checkpoint-3272/trainer_state.json
57.0 KB · 2342 lines · json Raw
1 {
2 "best_global_step": 3272,
3 "best_metric": 0.9839899931917298,
4 "best_model_checkpoint": "voice_gender_classification/checkpoint-3272",
5 "epoch": 1.0,
6 "eval_steps": 500,
7 "global_step": 3272,
8 "is_hyper_param_search": false,
9 "is_local_process_zero": true,
10 "is_world_process_zero": true,
11 "log_history": [
12 {
13 "epoch": 0.0003056234718826406,
14 "grad_norm": 1.0770392417907715,
15 "learning_rate": 6.000000000000001e-07,
16 "loss": 0.6859,
17 "step": 1
18 },
19 {
20 "epoch": 0.003056234718826406,
21 "grad_norm": 1.2716485261917114,
22 "learning_rate": 6e-06,
23 "loss": 0.7011,
24 "step": 10
25 },
26 {
27 "epoch": 0.006112469437652812,
28 "grad_norm": 1.0392799377441406,
29 "learning_rate": 1.2e-05,
30 "loss": 0.6944,
31 "step": 20
32 },
33 {
34 "epoch": 0.009168704156479218,
35 "grad_norm": 0.628126859664917,
36 "learning_rate": 1.8e-05,
37 "loss": 0.6592,
38 "step": 30
39 },
40 {
41 "epoch": 0.012224938875305624,
42 "grad_norm": 2.359282970428467,
43 "learning_rate": 2.4e-05,
44 "loss": 0.6314,
45 "step": 40
46 },
47 {
48 "epoch": 0.015281173594132029,
49 "grad_norm": 4.273380756378174,
50 "learning_rate": 3e-05,
51 "loss": 0.5976,
52 "step": 50
53 },
54 {
55 "epoch": 0.018337408312958436,
56 "grad_norm": 1.5336124897003174,
57 "learning_rate": 2.9906890130353816e-05,
58 "loss": 0.4168,
59 "step": 60
60 },
61 {
62 "epoch": 0.021393643031784843,
63 "grad_norm": 1.772170066833496,
64 "learning_rate": 2.9823091247672255e-05,
65 "loss": 0.3254,
66 "step": 70
67 },
68 {
69 "epoch": 0.02444987775061125,
70 "grad_norm": 1.347057580947876,
71 "learning_rate": 2.973929236499069e-05,
72 "loss": 0.1641,
73 "step": 80
74 },
75 {
76 "epoch": 0.02750611246943765,
77 "grad_norm": 0.8756974339485168,
78 "learning_rate": 2.9646182495344507e-05,
79 "loss": 0.2623,
80 "step": 90
81 },
82 {
83 "epoch": 0.030562347188264057,
84 "grad_norm": 0.594032347202301,
85 "learning_rate": 2.9553072625698325e-05,
86 "loss": 0.0924,
87 "step": 100
88 },
89 {
90 "epoch": 0.03361858190709047,
91 "grad_norm": 15.337114334106445,
92 "learning_rate": 2.946927374301676e-05,
93 "loss": 0.4399,
94 "step": 110
95 },
96 {
97 "epoch": 0.03667481662591687,
98 "grad_norm": 0.4488385319709778,
99 "learning_rate": 2.937616387337058e-05,
100 "loss": 0.044,
101 "step": 120
102 },
103 {
104 "epoch": 0.03973105134474328,
105 "grad_norm": 1.3066495656967163,
106 "learning_rate": 2.9283054003724395e-05,
107 "loss": 0.2036,
108 "step": 130
109 },
110 {
111 "epoch": 0.042787286063569685,
112 "grad_norm": 0.36801964044570923,
113 "learning_rate": 2.919925512104283e-05,
114 "loss": 0.1376,
115 "step": 140
116 },
117 {
118 "epoch": 0.04584352078239609,
119 "grad_norm": 0.35944098234176636,
120 "learning_rate": 2.910614525139665e-05,
121 "loss": 0.2113,
122 "step": 150
123 },
124 {
125 "epoch": 0.0488997555012225,
126 "grad_norm": 0.32566192746162415,
127 "learning_rate": 2.9013035381750465e-05,
128 "loss": 0.1613,
129 "step": 160
130 },
131 {
132 "epoch": 0.0519559902200489,
133 "grad_norm": 0.3194483816623688,
134 "learning_rate": 2.8919925512104284e-05,
135 "loss": 0.1142,
136 "step": 170
137 },
138 {
139 "epoch": 0.0550122249388753,
140 "grad_norm": 0.35201185941696167,
141 "learning_rate": 2.88268156424581e-05,
142 "loss": 0.2022,
143 "step": 180
144 },
145 {
146 "epoch": 0.05806845965770171,
147 "grad_norm": 0.3429718017578125,
148 "learning_rate": 2.8733705772811922e-05,
149 "loss": 0.1527,
150 "step": 190
151 },
152 {
153 "epoch": 0.061124694376528114,
154 "grad_norm": 0.3786994516849518,
155 "learning_rate": 2.8640595903165737e-05,
156 "loss": 0.1623,
157 "step": 200
158 },
159 {
160 "epoch": 0.06418092909535453,
161 "grad_norm": 0.4179794192314148,
162 "learning_rate": 2.8547486033519556e-05,
163 "loss": 0.1551,
164 "step": 210
165 },
166 {
167 "epoch": 0.06723716381418093,
168 "grad_norm": 83.91354370117188,
169 "learning_rate": 2.845437616387337e-05,
170 "loss": 0.2994,
171 "step": 220
172 },
173 {
174 "epoch": 0.07029339853300734,
175 "grad_norm": 0.2769756615161896,
176 "learning_rate": 2.8361266294227187e-05,
177 "loss": 0.1539,
178 "step": 230
179 },
180 {
181 "epoch": 0.07334963325183375,
182 "grad_norm": 0.2955838441848755,
183 "learning_rate": 2.8268156424581005e-05,
184 "loss": 0.1634,
185 "step": 240
186 },
187 {
188 "epoch": 0.07640586797066015,
189 "grad_norm": 0.2536284029483795,
190 "learning_rate": 2.817504655493482e-05,
191 "loss": 0.1137,
192 "step": 250
193 },
194 {
195 "epoch": 0.07946210268948656,
196 "grad_norm": 1.7733937501907349,
197 "learning_rate": 2.8081936685288643e-05,
198 "loss": 0.3204,
199 "step": 260
200 },
201 {
202 "epoch": 0.08251833740831296,
203 "grad_norm": 0.36826714873313904,
204 "learning_rate": 2.798882681564246e-05,
205 "loss": 0.1624,
206 "step": 270
207 },
208 {
209 "epoch": 0.08557457212713937,
210 "grad_norm": 0.3160645067691803,
211 "learning_rate": 2.7895716945996277e-05,
212 "loss": 0.1015,
213 "step": 280
214 },
215 {
216 "epoch": 0.08863080684596578,
217 "grad_norm": 0.24478046596050262,
218 "learning_rate": 2.7802607076350093e-05,
219 "loss": 0.1955,
220 "step": 290
221 },
222 {
223 "epoch": 0.09168704156479218,
224 "grad_norm": 14.753984451293945,
225 "learning_rate": 2.770949720670391e-05,
226 "loss": 0.1996,
227 "step": 300
228 },
229 {
230 "epoch": 0.09474327628361859,
231 "grad_norm": 1.5840058326721191,
232 "learning_rate": 2.7616387337057727e-05,
233 "loss": 0.1386,
234 "step": 310
235 },
236 {
237 "epoch": 0.097799511002445,
238 "grad_norm": 0.23863820731639862,
239 "learning_rate": 2.752327746741155e-05,
240 "loss": 0.2088,
241 "step": 320
242 },
243 {
244 "epoch": 0.1008557457212714,
245 "grad_norm": 1.5646458864212036,
246 "learning_rate": 2.7430167597765364e-05,
247 "loss": 0.16,
248 "step": 330
249 },
250 {
251 "epoch": 0.1039119804400978,
252 "grad_norm": 0.25270429253578186,
253 "learning_rate": 2.7337057728119183e-05,
254 "loss": 0.0747,
255 "step": 340
256 },
257 {
258 "epoch": 0.1069682151589242,
259 "grad_norm": 1.4299941062927246,
260 "learning_rate": 2.7243947858473e-05,
261 "loss": 0.1075,
262 "step": 350
263 },
264 {
265 "epoch": 0.1100244498777506,
266 "grad_norm": 20.335739135742188,
267 "learning_rate": 2.7150837988826817e-05,
268 "loss": 0.2605,
269 "step": 360
270 },
271 {
272 "epoch": 0.11308068459657701,
273 "grad_norm": 0.2346445471048355,
274 "learning_rate": 2.7057728119180633e-05,
275 "loss": 0.0954,
276 "step": 370
277 },
278 {
279 "epoch": 0.11613691931540342,
280 "grad_norm": 0.23116770386695862,
281 "learning_rate": 2.696461824953445e-05,
282 "loss": 0.0641,
283 "step": 380
284 },
285 {
286 "epoch": 0.11919315403422982,
287 "grad_norm": 1.6760469675064087,
288 "learning_rate": 2.687150837988827e-05,
289 "loss": 0.0678,
290 "step": 390
291 },
292 {
293 "epoch": 0.12224938875305623,
294 "grad_norm": 0.1794356405735016,
295 "learning_rate": 2.6778398510242085e-05,
296 "loss": 0.0609,
297 "step": 400
298 },
299 {
300 "epoch": 0.12530562347188265,
301 "grad_norm": 0.18168634176254272,
