509228706359863,
|
1771 |
-
"learning_rate": 1.233974358974359e-05,
|
1772 |
-
"loss": 0.1324,
|
1773 |
-
"step": 2430
|
1774 |
-
},
|
1775 |
-
{
|
1776 |
-
"epoch": 7.808,
|
1777 |
-
"grad_norm": 20.25900650024414,
|
1778 |
-
"learning_rate": 1.2161680911680912e-05,
|
1779 |
-
"loss": 0.1337,
|
1780 |
-
"step": 2440
|
1781 |
-
},
|
1782 |
-
{
|
1783 |
-
"epoch": 7.84,
|
1784 |
-
"grad_norm": 8.688305854797363,
|
1785 |
-
"learning_rate": 1.1983618233618234e-05,
|
1786 |
-
"loss": 0.1206,
|
1787 |
-
"step": 2450
|
1788 |
-
},
|
1789 |
-
{
|
1790 |
-
"epoch": 7.872,
|
1791 |
-
"grad_norm": 16.35320472717285,
|
1792 |
-
"learning_rate": 1.1805555555555555e-05,
|
1793 |
-
"loss": 0.133,
|
1794 |
-
"step": 2460
|
1795 |
-
},
|
1796 |
-
{
|
1797 |
-
"epoch": 7.904,
|
1798 |
-
"grad_norm": 35.276092529296875,
|
1799 |
-
"learning_rate": 1.1627492877492879e-05,
|
1800 |
-
"loss": 0.1425,
|
1801 |
-
"step": 2470
|
1802 |
-
},
|
1803 |
-
{
|
1804 |
-
"epoch": 7.936,
|
1805 |
-
"grad_norm": 16.226844787597656,
|
1806 |
-
"learning_rate": 1.14494301994302e-05,
|
1807 |
-
"loss": 0.1421,
|
1808 |
-
"step": 2480
|
1809 |
-
},
|
1810 |
-
{
|
1811 |
-
"epoch": 7.968,
|
1812 |
-
"grad_norm": 4.65103006362915,
|
1813 |
-
"learning_rate": 1.1271367521367522e-05,
|
1814 |
-
"loss": 0.1225,
|
1815 |
-
"step": 2490
|
1816 |
-
},
|
1817 |
-
{
|
1818 |
-
"epoch": 8.0,
|
1819 |
-
"grad_norm": 10.949529647827148,
|
1820 |
-
"learning_rate": 1.1093304843304844e-05,
|
1821 |
-
"loss": 0.1417,
|
1822 |
-
"step": 2500
|
1823 |
-
},
|
1824 |
-
{
|
1825 |
-
"epoch": 8.0,
|
1826 |
-
"eval_accuracy": 0.80625,
|
1827 |
-
"eval_loss": 0.662419319152832,
|
1828 |
-
"eval_runtime": 156.8602,
|
1829 |
-
"eval_samples_per_second": 25.5,
|
1830 |
-
"eval_steps_per_second": 0.797,
|
1831 |
-
"step": 2500
|
1832 |
-
},
|
1833 |
-
{
|
1834 |
-
"epoch": 8.032,
|
1835 |
-
"grad_norm": 14.452116966247559,
|
1836 |
-
"learning_rate": 1.0915242165242166e-05,
|
1837 |
-
"loss": 0.1422,
|
1838 |
-
"step": 2510
|
1839 |
-
},
|
1840 |
-
{
|
1841 |
-
"epoch": 8.064,
|
1842 |
-
"grad_norm": 26.71641731262207,
|
1843 |
-
"learning_rate": 1.0737179487179487e-05,
|
1844 |
-
"loss": 0.1337,
|
1845 |
-
"step": 2520
|
1846 |
-
},
|
1847 |
-
{
|
1848 |
-
"epoch": 8.096,
|
1849 |
-
"grad_norm": 8.862665176391602,
|
1850 |
-
"learning_rate": 1.0559116809116809e-05,
|
1851 |
-
"loss": 0.1349,
|
1852 |
-
"step": 2530
|
1853 |
-
},
|
1854 |
-
{
|
1855 |
-
"epoch": 8.128,
|
1856 |
-
"grad_norm": 12.3915433883667,
|
1857 |
-
"learning_rate": 1.038105413105413e-05,
|
1858 |
-
"loss": 0.1479,
|
1859 |
-
"step": 2540
|
1860 |
-
},
|
1861 |
-
{
|
1862 |
-
"epoch": 8.16,
|
1863 |
-
"grad_norm": 11.4617280960083,
|
1864 |
-
"learning_rate": 1.0202991452991454e-05,
|
1865 |
-
"loss": 0.1295,
|
1866 |
-
"step": 2550
|
1867 |
-
},
|
1868 |
-
{
|
1869 |
-
"epoch": 8.192,
|
1870 |
-
"grad_norm": 19.975990295410156,
|
1871 |
-
"learning_rate": 1.0024928774928776e-05,
|
1872 |
-
"loss": 0.1382,
|
1873 |
-
"step": 2560
|
1874 |
-
},
|
1875 |
-
{
|
1876 |
-
"epoch": 8.224,
|
1877 |
-
"grad_norm": 9.245658874511719,
|
1878 |
-
"learning_rate": 9.846866096866097e-06,
|
1879 |
-
"loss": 0.1308,
|
1880 |
-
"step": 2570
|
1881 |
-
},
|
1882 |
-
{
|
1883 |
-
"epoch": 8.256,
|
1884 |
-
"grad_norm": 5.326711177825928,
|
1885 |
-
"learning_rate": 9.66880341880342e-06,
|
1886 |
-
"loss": 0.1023,
|
1887 |
-
"step": 2580
|
1888 |
-
},
|
1889 |
-
{
|
1890 |
-
"epoch": 8.288,
|
1891 |
-
"grad_norm": 11.740446090698242,
|
1892 |
-
"learning_rate": 9.490740740740741e-06,
|
1893 |
-
"loss": 0.1262,
|
1894 |
-
"step": 2590
|
1895 |
-
},
|
1896 |
-
{
|
1897 |
-
"epoch": 8.32,
|
1898 |
-
"grad_norm": 10.618571281433105,
|
1899 |
-
"learning_rate": 9.312678062678064e-06,
|
1900 |
-
"loss": 0.1306,
|
1901 |
-
"step": 2600
|
1902 |
-
},
|
1903 |
-
{
|
1904 |
-
"epoch": 8.352,
|
1905 |
-
"grad_norm": 21.31345558166504,
|
1906 |
-
"learning_rate": 9.134615384615384e-06,
|
1907 |
-
"loss": 0.136,
|
1908 |
-
"step": 2610
|
1909 |
-
},
|
1910 |
-
{
|
1911 |
-
"epoch": 8.384,
|
1912 |
-
"grad_norm": 6.4540863037109375,
|
1913 |
-
"learning_rate": 8.956552706552706e-06,
|
1914 |
-
"loss": 0.1063,
|
1915 |
-
"step": 2620
|
1916 |
-
},
|
1917 |
-
{
|
1918 |
-
"epoch": 8.416,
|
1919 |
-
"grad_norm": 12.291879653930664,
