terry69/preference_p0.1_seed42_level2_syntaxcleanbatch16
07
1{2 "best_metric": null,3 "best_model_checkpoint": null,4 "epoch": 1.0,5 "eval_steps": 500,6 "global_step": 2408,7 "is_hyper_param_search": false,8 "is_local_process_zero": true,9 "is_world_process_zero": true,10 "log_history": [11 {12 "epoch": 0.0004152823920265781,13 "grad_norm": 25.422981813437236,14 "learning_rate": 4.1493775933609963e-08,15 "loss": 1.3975,16 "step": 117 },18 {19 "epoch": 0.0020764119601328905,20 "grad_norm": 23.65282908395334,21 "learning_rate": 2.074688796680498e-07,22 "loss": 1.4281,23 "step": 524 },25 {26 "epoch": 0.004152823920265781,27 "grad_norm": 16.38973942245371,28 "learning_rate": 4.149377593360996e-07,29 "loss": 1.3933,30 "step": 1031 },32 {33 "epoch": 0.006229235880398671,34 "grad_norm": 8.620332321861904,35 "learning_rate": 6.224066390041494e-07,36 "loss": 1.2986,37 "step": 1538 },39 {40 "epoch": 0.008305647840531562,41 "grad_norm": 10.289897317705874,42 "learning_rate": 8.298755186721992e-07,43 "loss": 1.1565,44 "step": 2045 },46 {47 "epoch": 0.010382059800664452,48 "grad_norm": 4.429779856244459,49 "learning_rate": 1.037344398340249e-06,50 "loss": 1.051,51 "step": 2552 },53 {54 "epoch": 0.012458471760797342,55 "grad_norm": 3.3098208738585213,56 "learning_rate": 1.2448132780082988e-06,57 "loss": 0.9902,58 "step": 3059 },60 {61 "epoch": 0.014534883720930232,62 "grad_norm": 3.4349888460346687,63 "learning_rate": 1.4522821576763488e-06,64 "loss": 0.9652,65 "step": 3566 },67 {68 "epoch": 0.016611295681063124,69 "grad_norm": 3.1515624301454133,70 "learning_rate": 1.6597510373443984e-06,71 "loss": 0.9415,72 "step": 4073 },74 {75 "epoch": 0.018687707641196014,76 "grad_norm": 3.1235312209606505,77 "learning_rate": 1.8672199170124482e-06,78 "loss": 0.93,79 "step": 4580 },81 {82 "epoch": 0.020764119601328904,83 "grad_norm": 3.1741829648141926,84 "learning_rate": 2.074688796680498e-06,85 "loss": 0.9238,86 "step": 5087 },88 {89 "epoch": 0.022840531561461794,90 "grad_norm": 3.232116295196654,91 "learning_rate": 2.282157676348548e-06,92 "loss": 0.9123,93 "step": 5594 },95 {96 "epoch": 0.024916943521594685,97 "grad_norm": 3.1515595029223396,98 "learning_rate": 2.4896265560165977e-06,99 "loss": 0.9031,100 "step": 60101 },102 {103 "epoch": 0.026993355481727575,104 "grad_norm": 3.1003061370301617,105 "learning_rate": 2.6970954356846475e-06,106 "loss": 0.8947,107 "step": 65108 },109 {110 "epoch": 0.029069767441860465,111 "grad_norm": 2.9767060692194844,112 "learning_rate": 2.9045643153526977e-06,113 "loss": 0.8919,114 "step": 70115 },116 {117 "epoch": 0.031146179401993355,118 "grad_norm": 3.0759553041103205,119 "learning_rate": 3.112033195020747e-06,120 "loss": 0.8702,121 "step": 75122 },123 {124 "epoch": 0.03322259136212625,125 "grad_norm": 3.285827319776166,126 "learning_rate": 3.319502074688797e-06,127 "loss": 0.8727,128 "step": 80129 },130 {131 "epoch": 0.03529900332225914,132 "grad_norm": 3.3462993523967186,133 "learning_rate": 3.526970954356847e-06,134 "loss": 0.8736,135 "step": 85136 },137 {138 "epoch": 0.03737541528239203,139 "grad_norm": 3.256004424550593,140 "learning_rate": 3.7344398340248965e-06,141 "loss": 0.8858,142 "step": 90143 },144 {145 "epoch": 0.03945182724252492,146 "grad_norm": 3.159488005717498,147 "learning_rate": 3.941908713692946e-06,148 "loss": 0.8559,149 "step": 95150 },151 {152 "epoch": 0.04152823920265781,153 "grad_norm": 2.9634363451500114,154 "learning_rate": 4.149377593360996e-06,155 "loss": 0.8586,156 "step": 100157 },158 {159 "epoch": 0.0436046511627907,160 "grad_norm": 3.159728031235876,161 "learning_rate": 4.356846473029046e-06,162 "loss": 0.8674,163 "step": 105164 },165 {166 "epoch": 0.04568106312292359,167 "grad_norm": 2.970452415217835,168 "learning_rate": 4.564315352697096e-06,169 "loss": 0.8542,170 "step": 110171 },172 {173 "epoch": 0.04775747508305648,174 "grad_norm": 3.1788047641427513,175 "learning_rate": 4.771784232365146e-06,176 "loss": 0.8701,177 "step": 115178 },179 {180 "epoch": 0.04983388704318937,181 "grad_norm": 3.451301944053267,182 "learning_rate": 4.979253112033195e-06,183 "loss": 0.8714,184 "step": 120185 },186 {187 "epoch": 0.05191029900332226,188 "grad_norm": 3.388326009403783,189 "learning_rate": 5.1867219917012455e-06,190 "loss": 0.8483,191 "step": 125192 },193 {194 "epoch": 0.05398671096345515,195 "grad_norm": 2.9959540593135645,196 "learning_rate": 5.394190871369295e-06,197 "loss": 0.8481,198 "step": 130199 },200 {201 "epoch": 0.05606312292358804,202 "grad_norm": 3.1213953476841856,203 "learning_rate": 5.601659751037345e-06,204 "loss": 0.8387,205 "step": 135206 },207 {208 "epoch": 0.05813953488372093,209 "grad_norm": 3.403963416369247,210 "learning_rate": 5.809128630705395e-06,211 "loss": 0.8399,212 "step": 140213 },214 {215 "epoch": 0.06021594684385382,216 "grad_norm": 3.0893053330914695,217 "learning_rate": 6.016597510373444e-06,218 "loss": 0.8386,219 "step": 145220 },221 {222 "epoch": 0.06229235880398671,223 "grad_norm": 3.166492177328262,224 "learning_rate": 6.224066390041494e-06,225 "loss": 0.8457,226 "step": 150227 },228 {229 "epoch": 0.0643687707641196,230 "grad_norm": 3.2851556967703117,231 "learning_rate": 6.431535269709544e-06,232 "loss": 0.8421,233 "step": 155234 },235 {236 "epoch": 0.0664451827242525,237 "grad_norm": 2.9899317822541454,238 "learning_rate": 6.639004149377594e-06,239 "loss": 0.8373,240 "step": 160241 },242 {243 "epoch": 0.06852159468438539,244 "grad_norm": 3.0509892785590456,245 "learning_rate": 6.846473029045644e-06,246 "loss": 0.8334,247 "step": 165248 },249 {250 "epoch": 0.07059800664451828,251 "grad_norm": 3.00742757115455,252 "learning_rate": 7.053941908713694e-06,253 "loss": 0.8233,254 "step": 170255 },256 {257 "epoch": 0.07267441860465117,258 "grad_norm": 3.0518393701751485,259 "learning_rate": 7.261410788381743e-06,260 "loss": 0.8296,261 "step": 175262 },263 {264 "epoch": 0.07475083056478406,265 "grad_norm": 3.1984146233667263,266 "learning_rate": 7.468879668049793e-06,267 "loss": 0.8155,268 "step": 180269 },270 {271 "epoch": 0.07682724252491695,272 "grad_norm": 3.05629449726749,273 "learning_rate": 7.676348547717844e-06,274 "loss": 0.8377,275 "step": 185276 },277 {278 "epoch": 0.07890365448504984,279 "grad_norm": 3.249088059891964,280 "learning_rate": 7.883817427385892e-06,281 "loss": 0.8432,282 "step": 190283 },284 {285 "epoch": 0.08098006644518273,286 "grad_norm": 3.0028481508425515,287 "learning_rate": 8.091286307053943e-06,288 "loss": 0.8173,289 "step": 195290 },291 {292 "epoch": 0.08305647840531562,293 "grad_norm": 3.059733445916786,294 "learning_rate": 8.298755186721992e-06,295 "loss": 0.8227,296 "step": 200297 },298 {299 "epoch": 0.08513289036544851,300 "grad_norm": 3.0867633236533365,301 "learning_rate": 8.506224066390042e-06,302 "loss": 0.8181,303 "step": 205304 },305 {306 "epoch": 0.0872093023255814,307 "grad_norm": 2.997953986592159,308 "learning_rate": 8.713692946058093e-06,309 "loss": 0.821,310 "step": 210311 },312 {313 "epoch": 0.08928571428571429,314 "grad_norm": 3.2351659520743072,315 "learning_rate": 8.921161825726142e-06,316 "loss": 0.8294,317 "step": 215318 },319 {320 "epoch": 0.09136212624584718,321 "grad_norm": 3.1494481731597586,322 "learning_rate": 9.128630705394191e-06,323 "loss": 0.8261,324 "step": 220325 },326 {327 "epoch": 0.09343853820598007,328 "grad_norm": 3.105511823234228,329 "learning_rate": 9.33609958506224e-06,330 "loss": 0.8165,331 "step": 225332 },333 {334 "epoch": 0.09551495016611296,335 "grad_norm": 3.023901781664328,336 "learning_rate": 9.543568464730292e-06,337 "loss": 0.8123,338 "step": 230339 },340 {341 "epoch": 0.09759136212624585,342 "grad_norm": 3.4303556589177187,343 "learning_rate": 9.751037344398341e-06,344 "loss": 0.8093,345 "step": 235346 },347 {348 "epoch": 0.09966777408637874,349 "grad_norm": 3.6054989714255408,350 "learning_rate": 9.95850622406639e-06,351 "loss": 0.8201,352 "step": 240353 },354 {355 "epoch": 0.10174418604651163,356 "grad_norm": 2.990225009601177,357 "learning_rate": 9.999915930067828e-06,358 "loss": 0.8208,359 "step": 245360 },361 {362 "epoch": 0.10382059800664452,363 "grad_norm": 2.9957103647324264,364 "learning_rate": 9.999574400813641e-06,365 "loss": 0.816,366 "step": 250367 },368 {369 "epoch": 0.10589700996677741,370 "grad_norm": 2.8988415018010287,371 "learning_rate": 9.998970175798065e-06,372 "loss": 0.8044,373 "step": 255374 },375 {376 "epoch": 0.1079734219269103,377 "grad_norm": 2.893907971746992,378 "learning_rate": 9.998103286769267e-06,379 "loss": 0.799,380 "step": 260381 },382 {383 "epoch": 0.11004983388704319,384 "grad_norm": 2.898946354458808,385 "learning_rate": 9.996973779276743e-06,386 "loss": 0.8113,387 "step": 265388 },389 {390 "epoch": 0.11212624584717608,391 "grad_norm": 3.0485697591450998,392 "learning_rate": 9.99558171266891e-06,393 "loss": 0.8194,394 "step": 270395 },396 {397 "epoch": 0.11420265780730897,398 "grad_norm": 2.933613250090363,399 "learning_rate": 9.993927160089991e-06,400 "loss": 0.7981,401 "step": 275402 },403 {404 "epoch": 0.11627906976744186,405 "grad_norm": 2.900283777987733,406 "learning_rate": 9.992010208476178e-06,407 "loss": 0.8114,408 "step": 280409 },410 {411 "epoch": 0.11835548172757475,412 "grad_norm": 2.869639926652705,413 "learning_rate": 9.989830958551058e-06,414 "loss": 0.8026,415 "step": 285416 },417 {418 "epoch": 0.12043189368770764,419 "grad_norm": 3.0764284732072236,420 "learning_rate": 9.98738952482032e-06,421 "loss": 0.7816,422 "step": 290423 },424 {425 "epoch": 0.12250830564784053,426 "grad_norm": 2.872848930860205,427 "learning_rate": 9.984686035565742e-06,428 "loss": 0.7851,429 "step": 295430 },431 {432 "epoch": 0.12458471760797342,433 "grad_norm": 2.7170384439590367,434 "learning_rate": 9.98172063283845e-06,435 "loss": 0.8054,436 "step": 300437 },438 {439 "epoch": 0.12666112956810632,440 "grad_norm": 2.785739578421159,441 "learning_rate": 9.978493472451451e-06,442 "loss": 0.7824,443 "step": 305444 },445 {446 "epoch": 0.1287375415282392,447 "grad_norm": 2.955753943035507,448 "learning_rate": 9.975004723971452e-06,449 "loss": 0.7788,450 "step": 310451 },452 {453 "epoch": 0.1308139534883721,454 "grad_norm": 