302 "learning_rate": 2.6685288640595904e-05,
303 "loss": 0.0611,
304 "step": 410
305 },
306 {
307 "epoch": 0.12836185819070906,
308 "grad_norm": 0.2647036612033844,
309 "learning_rate": 2.659217877094972e-05,
310 "loss": 0.0598,
311 "step": 420
312 },
313 {
314 "epoch": 0.13141809290953546,
315 "grad_norm": 0.2929624915122986,
316 "learning_rate": 2.649906890130354e-05,
317 "loss": 0.1172,
318 "step": 430
319 },
320 {
321 "epoch": 0.13447432762836187,
322 "grad_norm": 0.7274528741836548,
323 "learning_rate": 2.6405959031657354e-05,
324 "loss": 0.1237,
325 "step": 440
326 },
327 {
328 "epoch": 0.13753056234718827,
329 "grad_norm": 1.7421795129776,
330 "learning_rate": 2.6312849162011176e-05,
331 "loss": 0.2341,
332 "step": 450
333 },
334 {
335 "epoch": 0.14058679706601468,
336 "grad_norm": 0.6214892268180847,
337 "learning_rate": 2.621973929236499e-05,
338 "loss": 0.1648,
339 "step": 460
340 },
341 {
342 "epoch": 0.14364303178484109,
343 "grad_norm": 0.21029511094093323,
344 "learning_rate": 2.612662942271881e-05,
345 "loss": 0.0573,
346 "step": 470
347 },
348 {
349 "epoch": 0.1466992665036675,
350 "grad_norm": 0.181521475315094,
351 "learning_rate": 2.6033519553072625e-05,
352 "loss": 0.2243,
353 "step": 480
354 },
355 {
356 "epoch": 0.1497555012224939,
357 "grad_norm": 0.2883399724960327,
358 "learning_rate": 2.5940409683426444e-05,
359 "loss": 0.1662,
360 "step": 490
361 },
362 {
363 "epoch": 0.1528117359413203,
364 "grad_norm": 0.1559794545173645,
365 "learning_rate": 2.584729981378026e-05,
366 "loss": 0.1764,
367 "step": 500
368 },
369 {
370 "epoch": 0.1558679706601467,
371 "grad_norm": 0.4264441728591919,
372 "learning_rate": 2.575418994413408e-05,
373 "loss": 0.0789,
374 "step": 510
375 },
376 {
377 "epoch": 0.15892420537897312,
378 "grad_norm": 0.320130318403244,
379 "learning_rate": 2.5661080074487897e-05,
380 "loss": 0.1565,
381 "step": 520
382 },
383 {
384 "epoch": 0.16198044009779952,
385 "grad_norm": 6.056375503540039,
386 "learning_rate": 2.5567970204841716e-05,
387 "loss": 0.2424,
388 "step": 530
389 },
390 {
391 "epoch": 0.16503667481662593,
392 "grad_norm": 4.100960731506348,
393 "learning_rate": 2.547486033519553e-05,
394 "loss": 0.0672,
395 "step": 540
396 },
397 {
398 "epoch": 0.16809290953545233,
399 "grad_norm": 4.6566901206970215,
400 "learning_rate": 2.538175046554935e-05,
401 "loss": 0.1897,
402 "step": 550
403 },
404 {
405 "epoch": 0.17114914425427874,
406 "grad_norm": 3.481842279434204,
407 "learning_rate": 2.5288640595903165e-05,
408 "loss": 0.1389,
409 "step": 560
410 },
411 {
412 "epoch": 0.17420537897310515,
413 "grad_norm": 0.45788800716400146,
414 "learning_rate": 2.519553072625698e-05,
415 "loss": 0.0967,
416 "step": 570
417 },
418 {
419 "epoch": 0.17726161369193155,
420 "grad_norm": 0.1905224472284317,
421 "learning_rate": 2.5102420856610803e-05,
422 "loss": 0.0771,
423 "step": 580
424 },
425 {
426 "epoch": 0.18031784841075796,
427 "grad_norm": 0.1771785467863083,
428 "learning_rate": 2.5009310986964618e-05,
429 "loss": 0.1548,
430 "step": 590
431 },
432 {
433 "epoch": 0.18337408312958436,
434 "grad_norm": 0.18358071148395538,
435 "learning_rate": 2.4916201117318437e-05,
436 "loss": 0.2571,
437 "step": 600
438 },
439 {
440 "epoch": 0.18643031784841077,
441 "grad_norm": 0.3116374611854553,
442 "learning_rate": 2.4823091247672252e-05,
443 "loss": 0.1225,
444 "step": 610
445 },
446 {
447 "epoch": 0.18948655256723718,
448 "grad_norm": 0.2802579998970032,
449 "learning_rate": 2.472998137802607e-05,
450 "loss": 0.0988,
451 "step": 620
452 },
453 {
454 "epoch": 0.19254278728606358,
455 "grad_norm": 43.23971176147461,
456 "learning_rate": 2.4636871508379887e-05,
457 "loss": 0.2569,
458 "step": 630
459 },
460 {
461 "epoch": 0.19559902200489,
462 "grad_norm": 0.25425320863723755,
463 "learning_rate": 2.454376163873371e-05,
464 "loss": 0.0759,
465 "step": 640
466 },
467 {
468 "epoch": 0.1986552567237164,
469 "grad_norm": 0.16143210232257843,
470 "learning_rate": 2.4450651769087524e-05,
471 "loss": 0.0666,
472 "step": 650
473 },
474 {
475 "epoch": 0.2017114914425428,
476 "grad_norm": 0.15112750232219696,
477 "learning_rate": 2.4357541899441343e-05,
478 "loss": 0.1222,
479 "step": 660
480 },
481 {
482 "epoch": 0.2047677261613692,
483 "grad_norm": 0.1990450918674469,
484 "learning_rate": 2.4264432029795158e-05,
485 "loss": 0.1612,
486 "step": 670
487 },
488 {
489 "epoch": 0.2078239608801956,
490 "grad_norm": 0.23198798298835754,
491 "learning_rate": 2.4171322160148977e-05,
492 "loss": 0.0139,
493 "step": 680
494 },
495 {
496 "epoch": 0.210880195599022,
497 "grad_norm": 0.2512992024421692,
498 "learning_rate": 2.4078212290502792e-05,
499 "loss": 0.2073,
500 "step": 690
501 },
502 {
503 "epoch": 0.2139364303178484,
504 "grad_norm": 0.29186299443244934,
505 "learning_rate": 2.398510242085661e-05,
506 "loss": 0.0141,
507 "step": 700
508 },
509 {
510 "epoch": 0.2169926650366748,
511 "grad_norm": 0.20607629418373108,
512 "learning_rate": 2.389199255121043e-05,
513 "loss": 0.1886,
514 "step": 710
515 },
516 {
517 "epoch": 0.2200488997555012,
518 "grad_norm": 0.40326663851737976,
519 "learning_rate": 2.379888268156425e-05,
520 "loss": 0.082,
521 "step": 720
522 },
523 {
524 "epoch": 0.22310513447432762,
525 "grad_norm": 0.18325626850128174,
526 "learning_rate": 2.3705772811918064e-05,
527 "loss": 0.0628,
528 "step": 730
529 },
530 {
531 "epoch": 0.22616136919315402,
532 "grad_norm": 0.1717202067375183,
533 "learning_rate": 2.361266294227188e-05,
534 "loss": 0.1665,
535 "step": 740
536 },
537 {
538 "epoch": 0.22921760391198043,
539 "grad_norm": 1.470982313156128,
540 "learning_rate": 2.3519553072625698e-05,
541 "loss": 0.1568,
542 "step": 750
543 },
544 {
545 "epoch": 0.23227383863080683,
546 "grad_norm": 0.1533995270729065,
547 "learning_rate": 2.3426443202979514e-05,
548 "loss": 0.0565,
549 "step": 760
550 },
551 {
552 "epoch": 0.23533007334963324,
553 "grad_norm": 0.18549972772598267,
554 "learning_rate": 2.3333333333333336e-05,
555 "loss": 0.1877,
556 "step": 770
557 },
558 {
559 "epoch": 0.23838630806845965,
560 "grad_norm": 1.3424246311187744,
561 "learning_rate": 2.324022346368715e-05,
562 "loss": 0.1371,
563 "step": 780
564 },
565 {
566 "epoch": 0.24144254278728605,
567 "grad_norm": 1.0893691778182983,
568 "learning_rate": 2.314711359404097e-05,
569 "loss": 0.1286,
570 "step": 790
571 },
572 {
573 "epoch": 0.24449877750611246,
574 "grad_norm": 0.309053897857666,
575 "learning_rate": 2.3054003724394785e-05,
576 "loss": 0.0637,
577 "step": 800
578 },
579 {
580 "epoch": 0.24755501222493886,
581 "grad_norm": 0.2743883728981018,
582 "learning_rate": 2.2960893854748604e-05,
583 "loss": 0.0151,
584 "step": 810
585 },
586 {
587 "epoch": 0.2506112469437653,
588 "grad_norm": 1.4966446161270142,
589 "learning_rate": 2.286778398510242e-05,
590 "loss": 0.1031,
591 "step": 820
592 },
593 {
594 "epoch": 0.2536674816625917,
595 "grad_norm": 0.22224830090999603,
596 "learning_rate": 2.2774674115456238e-05,
597 "loss": 0.0618,
598 "step": 830
599 },
600 {
601 "epoch": 0.2567237163814181,
602 "grad_norm": 0.17919765412807465,
