XGenerationLab/XiYanSQL-QwenCoder-3B-2502
5201
1{2 "best_metric": null,3 "best_model_checkpoint": null,4 "epoch": 1.5296367112810707,5 "eval_steps": 500,6 "global_step": 4000,7 "is_hyper_param_search": false,8 "is_local_process_zero": true,9 "is_world_process_zero": true,10 "log_history": [11 {12 "epoch": 0.0003824091778202677,13 "grad_norm": 8.127729221609483,14 "learning_rate": 0.0,15 "loss": 0.7499,16 "step": 117 },18 {19 "epoch": 0.0007648183556405354,20 "grad_norm": 7.963595529643983,21 "learning_rate": 9.658899046395883e-08,22 "loss": 0.7745,23 "step": 224 },25 {26 "epoch": 0.001147227533460803,27 "grad_norm": 6.399453926631711,28 "learning_rate": 1.530899278678881e-07,29 "loss": 0.827,30 "step": 331 },32 {33 "epoch": 0.0015296367112810707,34 "grad_norm": 7.212906755583777,35 "learning_rate": 1.9317798092791765e-07,36 "loss": 0.9354,37 "step": 438 },39 {40 "epoch": 0.0019120458891013384,41 "grad_norm": 4.403583991640848,42 "learning_rate": 2.2427269061507353e-07,43 "loss": 0.8108,44 "step": 545 },46 {47 "epoch": 0.002294455066921606,48 "grad_norm": 9.30785115586022,49 "learning_rate": 2.496789183318469e-07,50 "loss": 0.8121,51 "step": 652 },53 {54 "epoch": 0.0026768642447418736,55 "grad_norm": 10.972546390514474,56 "learning_rate": 2.711595777955698e-07,57 "loss": 0.8276,58 "step": 759 },60 {61 "epoch": 0.0030592734225621415,62 "grad_norm": 7.964528600005935,63 "learning_rate": 2.8976697139187646e-07,64 "loss": 0.889,65 "step": 866 },67 {68 "epoch": 0.0034416826003824093,69 "grad_norm": 11.538619501059664,70 "learning_rate": 3.061798557357762e-07,71 "loss": 0.5456,72 "step": 973 },74 {75 "epoch": 0.0038240917782026767,76 "grad_norm": 9.923505802246833,77 "learning_rate": 3.208616810790324e-07,78 "loss": 0.5617,79 "step": 1080 },81 {82 "epoch": 0.004206500956022944,83 "grad_norm": 13.868230594571852,84 "learning_rate": 3.341430076232756e-07,85 "loss": 0.8914,86 "step": 1187 },88 {89 "epoch": 0.004588910133843212,90 "grad_norm": 15.82952470982823,91 "learning_rate": 3.4626790879580574e-07,92 "loss": 0.3943,93 "step": 1294 },95 {96 "epoch": 0.00497131931166348,97 "grad_norm": 13.881207151660286,98 "learning_rate": 3.5742173664798447e-07,99 "loss": 0.9886,100 "step": 13101 },102 {103 "epoch": 0.005353728489483747,104 "grad_norm": 9.351576777218634,105 "learning_rate": 3.677485682595286e-07,106 "loss": 1.2225,107 "step": 14108 },109 {110 "epoch": 0.0057361376673040155,111 "grad_norm": 6.921096792456155,112 "learning_rate": 3.773626184829616e-07,113 "loss": 0.6324,114 "step": 15115 },116 {117 "epoch": 0.006118546845124283,118 "grad_norm": 7.911741257096208,119 "learning_rate": 3.863559618558353e-07,120 "loss": 1.0263,121 "step": 16122 },123 {124 "epoch": 0.00650095602294455,125 "grad_norm": 9.023578698841147,126 "learning_rate": 3.948039093953151e-07,127 "loss": 0.8379,128 "step": 17129 },130 {131 "epoch": 0.006883365200764819,132 "grad_norm": 5.773447080611977,133 "learning_rate": 