2.7566534229071378,455 "learning_rate": 9.971254570709939e-06,456 "loss": 0.7804,457 "step": 315458 },459 {460 "epoch": 0.132890365448505,461 "grad_norm": 3.0399050026271945,462 "learning_rate": 9.967243209713563e-06,463 "loss": 0.7712,464 "step": 320465 },466 {467 "epoch": 0.13496677740863788,468 "grad_norm": 3.227011718605211,469 "learning_rate": 9.962970851753767e-06,470 "loss": 0.7852,471 "step": 325472 },473 {474 "epoch": 0.13704318936877077,475 "grad_norm": 2.894940556652265,476 "learning_rate": 9.95843772131573e-06,477 "loss": 0.767,478 "step": 330479 },480 {481 "epoch": 0.13911960132890366,482 "grad_norm": 3.137972193410393,483 "learning_rate": 9.95364405658655e-06,484 "loss": 0.77,485 "step": 335486 },487 {488 "epoch": 0.14119601328903655,489 "grad_norm": 2.7913612546678426,490 "learning_rate": 9.948590109442755e-06,491 "loss": 0.7768,492 "step": 340493 },494 {495 "epoch": 0.14327242524916944,496 "grad_norm": 2.893979747266515,497 "learning_rate": 9.94327614543704e-06,498 "loss": 0.7827,499 "step": 345500 },501 {502 "epoch": 0.14534883720930233,503 "grad_norm": 2.665071280290936,504 "learning_rate": 9.937702443784343e-06,505 "loss": 0.7474,506 "step": 350507 },508 {509 "epoch": 0.14742524916943522,510 "grad_norm": 2.741350083908129,511 "learning_rate": 9.931869297347146e-06,512 "loss": 0.7638,513 "step": 355514 },515 {516 "epoch": 0.14950166112956811,517 "grad_norm": 2.9878149207237357,518 "learning_rate": 9.925777012620111e-06,519 "loss": 0.7419,520 "step": 360521 },522 {523 "epoch": 0.151578073089701,524 "grad_norm": 2.801227928713699,525 "learning_rate": 9.919425909713958e-06,526 "loss": 0.769,527 "step": 365528 },529 {530 "epoch": 0.1536544850498339,531 "grad_norm": 3.023770968839729,532 "learning_rate": 9.912816322338659e-06,533 "loss": 0.7447,534 "step": 370535 },536 {537 "epoch": 0.15573089700996678,538 "grad_norm": 2.9927287523796715,539 "learning_rate": 9.905948597785888e-06,540 "loss": 0.754,541 "step": 375542 },543 {544 "epoch": 0.15780730897009967,545 "grad_norm": 11.785492453222856,546 "learning_rate": 9.89882309691079e-06,547 "loss": 0.7497,548 "step": 380549 },550 {551 "epoch": 0.15988372093023256,552 "grad_norm": 2.8752234411604682,553 "learning_rate": 9.891440194113008e-06,554 "loss": 0.7427,555 "step": 385556 },557 {558 "epoch": 0.16196013289036545,559 "grad_norm": 3.097207390376622,560 "learning_rate": 9.88380027731702e-06,561 "loss": 0.7542,562 "step": 390563 },564 {565 "epoch": 0.16403654485049834,566 "grad_norm": 2.921991118334764,567 "learning_rate": 9.875903747951742e-06,568 "loss": 0.7621,569 "step": 395570 },571 {572 "epoch": 0.16611295681063123,573 "grad_norm": 2.8395297947865963,574 "learning_rate": 9.867751020929454e-06,575 "loss": 0.735,576 "step": 400577 },578 {579 "epoch": 0.16818936877076412,580 "grad_norm": 2.726116425089643,581 "learning_rate": 9.859342524623985e-06,582 "loss": 0.7124,583 "step": 405584 },585 {586 "epoch": 0.17026578073089702,587 "grad_norm": 3.2173444091652943,588 "learning_rate": 9.850678700848208e-06,589 "loss": 0.7374,590 "step": 410591 },592 {593 "epoch": 0.1723421926910299,594 "grad_norm": 2.716930762983964,595 "learning_rate": 9.84176000483083e-06,596 "loss": 0.7138,597 "step": 415598 },599 {600 "epoch": 0.1744186046511628,601 "grad_norm": 2.985441779621083,602 "learning_rate": 9.832586905192469e-06,603 "loss": 0.731,604 "step": 