603 "learning_rate": 2.2681564245810057e-05,
604 "loss": 0.0593,
605 "step": 840
606 },
607 {
608 "epoch": 0.2597799511002445,
609 "grad_norm": 0.09819658100605011,
610 "learning_rate": 2.2588454376163876e-05,
611 "loss": 0.0781,
612 "step": 850
613 },
614 {
615 "epoch": 0.2628361858190709,
616 "grad_norm": 0.22454911470413208,
617 "learning_rate": 2.249534450651769e-05,
618 "loss": 0.1045,
619 "step": 860
620 },
621 {
622 "epoch": 0.2658924205378973,
623 "grad_norm": 49.3597526550293,
624 "learning_rate": 2.240223463687151e-05,
625 "loss": 0.087,
626 "step": 870
627 },
628 {
629 "epoch": 0.26894865525672373,
630 "grad_norm": 1.6358880996704102,
631 "learning_rate": 2.2309124767225325e-05,
632 "loss": 0.2507,
633 "step": 880
634 },
635 {
636 "epoch": 0.2720048899755501,
637 "grad_norm": 0.15142078697681427,
638 "learning_rate": 2.2216014897579144e-05,
639 "loss": 0.1193,
640 "step": 890
641 },
642 {
643 "epoch": 0.27506112469437655,
644 "grad_norm": 0.1405174732208252,
645 "learning_rate": 2.2122905027932963e-05,
646 "loss": 0.0527,
647 "step": 900
648 },
649 {
650 "epoch": 0.2781173594132029,
651 "grad_norm": 0.1312502920627594,
652 "learning_rate": 2.2029795158286778e-05,
653 "loss": 0.1309,
654 "step": 910
655 },
656 {
657 "epoch": 0.28117359413202936,
658 "grad_norm": 0.11085129529237747,
659 "learning_rate": 2.1936685288640597e-05,
660 "loss": 0.0598,
661 "step": 920
662 },
663 {
664 "epoch": 0.28422982885085574,
665 "grad_norm": 0.13175728917121887,
666 "learning_rate": 2.1843575418994412e-05,
667 "loss": 0.0601,
668 "step": 930
669 },
670 {
671 "epoch": 0.28728606356968217,
672 "grad_norm": 0.13237424194812775,
673 "learning_rate": 2.175046554934823e-05,
674 "loss": 0.1338,
675 "step": 940
676 },
677 {
678 "epoch": 0.29034229828850855,
679 "grad_norm": 1.7460986375808716,
680 "learning_rate": 2.1657355679702047e-05,
681 "loss": 0.1575,
682 "step": 950
683 },
684 {
685 "epoch": 0.293398533007335,
686 "grad_norm": 0.1591300666332245,
687 "learning_rate": 2.156424581005587e-05,
688 "loss": 0.0535,
689 "step": 960
690 },
691 {
692 "epoch": 0.29645476772616136,
693 "grad_norm": 0.1477421522140503,
694 "learning_rate": 2.1471135940409684e-05,
695 "loss": 0.0701,
696 "step": 970
697 },
698 {
699 "epoch": 0.2995110024449878,
700 "grad_norm": 0.15832221508026123,
701 "learning_rate": 2.1378026070763503e-05,
702 "loss": 0.2401,
703 "step": 980
704 },
705 {
706 "epoch": 0.3025672371638142,
707 "grad_norm": 0.2130369395017624,
708 "learning_rate": 2.1284916201117318e-05,
709 "loss": 0.2158,
710 "step": 990
711 },
712 {
713 "epoch": 0.3056234718826406,
714 "grad_norm": 0.1542048305273056,
715 "learning_rate": 2.1191806331471137e-05,
716 "loss": 0.0971,
717 "step": 1000
718 },
719 {
720 "epoch": 0.308679706601467,
721 "grad_norm": 3.364166021347046,
722 "learning_rate": 2.1098696461824952e-05,
723 "loss": 0.1079,
724 "step": 1010
725 },
726 {
727 "epoch": 0.3117359413202934,
728 "grad_norm": 0.14105622470378876,
729 "learning_rate": 2.100558659217877e-05,
730 "loss": 0.0597,
731 "step": 1020
732 },
733 {
734 "epoch": 0.3147921760391198,
735 "grad_norm": 0.1316232830286026,
736 "learning_rate": 2.091247672253259e-05,
737 "loss": 0.1159,
738 "step": 1030
739 },
740 {
741 "epoch": 0.31784841075794623,
742 "grad_norm": 0.48321789503097534,
743 "learning_rate": 2.081936685288641e-05,
744 "loss": 0.078,
745 "step": 1040
746 },
747 {
748 "epoch": 0.3209046454767726,
749 "grad_norm": 0.1567317247390747,
750 "learning_rate": 2.0726256983240224e-05,
751 "loss": 0.204,
752 "step": 1050
753 },
754 {
755 "epoch": 0.32396088019559904,
756 "grad_norm": 6.398591995239258,
757 "learning_rate": 2.0633147113594043e-05,
758 "loss": 0.3119,
759 "step": 1060
760 },
761 {
762 "epoch": 0.3270171149144254,
763 "grad_norm": 0.20879434049129486,
764 "learning_rate": 2.0540037243947858e-05,
765 "loss": 0.1167,
766 "step": 1070
767 },
768 {
769 "epoch": 0.33007334963325186,
770 "grad_norm": 0.23976916074752808,
771 "learning_rate": 2.0446927374301674e-05,
772 "loss": 0.1077,
773 "step": 1080
774 },
775 {
776 "epoch": 0.33312958435207823,
777 "grad_norm": 0.17524395883083344,
778 "learning_rate": 2.0353817504655496e-05,
779 "loss": 0.0814,
780 "step": 1090
781 },
782 {
783 "epoch": 0.33618581907090467,
784 "grad_norm": 0.36440449953079224,
785 "learning_rate": 2.026070763500931e-05,
786 "loss": 0.0589,
787 "step": 1100
788 },
789 {
790 "epoch": 0.33924205378973105,
791 "grad_norm": 0.15378913283348083,
792 "learning_rate": 2.016759776536313e-05,
793 "loss": 0.0112,
794 "step": 1110
795 },
796 {
797 "epoch": 0.3422982885085575,
798 "grad_norm": 0.14528077840805054,
799 "learning_rate": 2.0074487895716945e-05,
800 "loss": 0.1187,
801 "step": 1120
802 },
803 {
804 "epoch": 0.34535452322738386,
805 "grad_norm": 0.12367907911539078,
806 "learning_rate": 1.9990689013035385e-05,
807 "loss": 0.0722,
808 "step": 1130
809 },
810 {
811 "epoch": 0.3484107579462103,
812 "grad_norm": 0.09661256521940231,
813 "learning_rate": 1.98975791433892e-05,
814 "loss": 0.0052,
815 "step": 1140
816 },
817 {
818 "epoch": 0.35146699266503667,
819 "grad_norm": 0.7434007525444031,
820 "learning_rate": 1.980446927374302e-05,
821 "loss": 0.005,
822 "step": 1150
823 },
824 {
825 "epoch": 0.3545232273838631,
826 "grad_norm": 0.04170653969049454,
827 "learning_rate": 1.9711359404096834e-05,
828 "loss": 0.0036,
829 "step": 1160
830 },
831 {
832 "epoch": 0.3575794621026895,
833 "grad_norm": 0.07466982305049896,
834 "learning_rate": 1.961824953445065e-05,
835 "loss": 0.0684,
836 "step": 1170
837 },
838 {
839 "epoch": 0.3606356968215159,
840 "grad_norm": 0.05339043587446213,
841 "learning_rate": 1.952513966480447e-05,
842 "loss": 0.0064,
843 "step": 1180
844 },
845 {
846 "epoch": 0.3636919315403423,
847 "grad_norm": 4.3722052574157715,
848 "learning_rate": 1.9432029795158287e-05,
849 "loss": 0.295,
850 "step": 1190
851 },
852 {
853 "epoch": 0.36674816625916873,
854 "grad_norm": 0.04944440722465515,
855 "learning_rate": 1.9338919925512106e-05,
856 "loss": 0.0644,
857 "step": 1200
858 },
859 {
860 "epoch": 0.3698044009779951,
861 "grad_norm": 0.09254731982946396,
862 "learning_rate": 1.924581005586592e-05,
863 "loss": 0.2032,
864 "step": 1210
865 },
866 {
867 "epoch": 0.37286063569682154,
868 "grad_norm": 2.1284565925598145,
869 "learning_rate": 1.915270018621974e-05,
870 "loss": 0.1224,
871 "step": 1220
872 },
873 {
874 "epoch": 0.3759168704156479,
875 "grad_norm": 0.1618586629629135,
876 "learning_rate": 1.9059590316573555e-05,
877 "loss": 0.0694,
878 "step": 1230
879 },
880 {
881 "epoch": 0.37897310513447435,
882 "grad_norm": 0.20215436816215515,
883 "learning_rate": 1.8966480446927374e-05,
884 "loss": 0.0642,
885 "step": 1240
886 },
887 {
888 "epoch": 0.38202933985330073,
889 "grad_norm": 0.11647956073284149,
890 "learning_rate": 1.8873370577281193e-05,
891 "loss": 0.0268,
892 "step": 1250
893 },
894 {
895 "epoch": 0.38508557457212717,
896 "grad_norm": 2.9735450744628906,
897 "learning_rate": 1.878026070763501e-05,
898 "loss": 0.1907,
899 "step": 1260
900 },
901 {
902 "epoch": 0.38814180929095354,