|
1920 |
-
"learning_rate": 8.77849002849003e-06,
|
1921 |
-
"loss": 0.1222,
|
1922 |
-
"step": 2630
|
1923 |
-
},
|
1924 |
-
{
|
1925 |
-
"epoch": 8.448,
|
1926 |
-
"grad_norm": 7.860499858856201,
|
1927 |
-
"learning_rate": 8.600427350427351e-06,
|
1928 |
-
"loss": 0.1152,
|
1929 |
-
"step": 2640
|
1930 |
-
},
|
1931 |
-
{
|
1932 |
-
"epoch": 8.48,
|
1933 |
-
"grad_norm": 7.345857620239258,
|
1934 |
-
"learning_rate": 8.422364672364673e-06,
|
1935 |
-
"loss": 0.1194,
|
1936 |
-
"step": 2650
|
1937 |
-
},
|
1938 |
-
{
|
1939 |
-
"epoch": 8.512,
|
1940 |
-
"grad_norm": 25.202924728393555,
|
1941 |
-
"learning_rate": 8.244301994301995e-06,
|
1942 |
-
"loss": 0.1214,
|
1943 |
-
"step": 2660
|
1944 |
-
},
|
1945 |
-
{
|
1946 |
-
"epoch": 8.544,
|
1947 |
-
"grad_norm": 4.337811470031738,
|
1948 |
-
"learning_rate": 8.066239316239316e-06,
|
1949 |
-
"loss": 0.1316,
|
1950 |
-
"step": 2670
|
1951 |
-
},
|
1952 |
-
{
|
1953 |
-
"epoch": 8.576,
|
1954 |
-
"grad_norm": 9.332151412963867,
|
1955 |
-
"learning_rate": 7.88817663817664e-06,
|
1956 |
-
"loss": 0.1208,
|
1957 |
-
"step": 2680
|
1958 |
-
},
|
1959 |
-
{
|
1960 |
-
"epoch": 8.608,
|
1961 |
-
"grad_norm": 15.260229110717773,
|
1962 |
-
"learning_rate": 7.71011396011396e-06,
|
1963 |
-
"loss": 0.1223,
|
1964 |
-
"step": 2690
|
1965 |
-
},
|
1966 |
-
{
|
1967 |
-
"epoch": 8.64,
|
1968 |
-
"grad_norm": 4.976259708404541,
|
1969 |
-
"learning_rate": 7.532051282051282e-06,
|
1970 |
-
"loss": 0.111,
|
1971 |
-
"step": 2700
|
1972 |
-
},
|
1973 |
-
{
|
1974 |
-
"epoch": 8.672,
|
1975 |
-
"grad_norm": 18.121919631958008,
|
1976 |
-
"learning_rate": 7.353988603988604e-06,
|
1977 |
-
"loss": 0.1184,
|
1978 |
-
"step": 2710
|
1979 |
-
},
|
1980 |
-
{
|
1981 |
-
"epoch": 8.704,
|
1982 |
-
"grad_norm": 15.61289119720459,
|
1983 |
-
"learning_rate": 7.1759259259259266e-06,
|
1984 |
-
"loss": 0.125,
|
1985 |
-
"step": 2720
|
1986 |
-
},
|
1987 |
-
{
|
1988 |
-
"epoch": 8.736,
|
1989 |
-
"grad_norm": 14.504264831542969,
|
1990 |
-
"learning_rate": 6.997863247863248e-06,
|
1991 |
-
"loss": 0.1165,
|
1992 |
-
"step": 2730
|
1993 |
-
},
|
1994 |
-
{
|
1995 |
-
"epoch": 8.768,
|
1996 |
-
"grad_norm": 10.373973846435547,
|
1997 |
-
"learning_rate": 6.819800569800571e-06,
|
1998 |
-
"loss": 0.1117,
|
1999 |
-
"step": 2740
|
2000 |
-
},
|
2001 |
-
{
|
2002 |
-
"epoch": 8.8,
|
2003 |
-
"grad_norm": 32.67918014526367,
|
2004 |
-
"learning_rate": 6.6417378917378925e-06,
|
2005 |
-
"loss": 0.1345,
|
2006 |
-
"step": 2750
|
2007 |
-
},
|
2008 |
-
{
|
2009 |
-
"epoch": 8.832,
|
2010 |
-
"grad_norm": 11.788604736328125,
|
2011 |
-
"learning_rate": 6.463675213675214e-06,
|
2012 |
-
"loss": 0.1194,
|
2013 |
-
"step": 2760
|
2014 |
-
},
|
2015 |
-
{
|
2016 |
-
"epoch": 8.864,
|
2017 |
-
"grad_norm": 10.273115158081055,
|
2018 |
-
"learning_rate": 6.285612535612535e-06,
|
2019 |
-
"loss": 0.1089,
|
2020 |
-
"step": 2770
|
2021 |
-
},
|
2022 |
-
{
|
2023 |
-
"epoch": 8.896,
|
2024 |
-
"grad_norm": 18.808115005493164,
|
2025 |
-
"learning_rate": 6.1075498575498585e-06,
|
2026 |
-
"loss": 0.1106,
|
2027 |
-
"step": 2780
|
2028 |
-
},
|
2029 |
-
{
|
2030 |
-
"epoch": 8.928,
|
2031 |
-
"grad_norm": 8.078729629516602,
|
2032 |
-
"learning_rate": 5.929487179487179e-06,
|
2033 |
-
"loss": 0.1407,
|
2034 |
-
"step": 2790
|
2035 |
-
},
|
2036 |
-
{
|
2037 |
-
"epoch": 8.96,
|
2038 |
-
"grad_norm": 6.521687984466553,
|
2039 |
-
"learning_rate": 5.751424501424502e-06,
|
2040 |
-
"loss": 0.1195,
|
2041 |
-
"step": 2800
|
2042 |
-
},
|
2043 |
-
{
|
2044 |
-
"epoch": 8.992,
|
2045 |
-
"grad_norm": 7.088119029998779,
|
2046 |
-
"learning_rate": 5.573361823361824e-06,
|
2047 |
-
"loss": 0.1284,
|
2048 |
-
"step": 2810
|
2049 |
-
},
|
2050 |
-
{
|
2051 |
-
"epoch": 8.9984,
|
2052 |
-
"eval_accuracy": 0.739,
|
2053 |
-
"eval_loss": 1.1648138761520386,
|
2054 |
-
"eval_runtime": 157.4642,
|
2055 |
-
"eval_samples_per_second": 25.403,
|
2056 |
-
"eval_steps_per_second": 0.794,
|
2057 |
-
"step": 2812
|
2058 |
-
},
|
2059 |
-
{
|
2060 |
-
"epoch": 9.024,
|
2061 |
-
"grad_norm": 19.490074157714844,
|
2062 |
-
"learning_rate": 5.395299145299146e-06,
|
2063 |
-
"loss": 0.1084,
|
2064 |
-
"step": 2820
|
2065 |
-
},
|
2066 |
-
{
|
2067 |
-
"epoch": 9.056,
|
2068 |
-
"grad_norm": 6.817428112030029,
|
2069 |
-