4.02768846199735e-07,134 "loss": 0.6313,135 "step": 18136 },137 {138 "epoch": 0.007265774378585086,139 "grad_norm": 6.391540312564622,140 "learning_rate": 4.103030300875908e-07,141 "loss": 0.6562,142 "step": 19143 },144 {145 "epoch": 0.0076481835564053535,146 "grad_norm": 4.734698092574861,147 "learning_rate": 4.174506715429912e-07,148 "loss": 0.9944,149 "step": 20150 },151 {152 "epoch": 0.00803059273422562,153 "grad_norm": 5.543325744703233,154 "learning_rate": 4.242495056634579e-07,155 "loss": 0.8748,156 "step": 21157 },158 {159 "epoch": 0.008413001912045888,160 "grad_norm": 6.139313458259769,161 "learning_rate": 4.3073199808723443e-07,162 "loss": 0.6497,163 "step": 22164 },165 {166 "epoch": 0.008795411089866157,167 "grad_norm": 10.51882881230223,168 "learning_rate": 4.369262826367178e-07,169 "loss": 0.7235,170 "step": 23171 },172 {173 "epoch": 0.009177820267686425,174 "grad_norm": 9.14158481873066,175 "learning_rate": 4.428568992597646e-07,176 "loss": 0.4482,177 "step": 24178 },179 {180 "epoch": 0.009560229445506692,181 "grad_norm": 10.990275506131937,182 "learning_rate": 4.4854538123014706e-07,183 "loss": 0.8668,184 "step": 25185 },186 {187 "epoch": 0.00994263862332696,188 "grad_norm": 11.815499706960994,189 "learning_rate": 4.540107271119433e-07,190 "loss": 0.8262,191 "step": 26192 },193 {194 "epoch": 0.010325047801147227,195 "grad_norm": 7.542823857279405,196 "learning_rate": 4.592697836036643e-07,197 "loss": 0.9995,198 "step": 27199 },200 {201 "epoch": 0.010707456978967494,202 "grad_norm": 3.439453125,203 "learning_rate": 4.6433755872348744e-07,204 "loss": 0.652,205 "step": 28206 },207 {208 "epoch": 0.011089866156787764,209 "grad_norm": 4.447922408065247,210 "learning_rate": 4.6922748001247033e-07,211 "loss": 0.4116,212 "step": 29213 },214 {215 "epoch": 0.011472275334608031,216 "grad_norm": 3.5419146432268502,217 "learning_rate": 4.739516089469205e-07,218 "loss": 0.603,219 "step": 30220 },221 {222 "epoch": 0.011854684512428298,223 "grad_norm": 5.595453269399992,224 "learning_rate": 4.78520820180538e-07,225 "loss": 0.8754,226 "step": 31227 },228 {229 "epoch": 0.012237093690248566,230 "grad_norm": 4.664053298711087,231 "learning_rate": 4.829449523197942e-07,232 "loss": 0.728,233 "step": 32234 },235 {236 "epoch": 0.012619502868068833,237 "grad_norm": 4.184285994388638,238 "learning_rate": 4.872329354911637e-07,239 "loss": 0.7668,240 "step": 33241 },242 {243 "epoch": 0.0130019120458891,244 "grad_norm": 3.8462487759245723,245 "learning_rate": 4.91392899859274e-07,246 "loss": 0.7642,247 "step": 34248 },249 {250 "epoch": 0.01338432122370937,251 "grad_norm": 5.035267807203385,252 "learning_rate": 4.954322684106434e-07,253 "loss": 0.4231,254 "step": 35255 },256 {257 "epoch": 0.013766730401529637,258 "grad_norm": 4.700427932737028,259 "learning_rate": 4.993578366636938e-07,260 "loss": 0.4554,261 "step": 36262 },263 {264 "epoch": 0.014149139579349905,265 "grad_norm": 6.65272564780195,266 "learning_rate": 5.031758414551729e-07,267 "loss": 0.3467,268 "step": 37269 },270 {271 "epoch": 0.014531548757170172,272 "grad_norm": 6.540097900964421,273 "learning_rate": 5.068920205515497e-07,274 "loss": 0.5521,275 "step": 38276 },277 {278 "epoch": 0.01491395793499044,279 "grad_norm": 6.187611588521872,280 "learning_rate": 5.105116645158726e-07,281 "loss": 0.8477,282 "step": 39283 },284 {285 "epoch": 0.015296367112810707,286 "grad_norm": 4.212933408814397,287 "learning_rate": 5.140396620069501e-07,288 "loss": 0.8013,289 "step": 40290 },291 {292 "epoch": 0.015678776290630976,293 "grad_norm": 3.7306584334956354,294 "learning_rate": 5.174805394842197e-07,295 "loss": 0.34,296 "step": 41297 },298 {299 "epoch": 0.01606118546845124,300 "grad_norm": 3.8129084634065613,301 "learning_rate": 5.208384961274167e-07,302 "loss": 0.7891,303 "step": 42304 },305 {306 "epoch": 0.01644359464627151,307 "grad_norm": 3.521126067201788,308 "learning_rate": 5.241174346468368e-07,309 "loss": 0.8447,310 "step": 43311 },312 {313 "epoch": 0.016826003824091777,314 "grad_norm": 5.279756052566196,315 "learning_rate": 5.273209885511933e-07,316 "loss": 0.6151,317 "step": 44318 },319 {320 "epoch": 0.017208413001912046,321 "grad_norm": 4.009723289626729,322 "learning_rate": 5.304525463508497e-07,323 "loss": 0.6095,324 "step": 45325 },326 {327 "epoch": 0.017590822179732315,328 "grad_norm": 3.7728758842220267,329 "learning_rate": 5.335152731006765e-07,330 "loss": 0.3871,331 "step": 46332 },333 {334 "epoch": 0.01797323135755258,335 "grad_norm": 4.866870067648313,336 "learning_rate": 5.365121296259033e-07,337 "loss": 0.4796,338 "step": 47339 },340 {341 "epoch": 0.01835564053537285,342 "grad_norm": 4.8744050665499286,343 "learning_rate": 5.394458897237234e-07,344 "loss": 0.5812,345 "step": 48346 },347 {348 "epoch": 0.018738049713193115,349 "grad_norm": 9.108527378017467,350 "learning_rate": 5.423191555911396e-07,351 "loss": 0.319,352 "step": 49353 },354 {355 "epoch": 0.019120458891013385,356 "grad_norm": 9.158412014105785,357 "learning_rate": 5.45134371694106e-07,358 "loss": 0.7643,359 "step": 50360 },361 {362 "epoch": 0.019502868068833654,363 "grad_norm": 5.790040627239837,364 "learning_rate": 5.478938372632032e-07,365 "loss": 0.6109,366 "step": 51367 },368 {369 "epoch": 0.01988527724665392,370 "grad_norm": 6.096683676341199,371 "learning_rate": 5.505997175759021e-07,372 "loss": 0.9369,373 "step": 52374 },375 {376 "epoch": 0.02026768642447419,377 "grad_norm": 4.441546891668887,378 "learning_rate": 5.532540541641197e-07,379 "loss": 0.7858,380 "step": 53381 },382 {383 "epoch": 0.020650095602294454,384 "grad_norm": 2.8696014470488382,385 "learning_rate": 5.558587740676232e-07,386 "loss": 0.8667,387 "step": 54388 },389 {390 "epoch": 0.021032504780114723,391 "grad_norm": 3.2738861286399046,392 "learning_rate": 5.584156982383491e-07,393 "loss": 0.4673,394 "step": 55395 },396 {397 "epoch": 0.02141491395793499,398 "grad_norm": 