420605 },606 {607 "epoch": 0.17649501661129569,608 "grad_norm": 3.032790315651323,609 "learning_rate": 9.823159883921028e-06,610 "loss": 0.7215,611 "step": 425612 },613 {614 "epoch": 0.17857142857142858,615 "grad_norm": 2.6988344818168155,616 "learning_rate": 9.813479436346378e-06,617 "loss": 0.7183,618 "step": 430619 },620 {621 "epoch": 0.18064784053156147,622 "grad_norm": 2.973146607192177,623 "learning_rate": 9.803546071114323e-06,624 "loss": 0.7311,625 "step": 435626 },627 {628 "epoch": 0.18272425249169436,629 "grad_norm": 2.9093506646801344,630 "learning_rate": 9.793360310159878e-06,631 "loss": 0.7049,632 "step": 440633 },634 {635 "epoch": 0.18480066445182725,636 "grad_norm": 3.01100096145872,637 "learning_rate": 9.782922688679847e-06,638 "loss": 0.7118,639 "step": 445640 },641 {642 "epoch": 0.18687707641196014,643 "grad_norm": 2.716470652939527,644 "learning_rate": 9.772233755104695e-06,645 "loss": 0.7277,646 "step": 450647 },648 {649 "epoch": 0.18895348837209303,650 "grad_norm": 2.7134248053870165,651 "learning_rate": 9.761294071069736e-06,652 "loss": 0.7205,653 "step": 455654 },655 {656 "epoch": 0.19102990033222592,657 "grad_norm": 2.6251507638777163,658 "learning_rate": 9.750104211385625e-06,659 "loss": 0.7152,660 "step": 460661 },662 {663 "epoch": 0.1931063122923588,664 "grad_norm": 2.8023948010803483,665 "learning_rate": 9.738664764008149e-06,666 "loss": 0.7233,667 "step": 465668 },669 {670 "epoch": 0.1951827242524917,671 "grad_norm": 3.714290449563204,672 "learning_rate": 9.726976330007341e-06,673 "loss": 0.6998,674 "step": 470675 },676 {677 "epoch": 0.1972591362126246,678 "grad_norm": 2.8670419197216512,679 "learning_rate": 9.71503952353589e-06,680 "loss": 0.6985,681 "step": 475682 },683 {684 "epoch": 0.19933554817275748,685 "grad_norm": 3.1683988394439107,686 "learning_rate": 9.702854971796876e-06,687 "loss": 0.7089,688 "step": 480689 },690 {691 "epoch": 0.20141196013289037,692 "grad_norm": 3.2223078839261166,693 "learning_rate": 9.690423315010814e-06,694 "loss": 0.7053,695 "step": 485696 },697 {698 "epoch": 0.20348837209302326,699 "grad_norm": 2.77875488832717,700 "learning_rate": 9.677745206382014e-06,701 "loss": 0.7271,702 "step": 490703 },704 {705 "epoch": 0.20556478405315615,706 "grad_norm": 2.888271933836237,707 "learning_rate": 9.664821312064258e-06,708 "loss": 0.7018,709 "step": 495710 },711 {712 "epoch": 0.20764119601328904,713 "grad_norm": 3.2746008040723815,714 "learning_rate": 9.651652311125803e-06,715 "loss": 0.6991,716 "step": 500717 },718 {719 "epoch": 0.20971760797342193,720 "grad_norm": 2.76622547311742,721 "learning_rate": 9.638238895513687e-06,722 "loss": 0.7075,723 "step": 505724 },725 {726 "epoch": 0.21179401993355482,727 "grad_norm": 2.9972446036957114,728 "learning_rate": 9.624581770017392e-06,729 "loss": 0.6857,730 "step": 510731 },732 {733 "epoch": 0.2138704318936877,734 "grad_norm": 2.869516499460042,735 "learning_rate": 9.610681652231794e-06,736 "loss": 0.6916,737 "step": 515738 },739 {740 "epoch": 0.2159468438538206,741 "grad_norm": 2.742923434452921,742 "learning_rate": 9.596539272519468e-06,743 "loss": 0.6811,744 "step": 520745 },746 {747 "epoch": 0.2180232558139535,748 "grad_norm": 2.8482023108565677,749 "learning_rate": 9.582155373972303e-06,750 "loss": 0.6744,751 "step": 525752 },753 {754 "epoch": 0.22009966777408638,755 "grad_norm": 2.9348099403663124,756 "learning_rate": 9.56753071237247e-06,757 "loss": 0.6776,758 "step": 530759 },760 {761 "epoch": 0.22217607973421927,762 "grad_norm": 2.786772996017183,763 "learning_rate": 9.552666056152704e-06,764 "loss": 0.6798,765 "step": 535766 },767 {768 "epoch": 0.22425249169435216,769 "grad_norm": 2.92722689041533,770 "learning_rate": 9.537562186355918e-06,771 "loss": 0.6843,772 "step": 540773 },774 {775 "epoch": 0.22632890365448505,776 "grad_norm": 2.7694998172195207,777 "learning_rate": 9.52221989659418e-06,778 "loss": 0.6938,779 "step": 545780 },781 {782 "epoch": 0.22840531561461794,783 "grad_norm": 2.9300442858036244,784 "learning_rate": 9.506639993007012e-06,785 "loss": 0.6944,786 "step": 550787 },788 {789 "epoch": 0.23048172757475083,790 "grad_norm": 3.1035204783454993,791 "learning_rate": 9.490823294219015e-06,792 "loss": 0.672,793 "step": 555794 },795 {796 "epoch": 0.23255813953488372,797 "grad_norm": 2.6193387690961245,798 "learning_rate": 9.474770631296882e-06,799 "loss": 0.6561,800 "step": 560801 },802 {803 "epoch": 0.2346345514950166,804 "grad_norm": 2.61646550507026,805 "learning_rate": 9.458482847705705e-06,806 "loss": 0.6576,807 "step": 565808 },809 {810 "epoch": 0.2367109634551495,811 "grad_norm": 2.756473668019519,812 "learning_rate": 9.441960799264678e-06,813 "loss": 0.6851,814 "step": 570815 },816 {817 "epoch": 0.2387873754152824,818 "grad_norm": 2.6995089678231614,819 "learning_rate": 9.425205354102111e-06,820 "loss": 0.6648,821 "step": 575822 },823 {824 "epoch": 0.24086378737541528,825 "grad_norm": 2.7140254791209677,826 "learning_rate": 9.408217392609831e-06,827 "loss": 0.6451,828 "step": 580829 },830 {831 "epoch": 0.24294019933554817,832 "grad_norm": 2.607599787114018,833 "learning_rate": 9.390997807396912e-06,834 "loss": 0.67,835 "step": 585836 },837 {838 "epoch": 0.24501661129568106,839 "grad_norm": 2.8420050898692764,840 "learning_rate": 9.373547503242775e-06,841 "loss": 0.6657,842 "step": 590843 },844 {845 "epoch": 0.24709302325581395,846 "grad_norm": 2.9228965685399095,847 "learning_rate": 9.355867397049658e-06,848 "loss": 0.6566,849 "step": 595850 },851 {852 "epoch": 0.24916943521594684,853 "grad_norm": 2.8048600929777403,854 "learning_rate": 9.337958417794425e-06,855 "loss": 0.6457,856 "step": 600857 },858 {859 "epoch": 0.25124584717607973,860 "grad_norm": 2.6983485281997415,861 "learning_rate": 9.319821506479762e-06,862 "loss": 0.6376,863 "step": 605864 },865 {866 "epoch": 0.25332225913621265,867 "grad_norm": 2.801805288954333,868 "learning_rate": 9.301457616084733e-06,869 "loss": 0.6523,870 "step": 610871 },872 {873 "epoch": 0.2553986710963455,874 "grad_norm": 2.820864396273499,875 "learning_rate": 9.282867711514703e-06,876 "loss": 0.6365,877 "step": 615878 },879 {880 "epoch": 0.2574750830564784,881 "grad_norm": 2.9932167823643043,882 "learning_rate": 9.264052769550643e-06,883 "loss": 0.6425,884 "step": 620885 },886 {887 "epoch": 0.2595514950166113,888 "grad_norm": 2.6556108045628544,889 "learning_rate": 9.245013778797802e-06,890 "loss": 0.6562,891 "step": 625892 },893 {894 "epoch": 0.2616279069767442,895 "grad_norm": 2.676416816690246,896 "learning_rate": 9.225751739633772e-06,897 "loss": 0.6387,898 "step": 630899 },900 {901 "epoch": 0.26370431893687707,902 "grad_norm": 2.702226526508375,903 "learning_rate": 9.206267664155906e-06,904 "loss": 0.6348,905 "step": 635906 },907 {908 "epoch": 