"learning_rate": 5.217236467236467e-06,
|
2070 |
-
"loss": 0.1194,
|
2071 |
-
"step": 2830
|
2072 |
-
},
|
2073 |
-
{
|
2074 |
-
"epoch": 9.088,
|
2075 |
-
"grad_norm": 8.714266777038574,
|
2076 |
-
"learning_rate": 5.03917378917379e-06,
|
2077 |
-
"loss": 0.1126,
|
2078 |
-
"step": 2840
|
2079 |
-
},
|
2080 |
-
{
|
2081 |
-
"epoch": 9.12,
|
2082 |
-
"grad_norm": 15.373977661132812,
|
2083 |
-
"learning_rate": 4.861111111111111e-06,
|
2084 |
-
"loss": 0.1147,
|
2085 |
-
"step": 2850
|
2086 |
-
},
|
2087 |
-
{
|
2088 |
-
"epoch": 9.152,
|
2089 |
-
"grad_norm": 8.224656105041504,
|
2090 |
-
"learning_rate": 4.683048433048434e-06,
|
2091 |
-
"loss": 0.1023,
|
2092 |
-
"step": 2860
|
2093 |
-
},
|
2094 |
-
{
|
2095 |
-
"epoch": 9.184,
|
2096 |
-
"grad_norm": 14.560908317565918,
|
2097 |
-
"learning_rate": 4.504985754985755e-06,
|
2098 |
-
"loss": 0.1156,
|
2099 |
-
"step": 2870
|
2100 |
-
},
|
2101 |
-
{
|
2102 |
-
"epoch": 9.216,
|
2103 |
-
"grad_norm": 9.510125160217285,
|
2104 |
-
"learning_rate": 4.326923076923077e-06,
|
2105 |
-
"loss": 0.1123,
|
2106 |
-
"step": 2880
|
2107 |
-
},
|
2108 |
-
{
|
2109 |
-
"epoch": 9.248,
|
2110 |
-
"grad_norm": 11.263538360595703,
|
2111 |
-
"learning_rate": 4.148860398860399e-06,
|
2112 |
-
"loss": 0.1126,
|
2113 |
-
"step": 2890
|
2114 |
-
},
|
2115 |
-
{
|
2116 |
-
"epoch": 9.28,
|
2117 |
-
"grad_norm": 7.910732269287109,
|
2118 |
-
"learning_rate": 3.970797720797722e-06,
|
2119 |
-
"loss": 0.1075,
|
2120 |
-
"step": 2900
|
2121 |
-
},
|
2122 |
-
{
|
2123 |
-
"epoch": 9.312,
|
2124 |
-
"grad_norm": 6.68363618850708,
|
2125 |
-
"learning_rate": 3.7927350427350425e-06,
|
2126 |
-
"loss": 0.1001,
|
2127 |
-
"step": 2910
|
2128 |
-
},
|
2129 |
-
{
|
2130 |
-
"epoch": 9.344,
|
2131 |
-
"grad_norm": 7.76242733001709,
|
2132 |
-
"learning_rate": 3.6146723646723646e-06,
|
2133 |
-
"loss": 0.0995,
|
2134 |
-
"step": 2920
|
2135 |
-
},
|
2136 |
-
{
|
2137 |
-
"epoch": 9.376,
|
2138 |
-
"grad_norm": 12.8687744140625,
|
2139 |
-
"learning_rate": 3.4366096866096867e-06,
|
2140 |
-
"loss": 0.1449,
|
2141 |
-
"step": 2930
|
2142 |
-
},
|
2143 |
-
{
|
2144 |
-
"epoch": 9.408,
|
2145 |
-
"grad_norm": 12.708325386047363,
|
2146 |
-
"learning_rate": 3.258547008547009e-06,
|
2147 |
-
"loss": 0.1299,
|
2148 |
-
"step": 2940
|
2149 |
-
},
|
2150 |
-
{
|
2151 |
-
"epoch": 9.44,
|
2152 |
-
"grad_norm": 14.307499885559082,
|
2153 |
-
"learning_rate": 3.0804843304843306e-06,
|
2154 |
-
"loss": 0.1259,
|
2155 |
-
"step": 2950
|
2156 |
-
},
|
2157 |
-
{
|
2158 |
-
"epoch": 9.472,
|
2159 |
-
"grad_norm": 16.907052993774414,
|
2160 |
-
"learning_rate": 2.9024216524216523e-06,
|
2161 |
-
"loss": 0.1014,
|
2162 |
-
"step": 2960
|
2163 |
-
},
|
2164 |
-
{
|
2165 |
-
"epoch": 9.504,
|
2166 |
-
"grad_norm": 20.69292640686035,
|
2167 |
-
"learning_rate": 2.7243589743589744e-06,
|
2168 |
-
"loss": 0.1059,
|
2169 |
-
"step": 2970
|
2170 |
-
},
|
2171 |
-
{
|
2172 |
-
"epoch": 9.536,
|
2173 |
-
"grad_norm": 7.013891696929932,
|
2174 |
-
"learning_rate": 2.546296296296296e-06,
|
2175 |
-
"loss": 0.1025,
|
2176 |
-
"step": 2980
|
2177 |
-
},
|
2178 |
-
{
|
2179 |
-
"epoch": 9.568,
|
2180 |
-
"grad_norm": 10.161940574645996,
|
2181 |
-
"learning_rate": 2.3682336182336183e-06,
|
2182 |
-
"loss": 0.1132,
|
2183 |
-
"step": 2990
|
2184 |
-
},
|
2185 |
-
{
|
2186 |
-
"epoch": 9.6,
|
2187 |
-
"grad_norm": 9.757913589477539,
|
2188 |
-
"learning_rate": 2.19017094017094e-06,
|
2189 |
-
"loss": 0.1053,
|
2190 |
-
"step": 3000
|
2191 |
-
},
|
2192 |
-
{
|
2193 |
-
"epoch": 9.632,
|
2194 |
-
"grad_norm": 13.03531265258789,
|
2195 |
-
"learning_rate": 2.012108262108262e-06,
|
2196 |
-
"loss": 0.0961,
|
2197 |
-
"step": 3010
|
2198 |
-
},
|
2199 |
-
{
|
2200 |
-
"epoch": 9.664,
|
2201 |
-
"grad_norm": 8.091846466064453,
|
2202 |
-
"learning_rate": 1.834045584045584e-06,
|
2203 |
-
"loss": 0.0904,
|
2204 |
-
"step": 3020
|
2205 |
-
},
|
2206 |
-
{
|
2207 |
-
"epoch": 9.696,
|
2208 |
-
"grad_norm": 6.219699382781982,
|
2209 |
-
"learning_rate": 1.6559829059829062e-06,
|
2210 |
-
"loss": 0.0947,
|
2211 |
-
"step": 3030
|
2212 |
-
},
|
2213 |
-
{
|
2214 |
-
"epoch": 9.728,
|
2215 |
-
"grad_norm": 8.676128387451172,
|
2216 |
-
"learning_rate": 1.477920227920228e-06,
|
2217 |
-
"loss": 0.095,
|
2218 |
-
"step": 3040
|
2219 |
-
},