903 "grad_norm": 0.16742634773254395,
904 "learning_rate": 1.8687150837988827e-05,
905 "loss": 0.068,
906 "step": 1270
907 },
908 {
909 "epoch": 0.39119804400978,
910 "grad_norm": 0.1201523020863533,
911 "learning_rate": 1.8594040968342646e-05,
912 "loss": 0.0627,
913 "step": 1280
914 },
915 {
916 "epoch": 0.39425427872860636,
917 "grad_norm": 0.07464749366044998,
918 "learning_rate": 1.850093109869646e-05,
919 "loss": 0.0132,
920 "step": 1290
921 },
922 {
923 "epoch": 0.3973105134474328,
924 "grad_norm": 0.12164530158042908,
925 "learning_rate": 1.840782122905028e-05,
926 "loss": 0.2341,
927 "step": 1300
928 },
929 {
930 "epoch": 0.40036674816625917,
931 "grad_norm": 0.09393124282360077,
932 "learning_rate": 1.83147113594041e-05,
933 "loss": 0.061,
934 "step": 1310
935 },
936 {
937 "epoch": 0.4034229828850856,
938 "grad_norm": 8.743972778320312,
939 "learning_rate": 1.8221601489757917e-05,
940 "loss": 0.1509,
941 "step": 1320
942 },
943 {
944 "epoch": 0.406479217603912,
945 "grad_norm": 0.22810806334018707,
946 "learning_rate": 1.8128491620111733e-05,
947 "loss": 0.0187,
948 "step": 1330
949 },
950 {
951 "epoch": 0.4095354523227384,
952 "grad_norm": 0.15852580964565277,
953 "learning_rate": 1.8035381750465548e-05,
954 "loss": 0.051,
955 "step": 1340
956 },
957 {
958 "epoch": 0.4125916870415648,
959 "grad_norm": 2.218247652053833,
960 "learning_rate": 1.7942271880819367e-05,
961 "loss": 0.151,
962 "step": 1350
963 },
964 {
965 "epoch": 0.4156479217603912,
966 "grad_norm": 0.0791810154914856,
967 "learning_rate": 1.7849162011173182e-05,
968 "loss": 0.352,
969 "step": 1360
970 },
971 {
972 "epoch": 0.4187041564792176,
973 "grad_norm": 0.2023414820432663,
974 "learning_rate": 1.7756052141527e-05,
975 "loss": 0.2002,
976 "step": 1370
977 },
978 {
979 "epoch": 0.421760391198044,
980 "grad_norm": 4.077162265777588,
981 "learning_rate": 1.766294227188082e-05,
982 "loss": 0.0924,
983 "step": 1380
984 },
985 {
986 "epoch": 0.4248166259168704,
987 "grad_norm": 0.16036756336688995,
988 "learning_rate": 1.756983240223464e-05,
989 "loss": 0.1015,
990 "step": 1390
991 },
992 {
993 "epoch": 0.4278728606356968,
994 "grad_norm": 12.642723083496094,
995 "learning_rate": 1.7476722532588454e-05,
996 "loss": 0.0885,
997 "step": 1400
998 },
999 {
1000 "epoch": 0.43092909535452323,
1001 "grad_norm": 0.20661431550979614,
1002 "learning_rate": 1.7383612662942273e-05,
1003 "loss": 0.1037,
1004 "step": 1410
1005 },
1006 {
1007 "epoch": 0.4339853300733496,
1008 "grad_norm": 0.14822793006896973,
1009 "learning_rate": 1.7290502793296088e-05,
1010 "loss": 0.0837,
1011 "step": 1420
1012 },
1013 {
1014 "epoch": 0.43704156479217604,
1015 "grad_norm": 0.12976513803005219,
1016 "learning_rate": 1.7197392923649907e-05,
1017 "loss": 0.2077,
1018 "step": 1430
1019 },
1020 {
1021 "epoch": 0.4400977995110024,
1022 "grad_norm": 0.045029882341623306,
1023 "learning_rate": 1.7104283054003726e-05,
1024 "loss": 0.0824,
1025 "step": 1440
1026 },
1027 {
1028 "epoch": 0.44315403422982885,
1029 "grad_norm": 0.19809339940547943,
1030 "learning_rate": 1.7011173184357544e-05,
1031 "loss": 0.1054,
1032 "step": 1450
1033 },
1034 {
1035 "epoch": 0.44621026894865523,
1036 "grad_norm": 0.2574075162410736,
1037 "learning_rate": 1.691806331471136e-05,
1038 "loss": 0.0878,
1039 "step": 1460
1040 },
1041 {
1042 "epoch": 0.44926650366748166,
1043 "grad_norm": 1.8627293109893799,
1044 "learning_rate": 1.682495344506518e-05,
1045 "loss": 0.1102,
1046 "step": 1470
1047 },
1048 {
1049 "epoch": 0.45232273838630804,
1050 "grad_norm": 0.1813037395477295,
1051 "learning_rate": 1.6731843575418994e-05,
1052 "loss": 0.1346,
1053 "step": 1480
1054 },
1055 {
1056 "epoch": 0.4553789731051345,
1057 "grad_norm": 0.21230725944042206,
1058 "learning_rate": 1.6638733705772813e-05,
1059 "loss": 0.0377,
1060 "step": 1490
1061 },
1062 {
1063 "epoch": 0.45843520782396086,
1064 "grad_norm": 0.17976613342761993,
1065 "learning_rate": 1.654562383612663e-05,
1066 "loss": 0.1334,
1067 "step": 1500
1068 },
1069 {
1070 "epoch": 0.4614914425427873,
1071 "grad_norm": 0.13542696833610535,
1072 "learning_rate": 1.6452513966480447e-05,
1073 "loss": 0.0935,
1074 "step": 1510
1075 },
1076 {
1077 "epoch": 0.46454767726161367,
1078 "grad_norm": 0.1337755024433136,
1079 "learning_rate": 1.6359404096834266e-05,
1080 "loss": 0.0073,
1081 "step": 1520
1082 },
1083 {
1084 "epoch": 0.4676039119804401,
1085 "grad_norm": 0.052432287484407425,
1086 "learning_rate": 1.626629422718808e-05,
1087 "loss": 0.1279,
1088 "step": 1530
1089 },
1090 {
1091 "epoch": 0.4706601466992665,
1092 "grad_norm": 0.06683123856782913,
1093 "learning_rate": 1.61731843575419e-05,
1094 "loss": 0.0042,
1095 "step": 1540
1096 },
1097 {
1098 "epoch": 0.4737163814180929,
1099 "grad_norm": 0.09111201763153076,
1100 "learning_rate": 1.6080074487895715e-05,
1101 "loss": 0.0046,
1102 "step": 1550
1103 },
1104 {
1105 "epoch": 0.4767726161369193,
1106 "grad_norm": 0.053649771958589554,
1107 "learning_rate": 1.5986964618249534e-05,
1108 "loss": 0.1305,
1109 "step": 1560
1110 },
1111 {
1112 "epoch": 0.4798288508557457,
1113 "grad_norm": 0.10080774873495102,
1114 "learning_rate": 1.5893854748603353e-05,
1115 "loss": 0.1217,
1116 "step": 1570
1117 },
1118 {
1119 "epoch": 0.4828850855745721,
1120 "grad_norm": 34.571510314941406,
1121 "learning_rate": 1.580074487895717e-05,
1122 "loss": 0.036,
1123 "step": 1580
1124 },
1125 {
1126 "epoch": 0.48594132029339854,
1127 "grad_norm": 0.07344861328601837,
1128 "learning_rate": 1.5707635009310987e-05,
1129 "loss": 0.1111,
1130 "step": 1590
1131 },
1132 {
1133 "epoch": 0.4889975550122249,
1134 "grad_norm": 4.016622543334961,
1135 "learning_rate": 1.5614525139664806e-05,
1136 "loss": 0.1141,
1137 "step": 1600
1138 },
1139 {
1140 "epoch": 0.49205378973105135,
1141 "grad_norm": 0.12781716883182526,
1142 "learning_rate": 1.552141527001862e-05,
1143 "loss": 0.1436,
1144 "step": 1610
1145 },
1146 {
1147 "epoch": 0.49511002444987773,
1148 "grad_norm": 0.28127339482307434,
1149 "learning_rate": 1.542830540037244e-05,
1150 "loss": 0.0141,
1151 "step": 1620
1152 },
1153 {
1154 "epoch": 0.49816625916870416,
1155 "grad_norm": 110.2169418334961,
1156 "learning_rate": 1.533519553072626e-05,
1157 "loss": 0.1011,
1158 "step": 1630
1159 },
1160 {
1161 "epoch": 0.5012224938875306,
1162 "grad_norm": 0.11259926855564117,
1163 "learning_rate": 1.5242085661080076e-05,
1164 "loss": 0.1317,
1165 "step": 1640
1166 },
1167 {
1168 "epoch": 0.5042787286063569,
1169 "grad_norm": 54.546661376953125,
1170 "learning_rate": 1.5148975791433893e-05,
1171 "loss": 0.1059,
1172 "step": 1650
1173 },
1174 {
1175 "epoch": 0.5073349633251834,
1176 "grad_norm": 0.039869241416454315,
1177 "learning_rate": 1.5055865921787711e-05,
1178 "loss": 0.0576,
1179 "step": 1660
1180 },
1181 {
1182 "epoch": 0.5103911980440098,
1183 "grad_norm": 0.09830798953771591,
1184 "learning_rate": 1.4962756052141527e-05,
1185 "loss": 0.0096,
1186 "step": 1670
1187 },
1188 {
1189 "epoch": 0.5134474327628362,
1190 "grad_norm": 0.07597620040178299,
1191 "learning_rate": 1.4869646182495346e-05,