0.26578073089701,909 "grad_norm": 2.637563222880754,910 "learning_rate": 9.186562576128159e-06,911 "loss": 0.6263,912 "step": 640913 },914 {915 "epoch": 0.26785714285714285,916 "grad_norm": 2.7815352111724603,917 "learning_rate": 9.16663751092728e-06,918 "loss": 0.6362,919 "step": 645920 },921 {922 "epoch": 0.26993355481727577,923 "grad_norm": 2.8822755136904528,924 "learning_rate": 9.146493515488418e-06,925 "loss": 0.6164,926 "step": 650927 },928 {929 "epoch": 0.27200996677740863,930 "grad_norm": 2.5755107274498146,931 "learning_rate": 9.126131648250112e-06,932 "loss": 0.6342,933 "step": 655934 },935 {936 "epoch": 0.27408637873754155,937 "grad_norm": 2.584492766117294,938 "learning_rate": 9.105552979098675e-06,939 "loss": 0.6329,940 "step": 660941 },942 {943 "epoch": 0.2761627906976744,944 "grad_norm": 2.6805388863449036,945 "learning_rate": 9.084758589311977e-06,946 "loss": 0.6307,947 "step": 665948 },949 {950 "epoch": 0.2782392026578073,951 "grad_norm": 2.7584115266730693,952 "learning_rate": 9.063749571502633e-06,953 "loss": 0.6374,954 "step": 670955 },956 {957 "epoch": 0.2803156146179402,958 "grad_norm": 2.8092430217085145,959 "learning_rate": 9.04252702956059e-06,960 "loss": 0.6282,961 "step": 675962 },963 {964 "epoch": 0.2823920265780731,965 "grad_norm": 2.6353604501522168,966 "learning_rate": 9.021092078595132e-06,967 "loss": 0.6332,968 "step": 680969 },970 {971 "epoch": 0.28446843853820597,972 "grad_norm": 2.7859177417571486,973 "learning_rate": 8.999445844876276e-06,974 "loss": 0.6381,975 "step": 685976 },977 {978 "epoch": 0.2865448504983389,979 "grad_norm": 2.6603634875986457,980 "learning_rate": 8.977589465775607e-06,981 "loss": 0.6312,982 "step": 690983 },984 {985 "epoch": 0.28862126245847175,986 "grad_norm": 2.6293766795824354,987 "learning_rate": 8.955524089706506e-06,988 "loss": 0.5999,989 "step": 695990 },991 {992 "epoch": 0.29069767441860467,993 "grad_norm": 2.8986723382239967,994 "learning_rate": 8.933250876063815e-06,995 "loss": 0.6297,996 "step": 700997 },998 {999 "epoch": 0.29277408637873753,1000 "grad_norm": 2.6589365161649834,1001 "learning_rate": 8.910770995162913e-06,1002 "loss": 0.6303,1003 "step": 7051004 },1005 {1006 "epoch": 0.29485049833887045,1007 "grad_norm": 2.64992234535583,1008 "learning_rate": 8.88808562817823e-06,1009 "loss": 0.6114,1010 "step": 7101011 },1012 {1013 "epoch": 0.2969269102990033,1014 "grad_norm": 2.7322760412568776,1015 "learning_rate": 8.865195967081174e-06,1016 "loss": 0.6215,1017 "step": 7151018 },1019 {1020 "epoch": 0.29900332225913623,1021 "grad_norm": 2.576473302210113,1022 "learning_rate": 8.842103214577511e-06,1023 "loss": 0.6147,1024 "step": 7201025 },1026 {1027 "epoch": 0.3010797342192691,1028 "grad_norm": 2.507546434543662,1029 "learning_rate": 8.818808584044163e-06,1030 "loss": 0.6089,1031 "step": 7251032 },1033 {1034 "epoch": 0.303156146179402,1035 "grad_norm": 2.953501799132662,1036 "learning_rate": 8.795313299465455e-06,1037 "loss": 0.6147,1038 "step": 7301039 },1040 {1041 "epoch": 0.30523255813953487,1042 "grad_norm": 2.58266860044093,1043 "learning_rate": 8.771618595368806e-06,1044 "loss": 0.6024,1045 "step": 7351046 },1047 {1048 "epoch": 0.3073089700996678,1049 "grad_norm": 2.7291039422306613,1050 "learning_rate": 8.747725716759859e-06,1051 "loss": 0.6152,1052 "step": 7401053 },1054 {1055 "epoch": 0.30938538205980065,1056 "grad_norm": 