|
2220 |
-
{
|
2221 |
-
"epoch": 9.76,
|
2222 |
-
"grad_norm": 12.65188980102539,
|
2223 |
-
"learning_rate": 1.29985754985755e-06,
|
2224 |
-
"loss": 0.1245,
|
2225 |
-
"step": 3050
|
2226 |
-
},
|
2227 |
-
{
|
2228 |
-
"epoch": 9.792,
|
2229 |
-
"grad_norm": 9.408976554870605,
|
2230 |
-
"learning_rate": 1.121794871794872e-06,
|
2231 |
-
"loss": 0.112,
|
2232 |
-
"step": 3060
|
2233 |
-
},
|
2234 |
-
{
|
2235 |
-
"epoch": 9.824,
|
2236 |
-
"grad_norm": 5.554480075836182,
|
2237 |
-
"learning_rate": 9.437321937321938e-07,
|
2238 |
-
"loss": 0.0919,
|
2239 |
-
"step": 3070
|
2240 |
-
},
|
2241 |
-
{
|
2242 |
-
"epoch": 9.856,
|
2243 |
-
"grad_norm": 20.830698013305664,
|
2244 |
-
"learning_rate": 7.656695156695157e-07,
|
2245 |
-
"loss": 0.1212,
|
2246 |
-
"step": 3080
|
2247 |
-
},
|
2248 |
-
{
|
2249 |
-
"epoch": 9.888,
|
2250 |
-
"grad_norm": 8.195901870727539,
|
2251 |
-
"learning_rate": 5.876068376068376e-07,
|
2252 |
-
"loss": 0.1122,
|
2253 |
-
"step": 3090
|
2254 |
-
},
|
2255 |
-
{
|
2256 |
-
"epoch": 9.92,
|
2257 |
-
"grad_norm": 14.159358024597168,
|
2258 |
-
"learning_rate": 4.0954415954415953e-07,
|
2259 |
-
"loss": 0.1236,
|
2260 |
-
"step": 3100
|
2261 |
-
},
|
2262 |
-
{
|
2263 |
-
"epoch": 9.952,
|
2264 |
-
"grad_norm": 8.97334098815918,
|
2265 |
-
"learning_rate": 2.3148148148148148e-07,
|
2266 |
-
"loss": 0.1078,
|
2267 |
-
"step": 3110
|
2268 |
-
},
|
2269 |
-
{
|
2270 |
-
"epoch": 9.984,
|
2271 |
-
"grad_norm": 8.013399124145508,
|
2272 |
-
"learning_rate": 5.341880341880343e-08,
|
2273 |
-
"loss": 0.0805,
|
2274 |
-
"step": 3120
|
2275 |
-
},
|
2276 |
-
{
|
2277 |
-
"epoch": 9.984,
|
2278 |
-
"eval_accuracy": 0.74475,
|
2279 |
-
"eval_loss": 1.1793317794799805,
|
2280 |
-
"eval_runtime": 151.3727,
|
2281 |
-
"eval_samples_per_second": 26.425,
|
2282 |
-
"eval_steps_per_second": 0.826,
|
2283 |
-
"step": 3120
|
2284 |
},
|
2285 |
{
|
2286 |
-
"epoch":
|
2287 |
-
"step":
|
2288 |
-
"total_flos":
|
2289 |
-
"train_loss": 0.
|
2290 |
-
"train_runtime":
|
2291 |
-
"train_samples_per_second": 2.
|
2292 |
"train_steps_per_second": 0.023
|
2293 |
}
|
2294 |
],
|
2295 |
"logging_steps": 10,
|
2296 |
-
"max_steps":
|
2297 |
"num_input_tokens_seen": 0,
|
2298 |
-
"num_train_epochs":
|
2299 |
"save_steps": 500,
|
2300 |
"stateful_callbacks": {
|
2301 |
"TrainerControl": {
|
@@ -2309,7 +713,7 @@
|
|
2309 |
"attributes": {}
|
2310 |
}
|
2311 |
},
|
2312 |
-
"total_flos":
|
2313 |
"train_batch_size": 32,
|
2314 |
"trial_name": null,
|
2315 |
"trial_params": null
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.7025,
|
3 |
+
"best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-leukemia-08-2024.v1.2\\checkpoint-625",
|
4 |
+
"epoch": 2.9952,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 936,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 0.032,
|
13 |
+
"grad_norm": 5.205770969390869,
|
14 |
+
"learning_rate": 5.319148936170213e-06,
|
15 |
+
"loss": 0.6922,
|
16 |
"step": 10
|
17 |
},
|
18 |
{
|
19 |
"epoch": 0.064,
|
20 |
+
"grad_norm": 5.621664524078369,
|
21 |
+
"learning_rate": 1.0638297872340426e-05,
|
22 |
+
"loss": 0.6735,
|
23 |
"step": 20
|
24 |
},
|
25 |
{
|
26 |
"epoch": 0.096,
|
27 |
+
"grad_norm": 4.5251078605651855,
|
28 |
+
"learning_rate": 1.595744680851064e-05,
|
29 |
+
"loss": 0.6799,
|
30 |
"step": 30
|
31 |
},
|
32 |
{
|
33 |
"epoch": 0.128,
|
34 |
+
"grad_norm": 6.694261074066162,
|
35 |
+
"learning_rate": 2.1276595744680852e-05,
|
36 |
+
"loss": 0.6703,
|
37 |
"step": 40
|
38 |
},
|
39 |
{
|
40 |
"epoch": 0.16,
|
41 |
+
"grad_norm": 5.9614176750183105,
|
42 |
+
"learning_rate": 2.6595744680851064e-05,
|
43 |
+
"loss": 0.6335,
|
44 |
"step": 50
|
45 |
},
|
46 |
{
|
47 |
"epoch": 0.192,
|
48 |
+
"grad_norm": 9.891090393066406,
|
49 |
+
"learning_rate": 3.191489361702128e-05,
|
50 |
+
"loss": 0.6442,
|
51 |
"step": 60
|
52 |
},
|
53 |
{
|
54 |
"epoch": 0.224,
|
55 |
+
"grad_norm": 5.152803897857666,
|
56 |
+
"learning_rate": 3.723404255319149e-05,
|
57 |
+
"loss": 0.6192,
|
58 |
"step": 70
|
59 |
},
|
60 |
{
|
61 |
"epoch": 0.256,
|
62 |
+
"grad_norm": 19.81842613220215,
|
63 |
+
"learning_rate": 4.2553191489361704e-05,
|
64 |
+
"loss": 0.6061,
|
65 |
"step": 80
|
66 |
},
|
67 |
{
|
68 |
"epoch": 0.288,
|
69 |
+
"grad_norm": 12.223305702209473,
|
70 |
+