1192 "loss": 0.051,
1193 "step": 1680
1194 },
1195 {
1196 "epoch": 0.5165036674816625,
1197 "grad_norm": 0.05168398097157478,
1198 "learning_rate": 1.4776536312849163e-05,
1199 "loss": 0.1166,
1200 "step": 1690
1201 },
1202 {
1203 "epoch": 0.519559902200489,
1204 "grad_norm": 93.93925476074219,
1205 "learning_rate": 1.468342644320298e-05,
1206 "loss": 0.0429,
1207 "step": 1700
1208 },
1209 {
1210 "epoch": 0.5226161369193154,
1211 "grad_norm": 0.9389681220054626,
1212 "learning_rate": 1.4590316573556799e-05,
1213 "loss": 0.0065,
1214 "step": 1710
1215 },
1216 {
1217 "epoch": 0.5256723716381418,
1218 "grad_norm": 0.09277983009815216,
1219 "learning_rate": 1.4497206703910616e-05,
1220 "loss": 0.0635,
1221 "step": 1720
1222 },
1223 {
1224 "epoch": 0.5287286063569682,
1225 "grad_norm": 0.09135166555643082,
1226 "learning_rate": 1.4404096834264433e-05,
1227 "loss": 0.1598,
1228 "step": 1730
1229 },
1230 {
1231 "epoch": 0.5317848410757946,
1232 "grad_norm": 0.07920227199792862,
1233 "learning_rate": 1.431098696461825e-05,
1234 "loss": 0.0586,
1235 "step": 1740
1236 },
1237 {
1238 "epoch": 0.534841075794621,
1239 "grad_norm": 0.0759720727801323,
1240 "learning_rate": 1.4217877094972069e-05,
1241 "loss": 0.1053,
1242 "step": 1750
1243 },
1244 {
1245 "epoch": 0.5378973105134475,
1246 "grad_norm": 0.07295390963554382,
1247 "learning_rate": 1.4124767225325884e-05,
1248 "loss": 0.0636,
1249 "step": 1760
1250 },
1251 {
1252 "epoch": 0.5409535452322738,
1253 "grad_norm": 0.1437848061323166,
1254 "learning_rate": 1.4031657355679701e-05,
1255 "loss": 0.1748,
1256 "step": 1770
1257 },
1258 {
1259 "epoch": 0.5440097799511002,
1260 "grad_norm": 0.0709507092833519,
1261 "learning_rate": 1.393854748603352e-05,
1262 "loss": 0.0113,
1263 "step": 1780
1264 },
1265 {
1266 "epoch": 0.5470660146699267,
1267 "grad_norm": 0.06068747863173485,
1268 "learning_rate": 1.3845437616387337e-05,
1269 "loss": 0.1883,
1270 "step": 1790
1271 },
1272 {
1273 "epoch": 0.5501222493887531,
1274 "grad_norm": 0.2686610221862793,
1275 "learning_rate": 1.3752327746741154e-05,
1276 "loss": 0.1691,
1277 "step": 1800
1278 },
1279 {
1280 "epoch": 0.5531784841075794,
1281 "grad_norm": 0.09411593526601791,
1282 "learning_rate": 1.3659217877094973e-05,
1283 "loss": 0.0983,
1284 "step": 1810
1285 },
1286 {
1287 "epoch": 0.5562347188264058,
1288 "grad_norm": 0.22031310200691223,
1289 "learning_rate": 1.356610800744879e-05,
1290 "loss": 0.0591,
1291 "step": 1820
1292 },
1293 {
1294 "epoch": 0.5592909535452323,
1295 "grad_norm": 0.09072048962116241,
1296 "learning_rate": 1.3472998137802607e-05,
1297 "loss": 0.0111,
1298 "step": 1830
1299 },
1300 {
1301 "epoch": 0.5623471882640587,
1302 "grad_norm": 0.10641049593687057,
1303 "learning_rate": 1.3379888268156426e-05,
1304 "loss": 0.1188,
1305 "step": 1840
1306 },
1307 {
1308 "epoch": 0.565403422982885,
1309 "grad_norm": 0.07925461232662201,
1310 "learning_rate": 1.3286778398510243e-05,
1311 "loss": 0.0603,
1312 "step": 1850
1313 },
1314 {
1315 "epoch": 0.5684596577017115,
1316 "grad_norm": 0.16015411913394928,
1317 "learning_rate": 1.319366852886406e-05,
1318 "loss": 0.062,
1319 "step": 1860
1320 },
1321 {
1322 "epoch": 0.5715158924205379,
1323 "grad_norm": 1.8697471618652344,
1324 "learning_rate": 1.3100558659217879e-05,
1325 "loss": 0.1608,
1326 "step": 1870
1327 },
1328 {
1329 "epoch": 0.5745721271393643,
1330 "grad_norm": 0.1413722038269043,
1331 "learning_rate": 1.3007448789571696e-05,
1332 "loss": 0.061,
1333 "step": 1880
1334 },
1335 {
1336 "epoch": 0.5776283618581907,
1337 "grad_norm": 0.1544978767633438,
1338 "learning_rate": 1.2914338919925513e-05,
1339 "loss": 0.0677,
1340 "step": 1890
1341 },
1342 {
1343 "epoch": 0.5806845965770171,
1344 "grad_norm": 0.1221311092376709,
1345 "learning_rate": 1.282122905027933e-05,
1346 "loss": 0.1088,
1347 "step": 1900
1348 },
1349 {
1350 "epoch": 0.5837408312958435,
1351 "grad_norm": 1.616721272468567,
1352 "learning_rate": 1.2728119180633149e-05,
1353 "loss": 0.1209,
1354 "step": 1910
1355 },
1356 {
1357 "epoch": 0.58679706601467,
1358 "grad_norm": 6.469699382781982,
1359 "learning_rate": 1.2635009310986966e-05,
1360 "loss": 0.1047,
1361 "step": 1920
1362 },
1363 {
1364 "epoch": 0.5898533007334963,
1365 "grad_norm": 0.13257375359535217,
1366 "learning_rate": 1.2541899441340781e-05,
1367 "loss": 0.1008,
1368 "step": 1930
1369 },
1370 {
1371 "epoch": 0.5929095354523227,
1372 "grad_norm": 5.230037689208984,
1373 "learning_rate": 1.24487895716946e-05,
1374 "loss": 0.1843,
1375 "step": 1940
1376 },
1377 {
1378 "epoch": 0.5959657701711492,
1379 "grad_norm": 5.640364170074463,
1380 "learning_rate": 1.2355679702048417e-05,
1381 "loss": 0.1216,
1382 "step": 1950
1383 },
1384 {
1385 "epoch": 0.5990220048899756,
1386 "grad_norm": 0.05993470177054405,
1387 "learning_rate": 1.2262569832402234e-05,
1388 "loss": 0.0991,
1389 "step": 1960
1390 },
1391 {
1392 "epoch": 0.6020782396088019,
1393 "grad_norm": 0.12238603085279465,
1394 "learning_rate": 1.2169459962756053e-05,
1395 "loss": 0.0109,
1396 "step": 1970
1397 },
1398 {
1399 "epoch": 0.6051344743276283,
1400 "grad_norm": 0.16531647741794586,
1401 "learning_rate": 1.207635009310987e-05,
1402 "loss": 0.0523,
1403 "step": 1980
1404 },
1405 {
1406 "epoch": 0.6081907090464548,
1407 "grad_norm": 0.07218731194734573,
1408 "learning_rate": 1.1983240223463687e-05,
1409 "loss": 0.0054,
1410 "step": 1990
1411 },
1412 {
1413 "epoch": 0.6112469437652812,
1414 "grad_norm": 0.04113167151808739,
1415 "learning_rate": 1.1890130353817506e-05,
1416 "loss": 0.0674,
1417 "step": 2000
1418 },
1419 {
1420 "epoch": 0.6143031784841075,
1421 "grad_norm": 13.622767448425293,
1422 "learning_rate": 1.1797020484171323e-05,
1423 "loss": 0.1255,
1424 "step": 2010
1425 },
1426 {
1427 "epoch": 0.617359413202934,
1428 "grad_norm": 0.06221568211913109,
1429 "learning_rate": 1.170391061452514e-05,
1430 "loss": 0.0043,
1431 "step": 2020
1432 },
1433 {
1434 "epoch": 0.6204156479217604,
1435 "grad_norm": 0.04879147931933403,
1436 "learning_rate": 1.1610800744878958e-05,
1437 "loss": 0.0603,
1438 "step": 2030
1439 },
1440 {
1441 "epoch": 0.6234718826405868,
1442 "grad_norm": 0.09293865412473679,
1443 "learning_rate": 1.1517690875232776e-05,
1444 "loss": 0.1467,
1445 "step": 2040
1446 },
1447 {
1448 "epoch": 0.6265281173594132,
1449 "grad_norm": 0.10396531224250793,
1450 "learning_rate": 1.1424581005586593e-05,
1451 "loss": 0.0688,
1452 "step": 2050
1453 },
1454 {
1455 "epoch": 0.6295843520782396,
1456 "grad_norm": 0.2048744261264801,
1457 "learning_rate": 1.133147113594041e-05,
1458 "loss": 0.0627,
1459 "step": 2060
1460 },
1461 {
1462 "epoch": 0.632640586797066,
1463 "grad_norm": 0.05055548995733261,
1464 "learning_rate": 1.1238361266294228e-05,
1465 "loss": 0.0662,
1466 "step": 2070
1467 },
1468 {
1469 "epoch": 0.6356968215158925,
1470 "grad_norm": 2.0886988639831543,
1471 "learning_rate": 1.1145251396648046e-05,
1472 "loss": 0.1094,
1473 "step": 2080
1474 },
1475 {
1476 "epoch": 0.6387530562347188,
1477 "grad_norm": 0.11593261361122131,
1478 "learning_rate": 1.1052141527001861e-05,