2.696653736904745,1057 "learning_rate": 8.723635919057058e-06,1058 "loss": 0.6082,1059 "step": 7451060 },1061 {1062 "epoch": 0.31146179401993357,1063 "grad_norm": 2.639188973608746,1064 "learning_rate": 8.699350468025699e-06,1065 "loss": 0.5924,1066 "step": 7501067 },1068 {1069 "epoch": 0.31353820598006643,1070 "grad_norm": 2.5960120065556294,1071 "learning_rate": 8.674870639711403e-06,1072 "loss": 0.5871,1073 "step": 7551074 },1075 {1076 "epoch": 0.31561461794019935,1077 "grad_norm": 2.691098687645451,1078 "learning_rate": 8.650197720373091e-06,1079 "loss": 0.5937,1080 "step": 7601081 },1082 {1083 "epoch": 0.3176910299003322,1084 "grad_norm": 2.7922815680081947,1085 "learning_rate": 8.625333006415372e-06,1086 "loss": 0.5806,1087 "step": 7651088 },1089 {1090 "epoch": 0.31976744186046513,1091 "grad_norm": 2.5989983221444635,1092 "learning_rate": 8.600277804320452e-06,1093 "loss": 0.5889,1094 "step": 7701095 },1096 {1097 "epoch": 0.321843853820598,1098 "grad_norm": 2.7500580415708553,1099 "learning_rate": 8.575033430579465e-06,1100 "loss": 0.5929,1101 "step": 7751102 },1103 {1104 "epoch": 0.3239202657807309,1105 "grad_norm": 2.9863748696055485,1106 "learning_rate": 8.549601211623316e-06,1107 "loss": 0.5905,1108 "step": 7801109 },1110 {1111 "epoch": 0.32599667774086377,1112 "grad_norm": 2.7128601524461966,1113 "learning_rate": 8.523982483752973e-06,1114 "loss": 0.5838,1115 "step": 7851116 },1117 {1118 "epoch": 0.3280730897009967,1119 "grad_norm": 2.6273588590853727,1120 "learning_rate": 8.498178593069262e-06,1121 "loss": 0.579,1122 "step": 7901123 },1124 {1125 "epoch": 0.33014950166112955,1126 "grad_norm": 2.6424251208940714,1127 "learning_rate": 8.472190895402131e-06,1128 "loss": 0.568,1129 "step": 7951130 },1131 {1132 "epoch": 0.33222591362126247,1133 "grad_norm": 2.774060760650428,1134 "learning_rate": 8.446020756239418e-06,1135 "loss": 0.5881,1136 "step": 8001137 },1138 {1139 "epoch": 0.33430232558139533,1140 "grad_norm": 2.7429673227633193,1141 "learning_rate": 8.419669550655093e-06,1142 "loss": 0.5807,1143 "step": 8051144 },1145 {1146 "epoch": 0.33637873754152825,1147 "grad_norm": 2.4588138685140164,1148 "learning_rate": 8.393138663237015e-06,1149 "loss": 0.5699,1150 "step": 8101151 },1152 {1153 "epoch": 0.3384551495016611,1154 "grad_norm": 2.8894345220890845,1155 "learning_rate": 8.366429488014178e-06,1156 "loss": 0.5644,1157 "step": 8151158 },1159 {1160 "epoch": 0.34053156146179403,1161 "grad_norm": 2.6417969175920253,1162 "learning_rate": 8.339543428383467e-06,1163 "loss": 0.577,1164 "step": 8201165 },1166 {1167 "epoch": 0.3426079734219269,1168 "grad_norm": 2.639049529021501,1169 "learning_rate": 8.312481897035906e-06,1170 "loss": 0.5835,1171 "step": 8251172 },1173 {1174 "epoch": 0.3446843853820598,1175 "grad_norm": 2.791601353912272,1176 "learning_rate": 8.285246315882448e-06,1177 "loss": 0.5873,1178 "step": 8301179 },1180 {1181 "epoch": 0.3467607973421927,1182 "grad_norm": 2.760486538247162,1183 "learning_rate": 8.257838115979244e-06,1184 "loss": 0.5743,1185 "step": 8351186 },1187 {1188 "epoch": 0.3488372093023256,1189 "grad_norm": 2.6084506349864114,1190 "learning_rate": 8.230258737452473e-06,1191 "loss": 0.5835,1192 "step": 8401193 },1194 {1195 "epoch": 0.35091362126245845,1196 "grad_norm": 2.568077365967415,1197 "learning_rate": 8.202509629422647e-06,1198 "loss": 0.5663,1199 "step": 8451200 },