"learning_rate": 4.787234042553192e-05,
|
71 |
+
"loss": 0.5728,
|
72 |
"step": 90
|
73 |
},
|
74 |
{
|
75 |
"epoch": 0.32,
|
76 |
+
"grad_norm": 9.165103912353516,
|
77 |
+
"learning_rate": 4.96437054631829e-05,
|
78 |
+
"loss": 0.532,
|
79 |
"step": 100
|
80 |
},
|
81 |
{
|
82 |
"epoch": 0.352,
|
83 |
+
"grad_norm": 11.49010181427002,
|
84 |
+
"learning_rate": 4.90498812351544e-05,
|
85 |
+
"loss": 0.5192,
|
86 |
"step": 110
|
87 |
},
|
88 |
{
|
89 |
"epoch": 0.384,
|
90 |
+
"grad_norm": 16.154022216796875,
|
91 |
+
"learning_rate": 4.851543942992874e-05,
|
92 |
+
"loss": 0.5428,
|
93 |
"step": 120
|
94 |
},
|
95 |
{
|
96 |
"epoch": 0.416,
|
97 |
+
"grad_norm": 15.223276138305664,
|
98 |
+
"learning_rate": 4.792161520190024e-05,
|
99 |
+
"loss": 0.4874,
|
100 |
"step": 130
|
101 |
},
|
102 |
{
|
103 |
"epoch": 0.448,
|
104 |
+
"grad_norm": 15.795281410217285,
|
105 |
+
"learning_rate": 4.732779097387174e-05,
|
106 |
+
"loss": 0.4917,
|
107 |
"step": 140
|
108 |
},
|
109 |
{
|
110 |
"epoch": 0.48,
|
111 |
+
"grad_norm": 15.048002243041992,
|
112 |
+
"learning_rate": 4.673396674584323e-05,
|
113 |
+
"loss": 0.4914,
|
114 |
"step": 150
|
115 |
},
|
116 |
{
|
117 |
"epoch": 0.512,
|
118 |
+
"grad_norm": 9.113037109375,
|
119 |
+
"learning_rate": 4.6140142517814724e-05,
|
120 |
+
"loss": 0.4794,
|
121 |
"step": 160
|
122 |
},
|
123 |
{
|
124 |
"epoch": 0.544,
|
125 |
+
"grad_norm": 8.940505027770996,
|
126 |
+
"learning_rate": 4.5546318289786225e-05,
|
127 |
+
"loss": 0.467,
|
128 |
"step": 170
|
129 |
},
|
130 |
{
|
131 |
"epoch": 0.576,
|
132 |
+
"grad_norm": 10.99445915222168,
|
133 |
+
"learning_rate": 4.501187648456057e-05,
|
134 |
+
"loss": 0.4666,
|
135 |
"step": 180
|
136 |
},
|
137 |
{
|
138 |
"epoch": 0.608,
|
139 |
+
"grad_norm": 11.823775291442871,
|
140 |
+
"learning_rate": 4.441805225653207e-05,
|
141 |
+
"loss": 0.4256,
|
142 |
"step": 190
|
143 |
},
|
144 |
{
|
145 |
"epoch": 0.64,
|
146 |
+
"grad_norm": 32.22935485839844,
|
147 |
+
"learning_rate": 4.382422802850357e-05,
|
148 |
+
"loss": 0.412,
|
149 |
"step": 200
|
150 |
},
|
151 |
{
|
152 |
"epoch": 0.672,
|
153 |
+
"grad_norm": 31.271358489990234,
|
154 |
+
"learning_rate": 4.323040380047506e-05,
|
155 |
+
"loss": 0.493,
|
156 |
"step": 210
|
157 |
},
|
158 |
{
|
159 |
"epoch": 0.704,
|
160 |
+
"grad_norm": 7.7791900634765625,
|
161 |
+
"learning_rate": 4.263657957244656e-05,
|
162 |
+
"loss": 0.4073,
|
163 |
"step": 220
|
164 |
},
|
165 |
{
|
166 |
"epoch": 0.736,
|
167 |
+
"grad_norm": 10.995684623718262,
|
168 |
+
"learning_rate": 4.204275534441806e-05,
|
169 |
+
"loss": 0.3982,
|
170 |
"step": 230
|
171 |
},
|
172 |
{
|
173 |
"epoch": 0.768,
|
174 |
+
"grad_norm": 17.266887664794922,
|
175 |
+
"learning_rate": 4.144893111638955e-05,
|
176 |
+
"loss": 0.3794,
|
177 |
"step": 240
|
178 |
},
|
179 |
{
|
180 |
"epoch": 0.8,
|
181 |
+
"grad_norm": 7.906798362731934,
|
182 |
+
"learning_rate": 4.0855106888361044e-05,
|
183 |
+
"loss": 0.3911,
|
184 |
"step": 250
|
185 |
},
|
186 |
{
|
187 |
"epoch": 0.832,
|
188 |
+
"grad_norm": 13.777398109436035,
|
189 |
+
"learning_rate": 4.0261282660332545e-05,
|
190 |
+
"loss": 0.3829,
|
191 |
"step": 260
|
192 |
},
|
193 |
{
|
194 |
"epoch": 0.864,
|
195 |
+
"grad_norm": 15.481444358825684,
|
196 |
+
"learning_rate": 3.966745843230404e-05,
|
197 |
+
"loss": 0.3937,
|
198 |
"step": 270
|
199 |
},
|
200 |
{
|
201 |
"epoch": 0.896,
|
202 |
+
"grad_norm": 11.324766159057617,
|
203 |
+
"learning_rate": 3.907363420427554e-05,
|
204 |
+
"loss": 0.3503,
|
205 |
"step": 280
|
206 |
},
|
207 |
{
|
208 |
"epoch": 0.928,
|
209 |
+
"grad_norm": 19.757225036621094,
|
210 |
+
"learning_rate": 3.847980997624703e-05,
|
211 |
+
"loss": 0.3484,
|
212 |
"step": 290
|
213 |
},
|
214 |
{
|
215 |
"epoch": 0.96,
|
216 |
+
"grad_norm": 9.243185997009277,
|
217 |
+
"learning_rate": 3.7885985748218526e-05,
|
218 |
+
"loss": 0.3803,
|
219 |
"step": 300
|
220 |
},
|
221 |
{
|
222 |
"epoch": 0.992,
|
223 |
+
"grad_norm": 9.08983325958252,
|
224 |
+
"learning_rate": 3.7292161520190026e-05,
|
225 |
+
"loss": 0.3704,
|
226 |
"step": 310
|
227 |
},
|
228 |
{
|
229 |
"epoch": 0.9984,
|
230 |
+
"eval_accuracy": 0.63875,
|
231 |
+
"eval_loss": 0.7880752682685852,
|
232 |
+
"eval_runtime": 187.0148,
|
233 |
+
"eval_samples_per_second": 21.389,