1479 "loss": 0.3611,
1480 "step": 2090
1481 },
1482 {
1483 "epoch": 0.6418092909535452,
1484 "grad_norm": 0.07993444800376892,
1485 "learning_rate": 1.095903165735568e-05,
1486 "loss": 0.0055,
1487 "step": 2100
1488 },
1489 {
1490 "epoch": 0.6448655256723717,
1491 "grad_norm": 0.13663113117218018,
1492 "learning_rate": 1.0865921787709497e-05,
1493 "loss": 0.0692,
1494 "step": 2110
1495 },
1496 {
1497 "epoch": 0.6479217603911981,
1498 "grad_norm": 0.09905444830656052,
1499 "learning_rate": 1.0772811918063314e-05,
1500 "loss": 0.008,
1501 "step": 2120
1502 },
1503 {
1504 "epoch": 0.6509779951100244,
1505 "grad_norm": 0.06681796908378601,
1506 "learning_rate": 1.0679702048417133e-05,
1507 "loss": 0.007,
1508 "step": 2130
1509 },
1510 {
1511 "epoch": 0.6540342298288508,
1512 "grad_norm": 0.08089947700500488,
1513 "learning_rate": 1.058659217877095e-05,
1514 "loss": 0.0875,
1515 "step": 2140
1516 },
1517 {
1518 "epoch": 0.6570904645476773,
1519 "grad_norm": 0.11339417099952698,
1520 "learning_rate": 1.0493482309124767e-05,
1521 "loss": 0.1986,
1522 "step": 2150
1523 },
1524 {
1525 "epoch": 0.6601466992665037,
1526 "grad_norm": 0.05615869536995888,
1527 "learning_rate": 1.0400372439478586e-05,
1528 "loss": 0.1701,
1529 "step": 2160
1530 },
1531 {
1532 "epoch": 0.66320293398533,
1533 "grad_norm": 0.09708325564861298,
1534 "learning_rate": 1.0307262569832403e-05,
1535 "loss": 0.059,
1536 "step": 2170
1537 },
1538 {
1539 "epoch": 0.6662591687041565,
1540 "grad_norm": 0.08505033701658249,
1541 "learning_rate": 1.021415270018622e-05,
1542 "loss": 0.2405,
1543 "step": 2180
1544 },
1545 {
1546 "epoch": 0.6693154034229829,
1547 "grad_norm": 0.12613414227962494,
1548 "learning_rate": 1.0121042830540038e-05,
1549 "loss": 0.0379,
1550 "step": 2190
1551 },
1552 {
1553 "epoch": 0.6723716381418093,
1554 "grad_norm": 0.11440026015043259,
1555 "learning_rate": 1.0027932960893856e-05,
1556 "loss": 0.0096,
1557 "step": 2200
1558 },
1559 {
1560 "epoch": 0.6754278728606357,
1561 "grad_norm": 1.7247629165649414,
1562 "learning_rate": 9.934823091247673e-06,
1563 "loss": 0.2441,
1564 "step": 2210
1565 },
1566 {
1567 "epoch": 0.6784841075794621,
1568 "grad_norm": 0.2626115679740906,
1569 "learning_rate": 9.84171322160149e-06,
1570 "loss": 0.1266,
1571 "step": 2220
1572 },
1573 {
1574 "epoch": 0.6815403422982885,
1575 "grad_norm": 0.08933160454034805,
1576 "learning_rate": 9.748603351955308e-06,
1577 "loss": 0.0932,
1578 "step": 2230
1579 },
1580 {
1581 "epoch": 0.684596577017115,
1582 "grad_norm": 0.18056859076023102,
1583 "learning_rate": 9.655493482309126e-06,
1584 "loss": 0.0528,
1585 "step": 2240
1586 },
1587 {
1588 "epoch": 0.6876528117359413,
1589 "grad_norm": 0.719739556312561,
1590 "learning_rate": 9.562383612662943e-06,
1591 "loss": 0.0848,
1592 "step": 2250
1593 },
1594 {
1595 "epoch": 0.6907090464547677,
1596 "grad_norm": 0.11645276844501495,
1597 "learning_rate": 9.46927374301676e-06,
1598 "loss": 0.0073,
1599 "step": 2260
1600 },
1601 {
1602 "epoch": 0.6937652811735942,
1603 "grad_norm": 0.05040796473622322,
1604 "learning_rate": 9.376163873370577e-06,
1605 "loss": 0.1393,
1606 "step": 2270
1607 },
1608 {
1609 "epoch": 0.6968215158924206,
1610 "grad_norm": 17.505489349365234,
1611 "learning_rate": 9.283054003724394e-06,
1612 "loss": 0.0733,
1613 "step": 2280
1614 },
1615 {
1616 "epoch": 0.6998777506112469,
1617 "grad_norm": 1.6304818391799927,
1618 "learning_rate": 9.189944134078213e-06,
1619 "loss": 0.0513,
1620 "step": 2290
1621 },
1622 {
1623 "epoch": 0.7029339853300733,
1624 "grad_norm": 0.1561206579208374,
1625 "learning_rate": 9.09683426443203e-06,
1626 "loss": 0.0575,
1627 "step": 2300
1628 },
1629 {
1630 "epoch": 0.7059902200488998,
1631 "grad_norm": 0.07211354374885559,
1632 "learning_rate": 9.003724394785847e-06,
1633 "loss": 0.1541,
1634 "step": 2310
1635 },
1636 {
1637 "epoch": 0.7090464547677262,
1638 "grad_norm": 0.12376312911510468,
1639 "learning_rate": 8.910614525139666e-06,
1640 "loss": 0.0061,
1641 "step": 2320
1642 },
1643 {
1644 "epoch": 0.7121026894865525,
1645 "grad_norm": 0.3487497568130493,
1646 "learning_rate": 8.817504655493483e-06,
1647 "loss": 0.0553,
1648 "step": 2330
1649 },
1650 {
1651 "epoch": 0.715158924205379,
1652 "grad_norm": 0.10781729221343994,
1653 "learning_rate": 8.7243947858473e-06,
1654 "loss": 0.0069,
1655 "step": 2340
1656 },
1657 {
1658 "epoch": 0.7182151589242054,
1659 "grad_norm": 5.861069679260254,
1660 "learning_rate": 8.631284916201118e-06,
1661 "loss": 0.1086,
1662 "step": 2350
1663 },
1664 {
1665 "epoch": 0.7212713936430318,
1666 "grad_norm": 0.2850921154022217,
1667 "learning_rate": 8.538175046554935e-06,
1668 "loss": 0.0912,
1669 "step": 2360
1670 },
1671 {
1672 "epoch": 0.7243276283618582,
1673 "grad_norm": 3.2116122245788574,
1674 "learning_rate": 8.445065176908753e-06,
1675 "loss": 0.1441,
1676 "step": 2370
1677 },
1678 {
1679 "epoch": 0.7273838630806846,
1680 "grad_norm": 0.4662390947341919,
1681 "learning_rate": 8.35195530726257e-06,
1682 "loss": 0.1017,
1683 "step": 2380
1684 },
1685 {
1686 "epoch": 0.730440097799511,
1687 "grad_norm": 0.05119713768362999,
1688 "learning_rate": 8.258845437616388e-06,
1689 "loss": 0.0102,
1690 "step": 2390
1691 },
1692 {
1693 "epoch": 0.7334963325183375,
1694 "grad_norm": 0.0849548801779747,
1695 "learning_rate": 8.165735567970205e-06,
1696 "loss": 0.0061,
1697 "step": 2400
1698 },
1699 {
1700 "epoch": 0.7365525672371638,
1701 "grad_norm": 0.035956285893917084,
1702 "learning_rate": 8.072625698324023e-06,
1703 "loss": 0.0035,
1704 "step": 2410
1705 },
1706 {
1707 "epoch": 0.7396088019559902,
1708 "grad_norm": 0.06464776396751404,
1709 "learning_rate": 7.979515828677841e-06,
1710 "loss": 0.0466,
1711 "step": 2420
1712 },
1713 {
1714 "epoch": 0.7426650366748166,
1715 "grad_norm": 0.5524429082870483,
1716 "learning_rate": 7.886405959031657e-06,
1717 "loss": 0.0704,
1718 "step": 2430
1719 },
1720 {
1721 "epoch": 0.7457212713936431,
1722 "grad_norm": 0.04895765706896782,
1723 "learning_rate": 7.793296089385474e-06,
1724 "loss": 0.0036,
1725 "step": 2440
1726 },
1727 {
1728 "epoch": 0.7487775061124694,
1729 "grad_norm": 0.03954509273171425,
1730 "learning_rate": 7.700186219739293e-06,
1731 "loss": 0.0032,
1732 "step": 2450
1733 },
1734 {
1735 "epoch": 0.7518337408312958,
1736 "grad_norm": 13.51663589477539,
1737 "learning_rate": 7.60707635009311e-06,
1738 "loss": 0.063,
1739 "step": 2460
1740 },
1741 {
1742 "epoch": 0.7548899755501223,
1743 "grad_norm": 0.11551866680383682,
1744 "learning_rate": 7.5139664804469275e-06,
1745 "loss": 0.0028,
1746 "step": 2470
1747 },
1748 {
1749 "epoch": 0.7579462102689487,
1750 "grad_norm": 7.975831031799316,
1751 "learning_rate": 7.420856610800745e-06,
1752 "loss": 0.2147,
1753 "step": 2480
1754 },
1755 {
1756 "epoch": 0.761002444987775,
1757 "grad_norm": 0.10641849786043167,
1758 "learning_rate": 7.3277467411545625e-06,
1759 "loss": 0.1884,
1760 "step": 2490
1761 },
1762 {
1763 "epoch": 0.7640586797066015,
1764 "grad_norm": 0.3172556757926941,
1765 "learning_rate": 7.23463687150838e-06,
1766 "loss": 0.0692,
1767 "step": 2500
1768 },
1769 {
1770 "epoch": 0.7671149144254279,
1771 "grad_norm": 0.12498392164707184,
1772 "learning_rate": 7.1415270018621975e-06,
1773 "loss": 0.1202,
1774 "step": 2510
1775 },
1776 {
1777 "epoch": 0.7701711491442543,
1778 "grad_norm": 0.1759582906961441,
1779 "learning_rate": 7.0484171322160155e-06,
1780 "loss": 0.0624,
1781 "step": 2520
1782 },
1783 {
1784 "epoch": 0.7732273838630807,
1785 "grad_norm": 0.09976211935281754,
1786 "learning_rate": 6.9553072625698325e-06,
1787 "loss": 0.1345,
1788 "step": 2530
1789 },
1790 {
1791 "epoch": 0.7762836185819071,
1792 "grad_norm": 0.10823414474725723,
1793 "learning_rate": 6.8621973929236505e-06,
1794 "loss": 0.0963,
1795 "step": 2540
1796 },
1797 {
1798 "epoch": 0.7793398533007335,
1799 "grad_norm": 0.09119368344545364,
1800 "learning_rate": 6.7690875232774675e-06,
1801 "loss": 0.0065,
1802 "step": 2550
1803 },
1804 {
1805 "epoch": 0.78239608801956,
1806 "grad_norm": 0.22418056428432465,
1807 "learning_rate": 6.675977653631285e-06,
1808 "loss": 0.1055,
1809 "step": 2560
1810 },
1811 {
1812 "epoch": 0.7854523227383863,
1813 "grad_norm": 0.16302266716957092,
1814 "learning_rate": 6.5828677839851025e-06,
1815 "loss": 0.1129,
1816 "step": 2570
1817 },
1818 {
1819 "epoch": 0.7885085574572127,
1820 "grad_norm": 0.12889882922172546,
1821 "learning_rate": 6.48975791433892e-06,
1822 "loss": 0.1578,
1823 "step": 2580
1824 },
1825 {
1826 "epoch": 0.7915647921760391,
1827 "grad_norm": 0.4161546528339386,
1828 "learning_rate": 6.3966480446927375e-06,
1829 "loss": 0.0546,
1830 "step": 2590
1831 },
1832 {
1833 "epoch": 0.7946210268948656,
1834 "grad_norm": 0.4652371108531952,
1835 "learning_rate": 6.3035381750465554e-06,
1836 "loss": 0.0736,
1837 "step": 2600
1838 },
1839 {
1840 "epoch": 0.7976772616136919,
1841 "grad_norm": 0.1676311194896698,
1842 "learning_rate": 6.2104283054003725e-06,
1843 "loss": 0.1377,
1844 "step": 2610
1845 },
1846 {
1847 "epoch": 0.8007334963325183,
1848 "grad_norm": 0.49725136160850525,
1849 "learning_rate": 6.1173184357541904e-06,
1850 "loss": 0.0478,
1851 "step": 2620
1852 },
1853 {
1854 "epoch": 0.8037897310513448,
1855 "grad_norm": 0.3372693359851837,
1856 "learning_rate": 6.0242085661080075e-06,
1857 "loss": 0.107,
1858 "step": 2630
1859 },
1860 {
1861 "epoch": 0.8068459657701712,
1862 "grad_norm": 11.673458099365234,
1863 "learning_rate": 5.9310986964618254e-06,
1864 "loss": 0.1794,
1865 "step": 2640
1866 },
1867 {
1868 "epoch": 0.8099022004889975,
1869 "grad_norm": 0.13454145193099976,
1870 "learning_rate": 5.8379888268156425e-06,
1871 "loss": 0.005,
1872 "step": 2650
1873 },
1874 {
1875 "epoch": 0.812958435207824,
1876 "grad_norm": 0.10115115344524384,
1877 "learning_rate": 5.74487895716946e-06,
1878 "loss": 0.0428,
1879 "step": 2660
1880 },
1881 {
1882 "epoch": 0.8160146699266504,
1883 "grad_norm": 0.1693265289068222,
1884 "learning_rate": 5.6517690875232775e-06,
1885 "loss": 0.0553,
1886 "step": 2670
1887 },
1888 {
1889 "epoch": 0.8190709046454768,
1890 "grad_norm": 0.040254298597574234,
1891 "learning_rate": 5.558659217877095e-06,
1892 "loss": 0.0041,
1893 "step": 2680
1894 },
1895 {
1896 "epoch": 0.8221271393643031,
1897 "grad_norm": 0.06211506202816963,
1898 "learning_rate": 5.4655493482309125e-06,
1899 "loss": 0.0301,
1900 "step": 2690
1901 },
1902 {
1903 "epoch": 0.8251833740831296,
1904 "grad_norm": 0.03791659697890282,
1905 "learning_rate": 5.37243947858473e-06,
1906 "loss": 0.1197,
1907 "step": 2700
1908 },
1909 {
1910 "epoch": 0.828239608801956,
1911 "grad_norm": 0.06369005143642426,
1912 "learning_rate": 5.2793296089385475e-06,
1913 "loss": 0.065,
1914 "step": 2710
1915 },
1916 {
1917 "epoch": 0.8312958435207825,
1918 "grad_norm": 0.15808682143688202,
1919 "learning_rate": 5.186219739292365e-06,
1920 "loss": 0.0554,
1921 "step": 2720
1922 },
1923 {
1924 "epoch": 0.8343520782396088,
1925 "grad_norm": 0.05038405582308769,
1926 "learning_rate": 5.0931098696461825e-06,
1927 "loss": 0.082,
1928 "step": 2730
1929 },
1930 {
1931 "epoch": 0.8374083129584352,
1932 "grad_norm": 0.26016944646835327,
1933 "learning_rate": 4.9999999999999996e-06,
1934 "loss": 0.0731,
1935 "step": 2740
1936 },
1937 {
1938 "epoch": 0.8404645476772616,
1939 "grad_norm": 0.03949306160211563,
1940 "learning_rate": 4.9068901303538175e-06,
1941 "loss": 0.2097,
1942 "step": 2750
1943 },
1944 {
1945 "epoch": 0.843520782396088,
1946 "grad_norm": 0.10841944068670273,
1947 "learning_rate": 4.813780260707635e-06,
1948 "loss": 0.1337,
1949 "step": 2760
1950 },
1951 {
1952 "epoch": 0.8465770171149144,
1953 "grad_norm": 0.08589453250169754,
1954 "learning_rate": 4.7206703910614525e-06,
1955 "loss": 0.1875,
1956 "step": 2770
1957 },
1958 {
1959 "epoch": 0.8496332518337408,
1960 "grad_norm": 7.176298141479492,
1961 "learning_rate": 4.62756052141527e-06,
1962 "loss": 0.1634,
1963 "step": 2780
1964 },
1965 {
1966 "epoch": 0.8526894865525673,
1967 "grad_norm": 0.20672304928302765,
1968 "learning_rate": 4.5344506517690875e-06,
1969 "loss": 0.0686,
1970 "step": 2790
1971 },
1972 {
1973 "epoch": 0.8557457212713936,
1974 "grad_norm": 0.03756504878401756,
1975 "learning_rate": 4.441340782122905e-06,
1976 "loss": 0.121,
1977 "step": 2800
1978 },
1979 {
1980 "epoch": 0.85880195599022,
1981 "grad_norm": 0.509375274181366,
1982 "learning_rate": 4.348230912476723e-06,
1983 "loss": 0.0217,
1984 "step": 2810
1985 },
1986 {
1987 "epoch": 0.8618581907090465,
1988 "grad_norm": 13.620018005371094,
1989 "learning_rate": 4.2551210428305395e-06,
1990 "loss": 0.119,
1991 "step": 2820
1992 },
1993 {
1994 "epoch": 0.8649144254278729,
1995 "grad_norm": 0.0516686849296093,
1996 "learning_rate": 4.1620111731843575e-06,
1997 "loss": 0.0654,
1998 "step": 2830
1999 },
2000 {
2001 "epoch": 0.8679706601466992,
2002 "grad_norm": 0.04489528760313988,
2003 "learning_rate": 4.068901303538175e-06,
2004 "loss": 0.0572,
2005 "step": 2840
2006 },
2007 {
2008 "epoch": 0.8710268948655256,
2009 "grad_norm": 0.06253698468208313,
2010 "learning_rate": 3.9757914338919925e-06,
2011 "loss": 0.0448,
2012 "step": 2850
2013 },
2014 {
2015 "epoch": 0.8740831295843521,
2016 "grad_norm": 0.06379766017198563,
2017 "learning_rate": 3.88268156424581e-06,
2018 "loss": 0.0062,
2019 "step": 2860
2020 },
2021 {
2022 "epoch": 0.8771393643031785,
2023 "grad_norm": 0.058733634650707245,
2024 "learning_rate": 3.789571694599628e-06,
2025 "loss": 0.0691,
2026 "step": 2870
2027 },
2028 {
2029 "epoch": 0.8801955990220048,
2030 "grad_norm": 0.06447366625070572,
2031 "learning_rate": 3.696461824953445e-06,
2032 "loss": 0.1691,
2033 "step": 2880
2034 },
2035 {
2036 "epoch": 0.8832518337408313,
2037 "grad_norm": 0.04590935632586479,
2038 "learning_rate": 3.6033519553072625e-06,
2039 "loss": 0.1227,
2040 "step": 2890
2041 },
2042 {
2043 "epoch": 0.8863080684596577,
2044 "grad_norm": 0.39994603395462036,
2045 "learning_rate": 3.5102420856610804e-06,
2046 "loss": 0.1183,
2047 "step": 2900
2048 },
2049 {
2050 "epoch": 0.8893643031784841,
2051 "grad_norm": 0.396999329328537,
2052 "learning_rate": 3.417132216014898e-06,
2053 "loss": 0.0406,
2054 "step": 2910
2055 },