|
234 |
+
"eval_steps_per_second": 0.668,
|
235 |
"step": 312
|
236 |
},
|
237 |
{
|
238 |
"epoch": 1.024,
|
239 |
+
"grad_norm": 45.11425018310547,
|
240 |
+
"learning_rate": 3.669833729216152e-05,
|
241 |
+
"loss": 0.3855,
|
242 |
"step": 320
|
243 |
},
|
244 |
{
|
245 |
"epoch": 1.056,
|
246 |
+
"grad_norm": 26.53282928466797,
|
247 |
+
"learning_rate": 3.6104513064133013e-05,
|
248 |
+
"loss": 0.3922,
|
249 |
"step": 330
|
250 |
},
|
251 |
{
|
252 |
"epoch": 1.088,
|
253 |
+
"grad_norm": 9.646315574645996,
|
254 |
+
"learning_rate": 3.5510688836104514e-05,
|
255 |
+
"loss": 0.4238,
|
256 |
"step": 340
|
257 |
},
|
258 |
{
|
259 |
"epoch": 1.12,
|
260 |
+
"grad_norm": 21.654006958007812,
|
261 |
+
"learning_rate": 3.4916864608076014e-05,
|
262 |
+
"loss": 0.3569,
|
263 |
"step": 350
|
264 |
},
|
265 |
{
|
266 |
"epoch": 1.152,
|
267 |
+
"grad_norm": 10.348204612731934,
|
268 |
+
"learning_rate": 3.432304038004751e-05,
|
269 |
+
"loss": 0.3122,
|
270 |
"step": 360
|
271 |
},
|
272 |
{
|
273 |
"epoch": 1.184,
|
274 |
+
"grad_norm": 9.6500883102417,
|
275 |
+
"learning_rate": 3.372921615201901e-05,
|
276 |
+
"loss": 0.327,
|
277 |
"step": 370
|
278 |
},
|
279 |
{
|
280 |
"epoch": 1.216,
|
281 |
+
"grad_norm": 16.52385902404785,
|
282 |
+
"learning_rate": 3.31353919239905e-05,
|
283 |
+
"loss": 0.4007,
|
284 |
"step": 380
|
285 |
},
|
286 |
{
|
287 |
"epoch": 1.248,
|
288 |
+
"grad_norm": 13.382709503173828,
|
289 |
+
"learning_rate": 3.2541567695961995e-05,
|
290 |
+
"loss": 0.3414,
|
291 |
"step": 390
|
292 |
},
|
293 |
{
|
294 |
"epoch": 1.28,
|
295 |
+
"grad_norm": 58.999324798583984,
|
296 |
+
"learning_rate": 3.1947743467933496e-05,
|
297 |
+
"loss": 0.3501,
|
298 |
"step": 400
|
299 |
},
|
300 |
{
|
301 |
"epoch": 1.312,
|
302 |
+
"grad_norm": 7.981963157653809,
|
303 |
+
"learning_rate": 3.135391923990499e-05,
|
304 |
+
"loss": 0.3822,
|
305 |
"step": 410
|
306 |
},
|
307 |
{
|
308 |
"epoch": 1.3439999999999999,
|
309 |
+
"grad_norm": 9.790070533752441,
|
310 |
+
"learning_rate": 3.076009501187649e-05,
|
311 |
+
"loss": 0.3383,
|
312 |
"step": 420
|
313 |
},
|
314 |
{
|
315 |
"epoch": 1.376,
|
316 |
+
"grad_norm": 8.631843566894531,
|
317 |
+
"learning_rate": 3.0166270783847983e-05,
|
318 |
+
"loss": 0.335,
|
319 |
"step": 430
|
320 |
},
|
321 |
{
|
322 |
"epoch": 1.408,
|
323 |
+
"grad_norm": 9.060861587524414,
|
324 |
+
"learning_rate": 2.9572446555819477e-05,
|
325 |
+
"loss": 0.3282,
|
326 |
"step": 440
|
327 |
},
|
328 |
{
|
329 |
"epoch": 1.44,
|
330 |
+
"grad_norm": 15.991608619689941,
|
331 |
+
"learning_rate": 2.8978622327790977e-05,
|
332 |
+
"loss": 0.3037,
|
333 |
"step": 450
|
334 |
},
|
335 |
{
|
336 |
"epoch": 1.472,
|
337 |
+
"grad_norm": 35.242740631103516,
|
338 |
+
"learning_rate": 2.838479809976247e-05,
|
339 |
+
"loss": 0.3321,
|
340 |
"step": 460
|
341 |
},
|
342 |
{
|
343 |
"epoch": 1.504,
|
344 |
+
"grad_norm": 10.073174476623535,
|
345 |
+
"learning_rate": 2.7790973871733968e-05,
|
346 |
+
"loss": 0.3023,
|
347 |
"step": 470
|
348 |
},
|
349 |
{
|
350 |
"epoch": 1.536,
|
351 |
+
"grad_norm": 19.67376708984375,
|
352 |
+
"learning_rate": 2.7197149643705465e-05,
|
353 |
+
"loss": 0.3168,
|
354 |
"step": 480
|
355 |
},
|
356 |
{
|
357 |
"epoch": 1.568,
|
358 |
+
"grad_norm": 6.408135890960693,
|
359 |
+
"learning_rate": 2.6603325415676962e-05,
|
360 |
+
"loss": 0.306,
|
361 |
"step": 490
|
362 |
},
|
363 |
{
|
364 |
"epoch": 1.6,
|
365 |
+
"grad_norm": 8.863042831420898,
|
366 |
+
"learning_rate": 2.6009501187648455e-05,
|
367 |
+
"loss": 0.3103,
|
368 |
"step": 500
|
369 |
},
|
370 |
{
|
371 |
"epoch": 1.6320000000000001,
|
372 |
+
"grad_norm": 17.778520584106445,
|
373 |
+
"learning_rate": 2.5415676959619956e-05,
|
374 |
+
"loss": 0.3123,
|
375 |
"step": 510
|
376 |
},
|
377 |
{
|
378 |
"epoch": 1.6640000000000001,
|
379 |
+
"grad_norm": 11.152641296386719,
|
380 |
+
"learning_rate": 2.482185273159145e-05,
|
381 |
+
"loss": 0.304,
|
382 |
"step": 520
|
383 |
},
|
384 |
{
|
385 |
"epoch": 1.696,
|
386 |
+
"grad_norm": 18.749303817749023,
|
387 |
+
"learning_rate": 2.4228028503562946e-05,
|
388 |
+
"loss": 0.2905,
|
389 |
"step": 530
|
390 |
},
|
391 |
{
|
392 |
"epoch": 1.728,
|
393 |
+
"grad_norm": 26.219440460205078,
|
394 |
+
"learning_rate": 2.3634204275534443e-05,