2056 {
2057 "epoch": 0.8924205378973105,
2058 "grad_norm": 0.10507692396640778,
2059 "learning_rate": 3.324022346368715e-06,
2060 "loss": 0.053,
2061 "step": 2920
2062 },
2063 {
2064 "epoch": 0.8954767726161369,
2065 "grad_norm": 0.3681839406490326,
2066 "learning_rate": 3.2309124767225324e-06,
2067 "loss": 0.0048,
2068 "step": 2930
2069 },
2070 {
2071 "epoch": 0.8985330073349633,
2072 "grad_norm": 0.21031515300273895,
2073 "learning_rate": 3.1378026070763504e-06,
2074 "loss": 0.0053,
2075 "step": 2940
2076 },
2077 {
2078 "epoch": 0.9015892420537898,
2079 "grad_norm": 6.659632682800293,
2080 "learning_rate": 3.044692737430168e-06,
2081 "loss": 0.0991,
2082 "step": 2950
2083 },
2084 {
2085 "epoch": 0.9046454767726161,
2086 "grad_norm": 0.049913499504327774,
2087 "learning_rate": 2.951582867783985e-06,
2088 "loss": 0.0038,
2089 "step": 2960
2090 },
2091 {
2092 "epoch": 0.9077017114914425,
2093 "grad_norm": 0.09325915575027466,
2094 "learning_rate": 2.8584729981378024e-06,
2095 "loss": 0.0033,
2096 "step": 2970
2097 },
2098 {
2099 "epoch": 0.910757946210269,
2100 "grad_norm": 0.08547964692115784,
2101 "learning_rate": 2.7653631284916204e-06,
2102 "loss": 0.0684,
2103 "step": 2980
2104 },
2105 {
2106 "epoch": 0.9138141809290954,
2107 "grad_norm": 9.980097770690918,
2108 "learning_rate": 2.672253258845438e-06,
2109 "loss": 0.0735,
2110 "step": 2990
2111 },
2112 {
2113 "epoch": 0.9168704156479217,
2114 "grad_norm": 0.03152007982134819,
2115 "learning_rate": 2.5791433891992554e-06,
2116 "loss": 0.049,
2117 "step": 3000
2118 },
2119 {
2120 "epoch": 0.9199266503667481,
2121 "grad_norm": 0.039837196469306946,
2122 "learning_rate": 2.4860335195530724e-06,
2123 "loss": 0.0686,
2124 "step": 3010
2125 },
2126 {
2127 "epoch": 0.9229828850855746,
2128 "grad_norm": 12.22442626953125,
2129 "learning_rate": 2.3929236499068903e-06,
2130 "loss": 0.0554,
2131 "step": 3020
2132 },
2133 {
2134 "epoch": 0.926039119804401,
2135 "grad_norm": 0.13149628043174744,
2136 "learning_rate": 2.299813780260708e-06,
2137 "loss": 0.1026,
2138 "step": 3030
2139 },
2140 {
2141 "epoch": 0.9290953545232273,
2142 "grad_norm": 0.044785063713788986,
2143 "learning_rate": 2.2067039106145253e-06,
2144 "loss": 0.0042,
2145 "step": 3040
2146 },
2147 {
2148 "epoch": 0.9321515892420538,
2149 "grad_norm": 0.042241331189870834,
2150 "learning_rate": 2.1135940409683424e-06,
2151 "loss": 0.0681,
2152 "step": 3050
2153 },
2154 {
2155 "epoch": 0.9352078239608802,
2156 "grad_norm": 0.0555758960545063,
2157 "learning_rate": 2.0204841713221603e-06,
2158 "loss": 0.0282,
2159 "step": 3060
2160 },
2161 {
2162 "epoch": 0.9382640586797066,
2163 "grad_norm": 0.07724986225366592,
2164 "learning_rate": 1.927374301675978e-06,
2165 "loss": 0.1247,
2166 "step": 3070
2167 },
2168 {
2169 "epoch": 0.941320293398533,
2170 "grad_norm": 0.08575576543807983,
2171 "learning_rate": 1.8342644320297951e-06,
2172 "loss": 0.0689,
2173 "step": 3080
2174 },
2175 {
2176 "epoch": 0.9443765281173594,
2177 "grad_norm": 0.2837558686733246,
2178 "learning_rate": 1.7411545623836128e-06,
2179 "loss": 0.102,
2180 "step": 3090
2181 },
2182 {
2183 "epoch": 0.9474327628361858,
2184 "grad_norm": 0.1205146312713623,
2185 "learning_rate": 1.6480446927374301e-06,
2186 "loss": 0.1657,
2187 "step": 3100
2188 },
2189 {
2190 "epoch": 0.9504889975550123,
2191 "grad_norm": 0.16669517755508423,
2192 "learning_rate": 1.5549348230912478e-06,
2193 "loss": 0.0078,
2194 "step": 3110
2195 },
2196 {
2197 "epoch": 0.9535452322738386,
2198 "grad_norm": 15.247810363769531,
2199 "learning_rate": 1.4618249534450651e-06,
2200 "loss": 0.0175,
2201 "step": 3120
2202 },
2203 {
2204 "epoch": 0.956601466992665,
2205 "grad_norm": 0.20827865600585938,
2206 "learning_rate": 1.3687150837988828e-06,
2207 "loss": 0.0353,
2208 "step": 3130
2209 },
2210 {
2211 "epoch": 0.9596577017114915,
2212 "grad_norm": 0.07442985475063324,
2213 "learning_rate": 1.2756052141527001e-06,
2214 "loss": 0.0415,
2215 "step": 3140
2216 },
2217 {
2218 "epoch": 0.9627139364303179,
2219 "grad_norm": 0.040485650300979614,
2220 "learning_rate": 1.1824953445065178e-06,
2221 "loss": 0.1062,
2222 "step": 3150
2223 },
2224 {
2225 "epoch": 0.9657701711491442,
2226 "grad_norm": 0.12251528352499008,
2227 "learning_rate": 1.0893854748603353e-06,
2228 "loss": 0.043,
2229 "step": 3160
2230 },
2231 {
2232 "epoch": 0.9688264058679706,
2233 "grad_norm": 0.050180476158857346,
2234 "learning_rate": 9.962756052141528e-07,
2235 "loss": 0.03,
2236 "step": 3170
2237 },
2238 {
2239 "epoch": 0.9718826405867971,
2240 "grad_norm": 0.0506335012614727,
2241 "learning_rate": 9.031657355679702e-07,
2242 "loss": 0.2524,
2243 "step": 3180
2244 },
2245 {
2246 "epoch": 0.9749388753056235,
2247 "grad_norm": 28.770177841186523,
2248 "learning_rate": 8.100558659217877e-07,
2249 "loss": 0.0968,
2250 "step": 3190
2251 },
2252 {
2253 "epoch": 0.9779951100244498,
2254 "grad_norm": 0.03078501671552658,
2255 "learning_rate": 7.169459962756052e-07,
2256 "loss": 0.0758,
2257 "step": 3200
2258 },
2259 {
2260 "epoch": 0.9810513447432763,
2261 "grad_norm": 16.90084457397461,
2262 "learning_rate": 6.238361266294227e-07,
2263 "loss": 0.1547,
2264 "step": 3210
2265 },
2266 {
2267 "epoch": 0.9841075794621027,
2268 "grad_norm": 0.1555572897195816,
2269 "learning_rate": 5.307262569832402e-07,
2270 "loss": 0.0491,
2271 "step": 3220
2272 },
2273 {
2274 "epoch": 0.9871638141809291,
2275 "grad_norm": 0.04730997607111931,
2276 "learning_rate": 4.3761638733705775e-07,
2277 "loss": 0.0039,
2278 "step": 3230
2279 },
2280 {
2281 "epoch": 0.9902200488997555,
2282 "grad_norm": 3.162146806716919,
2283 "learning_rate": 3.4450651769087525e-07,
2284 "loss": 0.1387,
2285 "step": 3240
2286 },
2287 {
2288 "epoch": 0.9932762836185819,
2289 "grad_norm": 2.292222499847412,
2290 "learning_rate": 2.5139664804469275e-07,
2291 "loss": 0.1184,
2292 "step": 3250
2293 },
2294 {
2295 "epoch": 0.9963325183374083,
2296 "grad_norm": 0.054889462888240814,
2297 "learning_rate": 1.5828677839851024e-07,
2298 "loss": 0.0065,
2299 "step": 3260
2300 },
2301 {
2302 "epoch": 0.9993887530562348,
2303 "grad_norm": 0.310970664024353,
2304 "learning_rate": 6.517690875232774e-08,
2305 "loss": 0.0087,
2306 "step": 3270
2307 },
2308 {
2309 "epoch": 1.0,
2310 "eval_accuracy": 0.9845683728036669,
2311 "eval_f1": 0.9839899931917298,
2312 "eval_loss": 0.06383595615625381,
2313 "eval_model_preparation_time": 0.0032,
2314 "eval_runtime": 411.4096,
2315 "eval_samples_per_second": 15.909,
2316 "eval_steps_per_second": 1.991,
2317 "step": 3272
2318 }
2319 ],
2320 "logging_steps": 10,
2321 "max_steps": 3272,
2322 "num_input_tokens_seen": 0,
2323 "num_train_epochs": 1,
2324 "save_steps": 500,
2325 "stateful_callbacks": {
2326 "TrainerControl": {
2327 "args": {
2328 "should_epoch_stop": false,
2329 "should_evaluate": false,
2330 "should_log": false,
2331 "should_save": true,
2332 "should_training_stop": true
2333 },
2334 "attributes": {}
2335 }
2336 },
2337 "total_flos": 2.2958418080824934e+18,
2338 "train_batch_size": 8,
2339 "trial_name": null,
2340 "trial_params": null
2341 }
2342