|
395 |
+
"loss": 0.3163,
|
396 |
"step": 540
|
397 |
},
|
398 |
{
|
399 |
"epoch": 1.76,
|
400 |
+
"grad_norm": 16.212156295776367,
|
401 |
+
"learning_rate": 2.3040380047505937e-05,
|
402 |
+
"loss": 0.3073,
|
403 |
"step": 550
|
404 |
},
|
405 |
{
|
406 |
"epoch": 1.792,
|
407 |
+
"grad_norm": 11.084125518798828,
|
408 |
+
"learning_rate": 2.2446555819477437e-05,
|
409 |
+
"loss": 0.3166,
|
410 |
"step": 560
|
411 |
},
|
412 |
{
|
413 |
"epoch": 1.8239999999999998,
|
414 |
+
"grad_norm": 12.52534008026123,
|
415 |
+
"learning_rate": 2.1852731591448934e-05,
|
416 |
+
"loss": 0.293,
|
417 |
"step": 570
|
418 |
},
|
419 |
{
|
420 |
"epoch": 1.8559999999999999,
|
421 |
+
"grad_norm": 20.346576690673828,
|
422 |
+
"learning_rate": 2.1258907363420428e-05,
|
423 |
+
"loss": 0.2948,
|
424 |
"step": 580
|
425 |
},
|
426 |
{
|
427 |
"epoch": 1.888,
|
428 |
+
"grad_norm": 9.408613204956055,
|
429 |
+
"learning_rate": 2.0665083135391925e-05,
|
430 |
+
"loss": 0.3081,
|
431 |
"step": 590
|
432 |
},
|
433 |
{
|
434 |
"epoch": 1.92,
|
435 |
+
"grad_norm": 27.265464782714844,
|
436 |
+
"learning_rate": 2.0071258907363422e-05,
|
437 |
+
"loss": 0.3084,
|
438 |
"step": 600
|
439 |
},
|
440 |
{
|
441 |
"epoch": 1.952,
|
442 |
+
"grad_norm": 8.094114303588867,
|
443 |
+
"learning_rate": 1.947743467933492e-05,
|
444 |
+
"loss": 0.3047,
|
445 |
"step": 610
|
446 |
},
|
447 |
{
|
448 |
"epoch": 1.984,
|
449 |
+
"grad_norm": 13.893214225769043,
|
450 |
+
"learning_rate": 1.8883610451306412e-05,
|
451 |
+
"loss": 0.2726,
|
452 |
"step": 620
|
453 |
},
|
454 |
{
|
455 |
"epoch": 2.0,
|
456 |
+
"eval_accuracy": 0.7025,
|
457 |
+
"eval_loss": 0.830498456954956,
|
458 |
+
"eval_runtime": 154.4758,
|
459 |
+
"eval_samples_per_second": 25.894,
|
460 |
+
"eval_steps_per_second": 0.809,
|
461 |
"step": 625
|
462 |
},
|
463 |
{
|
464 |
"epoch": 2.016,
|
465 |
+
"grad_norm": 5.93212890625,
|
466 |
+
"learning_rate": 1.828978622327791e-05,
|
467 |
+
"loss": 0.2716,
|
468 |
"step": 630
|
469 |
},
|
470 |
{
|
471 |
"epoch": 2.048,
|
472 |
+
"grad_norm": 33.18698501586914,
|
473 |
+
"learning_rate": 1.7695961995249406e-05,
|
474 |
+
"loss": 0.2845,
|
475 |
"step": 640
|
476 |
},
|
477 |
{
|
478 |
"epoch": 2.08,
|
479 |
+
"grad_norm": 23.357540130615234,
|
480 |
+
"learning_rate": 1.7102137767220903e-05,
|
481 |
+
"loss": 0.3093,
|
482 |
"step": 650
|
483 |
},
|
484 |
{
|
485 |
"epoch": 2.112,
|
486 |
+
"grad_norm": 14.696449279785156,
|
487 |
+
"learning_rate": 1.65083135391924e-05,
|
488 |
+
"loss": 0.2886,
|
489 |
"step": 660
|
490 |
},
|
491 |
{
|
492 |
"epoch": 2.144,
|
493 |
+
"grad_norm": 32.7164192199707,
|
494 |
+
"learning_rate": 1.5914489311163897e-05,
|
495 |
+
"loss": 0.2865,
|
496 |
"step": 670
|
497 |
},
|
498 |
{
|
499 |
"epoch": 2.176,
|
500 |
+
"grad_norm": 11.33267879486084,
|
501 |
+
"learning_rate": 1.5320665083135394e-05,
|
502 |
+
"loss": 0.2616,
|
503 |
"step": 680
|
504 |
},
|
505 |
{
|
506 |
"epoch": 2.208,
|
507 |
+
"grad_norm": 10.119343757629395,
|
508 |
+
"learning_rate": 1.4726840855106888e-05,
|
509 |
+
"loss": 0.2768,
|
510 |
"step": 690
|
511 |
},
|
512 |
{
|
513 |
"epoch": 2.24,
|
514 |
+
"grad_norm": 12.684237480163574,
|
515 |
+
"learning_rate": 1.4133016627078385e-05,
|
516 |
+
"loss": 0.2332,
|
517 |
"step": 700
|
518 |
},
|
519 |
{
|
520 |
"epoch": 2.2720000000000002,
|
521 |
+
"grad_norm": 17.10093879699707,
|
522 |
+
"learning_rate": 1.3539192399049882e-05,
|
523 |
+
"loss": 0.2648,
|
524 |
"step": 710
|
525 |
},
|
526 |
{
|
527 |
"epoch": 2.304,
|
528 |
+
"grad_norm": 22.309677124023438,
|
529 |
+
"learning_rate": 1.2945368171021377e-05,
|
530 |
+
"loss": 0.2794,
|
531 |
"step": 720
|
532 |
},
|
533 |
{
|
534 |
"epoch": 2.336,
|
535 |
+
"grad_norm": 15.592386245727539,
|
536 |
+
"learning_rate": 1.2351543942992874e-05,
|
537 |
+
"loss": 0.2826,
|
538 |
"step": 730
|
539 |
},
|
540 |
{
|
541 |
"epoch": 2.368,
|
542 |
+
"grad_norm": 10.027396202087402,
|
543 |
+
"learning_rate": 1.1757719714964371e-05,
|
544 |
+
"loss": 0.2662,
|
545 |
"step": 740
|
546 |
},
|
547 |
{
|
548 |
"epoch": 2.4,
|
549 |
+
"grad_norm": 7.71236515045166,
|
550 |
+
"learning_rate": 1.1163895486935868e-05,
|
551 |
+
"loss": 0.2802,
|
552 |
"step": 750
|
553 |
},
|
554 |
{
|
555 |
"epoch": 2.432,
|
556 |
+
"grad_norm": 8.326642036437988,
|
557 |
+
"learning_rate": 1.0570071258907365e-05,
|
558 |
+
"loss": 0.2644,
|
559 |
"step": 760
|
560 |
},
|
561 |
{
|
562 |
"epoch": 2.464,
|
563 |
+
"grad_norm": 17.820472717285156,
|
564 |
+
"learning_rate": 9.97624703087886e-06,
|
565 |
+
"loss": 0.2649,
|
566 |
"step": 770
|
567 |
},
|
568 |
{
|
569 |
"epoch": 2.496,
|
570 |
+
"grad_norm": 10.056536674499512,
|
571 |
+
"learning_rate": 9.382422802850356e-06,
|
572 |
+
"loss": 0.248,
|
573 |
"step": 780
|
574 |
},
|
575 |
{
|
576 |
"epoch": 2.528,
|
577 |
+
"grad_norm": 30.569965362548828,
|
578 |
+
"learning_rate": 8.788598574821852e-06,
|
579 |
+
"loss": 0.2303,
|
580 |
"step": 790
|
581 |
},
|
582 |
{
|
583 |
"epoch": 2.56,
|
584 |
+
"grad_norm": 7.8709797859191895,
|
585 |
+
"learning_rate": 8.19477434679335e-06,
|
586 |
+
"loss": 0.2537,
|
587 |
"step": 800
|
588 |
},
|
589 |
{
|
590 |
"epoch": 2.592,
|
591 |
+
"grad_norm": 15.642313957214355,
|
592 |
+
"learning_rate": 7.6009501187648464e-06,
|
593 |
+
"loss": 0.2726,
|
594 |
"step": 810
|
595 |
},
|
596 |
{
|
597 |
"epoch": 2.624,
|
598 |
+
"grad_norm": 8.542279243469238,
|
599 |
+
"learning_rate": 7.007125890736342e-06,
|
600 |
+
"loss": 0.27,
|
601 |
"step": 820
|
602 |
},
|
603 |
{
|
604 |
"epoch": 2.656,
|
605 |
+
"grad_norm": 7.526516914367676,
|
606 |
+
"learning_rate": 6.4133016627078396e-06,
|
607 |
+
"loss": 0.2742,
|
608 |
"step": 830
|
609 |
},
|
610 |
{
|
611 |
"epoch": 2.6879999999999997,
|
612 |
+
"grad_norm": 9.183687210083008,
|
613 |
+
"learning_rate": 5.819477434679335e-06,
|
614 |
+
"loss": 0.2477,
|
615 |
"step": 840
|
616 |
},
|
617 |
{
|
618 |
"epoch": 2.7199999999999998,
|
619 |
+
"grad_norm": 13.24605655670166,
|
620 |
+
"learning_rate": 5.225653206650832e-06,
|
621 |
+
"loss": 0.2462,
|
622 |
"step": 850
|
623 |
},
|
624 |
{
|
625 |
"epoch": 2.752,
|
626 |
+
"grad_norm": 14.802075386047363,
|
627 |
+
"learning_rate": 4.631828978622328e-06,
|
628 |
+
"loss": 0.2451,
|
629 |
"step": 860
|
630 |
},
|
631 |
{
|
632 |
"epoch": 2.784,
|
633 |
+
"grad_norm": 8.528770446777344,
|
634 |
+
"learning_rate": 4.038004750593825e-06,
|
635 |
+
"loss": 0.2221,
|
636 |
"step": 870
|
637 |
},
|
638 |
{
|
639 |
"epoch": 2.816,
|
640 |
+
"grad_norm": 12.269448280334473,
|
641 |
+
"learning_rate": 3.4441805225653207e-06,
|
642 |
+
"loss": 0.2522,
|
643 |
"step": 880
|
644 |
},
|
645 |
{
|
646 |
"epoch": 2.848,
|
647 |
+
"grad_norm": 30.11684226989746,
|
648 |
+
"learning_rate": 2.850356294536817e-06,
|
649 |
+
"loss": 0.2509,
|
650 |
"step": 890
|
651 |
},
|
652 |
{
|
653 |
"epoch": 2.88,
|
654 |
+
"grad_norm": 25.639724731445312,
|
655 |
+
"learning_rate": 2.2565320665083133e-06,
|
656 |
+
"loss": 0.2423,
|
657 |
"step": 900
|
658 |
},
|
659 |
{
|
660 |
"epoch": 2.912,
|
661 |
+
"grad_norm": 10.527947425842285,
|
662 |
+
"learning_rate": 1.6627078384798101e-06,
|
663 |
+
"loss": 0.2528,
|
664 |
"step": 910
|
665 |
},
|
666 |
{
|
667 |
"epoch": 2.944,
|
668 |
+
"grad_norm": 18.0466365814209,
|
669 |
+
"learning_rate": 1.0688836104513065e-06,
|
670 |
+
"loss": 0.257,
|
671 |
"step": 920
|
672 |
},
|
673 |
{
|
674 |
"epoch": 2.976,
|
675 |
+
"grad_norm": 12.547955513000488,
|
676 |
+
"learning_rate": 4.750593824228029e-07,
|
677 |
+
"loss": 0.2103,
|
678 |
"step": 930
|
679 |
},
|
680 |
{
|
681 |
+
"epoch": 2.9952,
|
682 |
+
"eval_accuracy": 0.692,
|
683 |
+
"eval_loss": 0.870841920375824,
|
684 |
+
"eval_runtime": 155.4401,
|
685 |
+
"eval_samples_per_second": 25.733,
|
686 |
+
"eval_steps_per_second": 0.804,
|
687 |
+
"step": 936
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
688 |
},
|
689 |
{
|
690 |
+
"epoch": 2.9952,
|
691 |
+
"step": 936,
|
692 |
+
"total_flos": 2.9779463284174356e+18,
|
693 |
+
"train_loss": 0.36125070786374247,
|
694 |
+
"train_runtime": 41238.7802,
|
695 |
+
"train_samples_per_second": 2.91,
|
696 |
"train_steps_per_second": 0.023
|
697 |
}
|
698 |
],
|
699 |
"logging_steps": 10,
|
700 |
+
"max_steps": 936,
|
701 |
"num_input_tokens_seen": 0,
|
702 |
+
"num_train_epochs": 3,
|
703 |
"save_steps": 500,
|
704 |
"stateful_callbacks": {
|
705 |
"TrainerControl": {
|
|
|
713 |
"attributes": {}
|
714 |
}
|
715 |
},
|
716 |
+
"total_flos": 2.9779463284174356e+18,
|
717 |
"train_batch_size": 32,
|
718 |
"trial_name": null,
|
719 |
"trial_params": null
|