3.697994492835673,399 "learning_rate": 5.609265491874463e-07,400 "loss": 0.6328,401 "step": 56402 },403 {404 "epoch": 0.021797323135755258,405 "grad_norm": 3.125139767382225,406 "learning_rate": 5.633929579554789e-07,407 "loss": 0.4692,408 "step": 57409 },410 {411 "epoch": 0.022179732313575527,412 "grad_norm": 4.298443428237915,413 "learning_rate": 5.658164704764291e-07,414 "loss": 0.6768,415 "step": 58416 },417 {418 "epoch": 0.022562141491395793,419 "grad_norm": 4.398677297665077,420 "learning_rate": 5.681985533976847e-07,421 "loss": 0.6619,422 "step": 59423 },424 {425 "epoch": 0.022944550669216062,426 "grad_norm": 5.818330527160778,427 "learning_rate": 5.705405994108792e-07,428 "loss": 0.6268,429 "step": 60430 },431 {432 "epoch": 0.023326959847036328,433 "grad_norm": 5.36324426425703,434 "learning_rate": 5.728439321421063e-07,435 "loss": 0.3157,436 "step": 61437 },438 {439 "epoch": 0.023709369024856597,440 "grad_norm": 5.248781153565317,441 "learning_rate": 5.751098106444967e-07,442 "loss": 0.3525,443 "step": 62444 },445 {446 "epoch": 0.024091778202676863,447 "grad_norm": 6.7358008908634925,448 "learning_rate": 5.77339433531346e-07,449 "loss": 0.6165,450 "step": 63451 },452 {453 "epoch": 0.024474187380497132,454 "grad_norm": 5.135947096269937,455 "learning_rate": 5.795339427837529e-07,456 "loss": 0.9469,457 "step": 64458 },459 {460 "epoch": 0.0248565965583174,461 "grad_norm": 4.721443670581346,462 "learning_rate": 5.816944272630579e-07,463 "loss": 1.0586,464 "step": 65465 },466 {467 "epoch": 0.025239005736137667,468 "grad_norm": 4.560383135778482,469 "learning_rate": 5.838219259551224e-07,470 "loss": 0.6226,471 "step": 66472 },473 {474 "epoch": 0.025621414913957936,475 "grad_norm": 3.0533395900689975,476 "learning_rate": 5.859174309706495e-07,477 "loss": 0.5508,478 "step": 67479 },480 {481 "epoch": 0.0260038240917782,482 "grad_norm": 4.38574534050627,483 "learning_rate": 5.879818903232327e-07,484 "loss": 0.6519,485 "step": 68486 },487 {488 "epoch": 0.02638623326959847,489 "grad_norm": 3.1760295850591347,490 "learning_rate": 5.900162105046059e-07,491 "loss": 0.7339,492 "step": 69493 },494 {495 "epoch": 0.02676864244741874,496 "grad_norm": 3.243632755108415,497 "learning_rate": 5.920212588746022e-07,498 "loss": 0.4918,499 "step": 70500 },501 {502 "epoch": 0.027151051625239005,503 "grad_norm": 3.432834024756851,504 "learning_rate": 5.939978658815961e-07,505 "loss": 0.5308,506 "step": 71507 },508 {509 "epoch": 0.027533460803059275,510 "grad_norm": 3.677125723417059,511 "learning_rate": 5.959468271276526e-07,512 "loss": 0.3372,513 "step": 72514 },515 {516 "epoch": 0.02791586998087954,517 "grad_norm": 2.8933168816281856,518 "learning_rate": 5.978689052912401e-07,519 "loss": 0.3834,520 "step": 73521 },522 {523 "epoch": 0.02829827915869981,524 "grad_norm": 4.377616318007501,525 "learning_rate": 5.997648319191317e-07,526 "loss": 0.285,527 "step": 74528 },529 {530 "epoch": 0.028680688336520075,531 "grad_norm": 5.329271657395167,532 "learning_rate": 6.016353090980351e-07,533 "loss": 0.4062,534 "step": 75535 },536 {537 "epoch": 0.029063097514340344,538 "grad_norm": 7.870238302614566,539 "learning_rate": 6.034810110155085e-07,540 "loss": 0.7765,541 "step": 76542 },543 {544 "epoch": 0.029445506692160613,545 "grad_norm": 4.2678637096949625,546 "learning_rate": 6.053025854188454e-07,547 "loss": 0.7735,548 "step": 77549 },550 {551 "epoch": 0.02982791586998088,552 "grad_norm": 3.691418909121812,553 "learning_rate": 6.071006549798314e-07,554 "loss": 0.7857,555 "step": 78556 },557 {558 "epoch": 0.030210325047801148,559 "grad_norm": 2.8591930560816983,560 "learning_rate": 6.088758185725654e-07,561 "loss": 0.7999,562 "step": 79563 },564 {565 "epoch": 0.030592734225621414,566 "grad_norm": 3.9392785793477474,567 "learning_rate": 6.106286524709088e-07,568 "loss": 0.7652,569 "step": 80570 },571 {572 "epoch": 0.030975143403441683,573 "grad_norm": 4.318070062532729,574 "learning_rate": 6.123597114715524e-07,575 "loss": 0.6564,576 "step": 81577 },578 {579 "epoch": 0.03135755258126195,580 "grad_norm": 3.853141414610464,581 "learning_rate": 6.140695299481785e-07,582 "loss": 0.6589,583 "step": 82584 },585 {586 "epoch": 0.03173996175908222,587 "grad_norm": 3.454451500377018,588 "learning_rate": 6.157586228417297e-07,589 "loss": 0.6253,590 "step": 83591 },592 {593 "epoch": 0.03212237093690248,594 "grad_norm": 3.244831082924636,595 "learning_rate": 6.174274865913755e-07,596 "loss": 0.2258,597 "step": 84598 },599 {600 "epoch": 0.032504780114722756,601 "grad_norm": 4.9841997360937835,602 "learning_rate": 6.190766000103886e-07,603 "loss": 0.4724,604 "step": 85605 },606 {607 "epoch": 0.03288718929254302,608 "grad_norm": 4.358109922986119,609 "learning_rate": 6.207064251107957e-07,610 "loss": 0.3161,611 "step": 86612 },613 {614 "epoch": 0.03326959847036329,615 "grad_norm": 3.752613174388839,616 "learning_rate": 6.223174078803584e-07,617 "loss": 0.2185,618 "step": 87619 },620 {621 "epoch": 0.03365200764818355,622 "grad_norm": 6.035992748966501,623 "learning_rate": 6.239099790151521e-07,624 "loss": 0.7716,625 "step": 88626 },627 {628 "epoch": 0.034034416826003826,629 "grad_norm": 5.726135487788577,630 "learning_rate": 6.254845546107528e-07,631 "loss": 1.1187,632 "step": 89633 },634 {635 "epoch": 0.03441682600382409,636 "grad_norm": 4.188566157565,637 "learning_rate": 6.270415368148085e-07,638 "loss": 0.6178,639 "step": 90640 },641 {642 "epoch": 0.03479923518164436,643 "grad_norm": 1.8973622357764504,644 "learning_rate": 6.285813144435542e-07,645 "loss": 0.82,646 "step": 91647 },648 {649 "epoch": 0.03518164435946463,650 "grad_norm": 3.203917465804253,651 "learning_rate": 6.301042635646355e-07,652 "loss": 0.7247,653 "step": 92654 },655 {656 "epoch": 0.035564053537284895,657 "grad_norm": 4.069173173828255,658 "learning_rate": 6.316107480484261e-07,659 "loss": 0.9489,660 "step": 93661 },662 {663 "epoch": 0.03594646271510516,664 "grad_norm": 5.255571224914232,665 "learning_rate": 6.331011200898622e-07,666 "loss": 0.3546,667 "step": 94668 },669 {670 "epoch": 0.036328871892925434,671 "grad_norm": 2.9958642426559994,672 "learning_rate": 6.345757207026644e-07,673 "loss": 0.5467,674 "step": 95675 },676 {677 "epoch": 0.0367112810707457,678 "grad_norm": 3.6217326204556106,679 "learning_rate": 6.360348801876822e-07,680 "loss": 0.2342,681 "step": 96682 },683 {684 "epoch": 0.037093690248565965,685 "grad_norm": 4.380818231179398,686 "learning_rate": 6.374789185769719e-07,687 "loss": 0.2475,688 "step": 97689 },690 {691 "epoch": 0.03747609942638623,692 "grad_norm": 3.792650297804044,693 "learning_rate": 6.389081460550985e-07,694 "loss": 0.3622,695 "step": 98696 },697 {698 "epoch": 0.0378585086042065,699 "grad_norm": 4.318067191395842,700 "learning_rate": 6.403228633590517e-07,701 "loss": 0.2876,702 "step": 99703 },704 {705 "epoch": 0.03824091778202677,706 "grad_norm": 4.3852408306664685,707 "learning_rate": 6.417233621580648e-07,708 "loss": 0.3416,709 "step": 100710 },711 {712 "epoch": 0.038623326959847035,713 "grad_norm": 5.55577400837663,714 "learning_rate": 6.431099254145344e-07,715 "loss": 0.6162,716 "step": 101717 },718 {719 "epoch": 0.03900573613766731,720 "grad_norm": 4.679812829276982,721 "learning_rate": 6.44482827727162e-07,722 "loss": 0.8162,723 "step": 102724 },725 {726 "epoch": 0.03938814531548757,727 "grad_norm": 4.147621318751239,728 "learning_rate": 6.458423356573556e-07,729 "loss": 1.1291,730 "step": 103731 },732 {733 "epoch": 0.03977055449330784,734 "grad_norm": 3.2625495994568445,735 "learning_rate": 6.471887080398609e-07,736 "loss": 0.6366,737 "step": 104738 },739 {740 "epoch": 0.040152963671128104,741 "grad_norm": 3.2876321396179984,742 "learning_rate": 6.485221962785314e-07,743 "loss": 0.877,744 "step": 105745 },746 {747 "epoch": 0.04053537284894838,748 "grad_norm": 3.51584336344417,749 "learning_rate": 6.498430446280784e-07,750 "loss": 0.4054,751 "step": 106752 },753 {754 "epoch": 0.04091778202676864,755 "grad_norm": 3.4273604229777384,756 "learning_rate": 6.511514904625936e-07,757 "loss": 0.4965,758 "step": 107759 },760 {761 "epoch": 0.04130019120458891,762 "grad_norm": 3.827336942561945,763 "learning_rate": 6.52447764531582e-07,764 "loss": 0.5168,765 "step": 108766 },767 {768 "epoch": 0.04168260038240918,769 "grad_norm": 3.217148613937907,770 "learning_rate": 6.537320912041942e-07,771 "loss": 0.5542,772 "step": 109773 },774 {775 "epoch": 0.04206500956022945,776 "grad_norm": 5.320029812779465,777 "learning_rate": 6.55004688702308e-07,778 "loss": 0.4252,779 "step": 110780 },781 {782 "epoch": 0.04244741873804971,783 "grad_norm": 3.4351211639757118,784 "learning_rate": 6.562657693230609e-07,785 "loss": 0.3598,786 "step": 111787 },788 {789 "epoch": 0.04282982791586998,790 "grad_norm": 4.734678353137203,791 "learning_rate": 6.575155396514051e-07,792 "loss": 0.3292,793 "step": 112794 },795 {796 "epoch": 0.04321223709369025,797 "grad_norm": 6.986725618584486,798 "learning_rate": 6.587542007632132e-07,799 "loss": 0.8562,800 "step": 113801 },802 {803 "epoch": 0.043594646271510516,804 "grad_norm": 4.767342479966072,805 "learning_rate": 6.599819484194378e-07,806 "loss": 0.9069,807 "step": 114808 },809 {810 "epoch": 0.04397705544933078,811 "grad_norm": 5.708434092428296,812 "learning_rate": 6.611989732517913e-07,813 "loss": 0.7799,814 "step": 115815 },816 {817 "epoch": 0.044359464627151055,818 "grad_norm": 2.143719184142411,819 "learning_rate": 6.62405460940388e-07,820 "loss": 0.762,821 "step": 116822 },823 {824 "epoch": 0.04474187380497132,825 "grad_norm": 3.075753164708183,826 "learning_rate": 6.636015923837607e-07,827 "loss": 0.6046,828 "step": 117829 },830 {831 "epoch": 0.045124282982791586,832 "grad_norm": 4.041382586271456,833 "learning_rate": 6.647875438616435e-07,834 "loss": 0.9243,835 "step": 118836 },837 {838 "epoch": 0.04550669216061185,839 "grad_norm": 4.0781175167555865,840 "learning_rate": 6.659634871908849e-07,841 "loss": 0.6291,842 "step": 119843 },844 {845 "epoch": 0.045889101338432124,846 "grad_norm": 2.947617658316482,847 "learning_rate": 6.671295898748382e-07,848 "loss": 0.6508,849 "step": 120850 },851 {852 "epoch": 0.04627151051625239,853 "grad_norm": 3.7795709081852134,854 "learning_rate": 6.682860152465512e-07,855 "loss": 0.4124,856 "step": 121857 },858 {859 "epoch": 0.046653919694072656,860 "grad_norm": 3.5289342257995586,861 "learning_rate": 6.694329226060651e-07,862 "loss": 0.4992,863 "step": 122864 },865 {866 "epoch": 0.04703632887189293,867 "grad_norm": 4.706878684962685,868 "learning_rate": 6.705704673521077e-07,869 "loss": 0.2933,870 "step": 123871 },872 {873 "epoch": 0.047418738049713194,874 "grad_norm": 5.110256294049169,875 "learning_rate": 6.716988011084555e-07,876 "loss": 0.4063,877 "step": 124878 },879 {880 "epoch": 0.04780114722753346,881 "grad_norm": 4.520100734904757,882 "learning_rate": 6.728180718452207e-07,883 "loss": 0.4,884 "step": 125885 },886 {887 "epoch": 0.048183556405353725,888 "grad_norm": 7.432639962375598,889 "learning_rate": 6.739284239953048e-07,890 "loss": 0.695,891 "step": 126892 },893 {894 "epoch": 0.048565965583174,895 "grad_norm": 4.290670013747284,896 "learning_rate": 6.750299985662519e-07,897 "loss": 0.4686,898 "step": 127899 },900 {901 "epoch": 0.048948374760994263,902 "grad_norm": 4.121158313697451,903 "learning_rate": 6.761229332477118e-07,904 "loss": 0.6365,905 "step": 128906 },907 {908 "epoch": 0.04933078393881453,909 "grad_norm": 2.135314479957041,910 "learning_rate": 6.772073625147248e-07,911 "loss": 0.6846,912 "step": 129913 },914 {915 "epoch": 0.0497131931166348,916 "grad_norm": 2.342096991950194,917 "learning_rate": 6.782834177270168e-07,918 "loss": 0.6111,919 "step": 130920 },921 {922 "epoch": 0.05009560229445507,923 "grad_norm": 3.647894547310256,924 "learning_rate": 6.793512272244895e-07,925 "loss": 0.4072,926 "step": 131927 },928 {929 "epoch": 0.05047801147227533,930 "grad_norm": 4.354125440971228,931 "learning_rate": 6.804109164190813e-07,932 "loss": 0.6969,933 "step": 132934 },935 {936 "epoch": 0.050860420650095606,937 "grad_norm": 2.7346670812193903,938 "learning_rate": 6.814626078831606e-07,939 "loss": 0.5829,940 "step": 133941 },942 {943 "epoch": 0.05124282982791587,944 "grad_norm": 4.094534063602115,945 "learning_rate": 6.825064214346084e-07,946 "loss": 0.3749,947 "step": 134948 },949 {950 "epoch": 0.05162523900573614,951 "grad_norm": 3.382758089469319,952 "learning_rate": 6.835424742187379e-07,953 "loss": 0.3332,954 "step": 135955 },956 {957 "epoch": 0.0520076481835564,958 "grad_norm": 2.683021984242027,959 "learning_rate": 6.845708807871916e-07,960 "loss": 0.1764,961 "step": 136962 },963 {964 "epoch": 0.052390057361376675,965 "grad_norm": 4.509963555368485,966 "learning_rate": 6.855917531739482e-07,967 "loss": 0.2143,968 "step": 137969 },970 {971 "epoch": 0.05277246653919694,972 "grad_norm": 5.256642816586635,973 "learning_rate": 6.866052009685647e-07,974 "loss": 0.4351,975 "step": 138976 },977 {978 "epoch": 0.05315487571701721,979 "grad_norm": 5.834538834758634,980 "learning_rate": 6.876113313867756e-07,981 "loss": 0.6942,982 "step": 139983 },984 {985 "epoch": 0.05353728489483748,986 "grad_norm": 3.014295690206189,987 "learning_rate": 6.88610249338561e-07,988 "loss": 0.4463,989 "step": 140990 },991 {992 "epoch": 0.053919694072657745,993 "grad_norm": 2.2857408607164182,994 "learning_rate": 6.896020574937915e-07,995 "loss": 0.5019,996 "step": 141997 },998 {999 "epoch": 0.05430210325047801,1000 "grad_norm": 2.8317586412436886,1001 "learning_rate": 6.905868563455549e-07,1002 "loss": 0.551,1003 "step": 1421004 },1005 {1006 "epoch": 0.054684512428298276,1007 "grad_norm": 3.590807962694632,1008 "learning_rate": 6.9156474427126e-07,1009 "loss": 0.5001,1010 "step": 1431011 },1012 {1013 "epoch": 0.05506692160611855,1014 "grad_norm": 3.3447864387888737,1015 "learning_rate": 6.925358175916115e-07,1016 "loss": 0.6405,1017 "step": 1441018 },1019 {1020 "epoch": 0.055449330783938815,1021 "grad_norm": 3.5812648007285097,1022 "learning_rate": 6.935001706275438e-07,1023 "loss": 0.3969,1024 "step": 1451025 },1026 {1027 "epoch": 0.05583173996175908,1028 "grad_norm": 2.588371307068287,1029 "learning_rate": 6.944578957551989e-07,1030 "loss": 0.4425,1031 "step": 1461032 },1033 {1034 "epoch": 0.05621414913957935,1035 "grad_norm": 3.8032063258431816,1036 "learning_rate": 6.954090834590278e-07,1037 "loss": 0.5201,1038 "step": 1471039 },1040 {1041 "epoch": 0.05659655831739962,1042 "grad_norm": 3.972727066954228,1043 "learning_rate": 6.963538223830905e-07,1044 "loss": 0.4198,1045 "step": 1481046 },1047 {1048 "epoch": 0.056978967495219884,1049 "grad_norm": 4.727081825163079,1050 "learning_rate": 6.972921993806316e-07,1051 "loss": 0.3914,1052 "step": 1491053 },1054 {1055 "epoch": 0.05736137667304015,1056 "grad_norm": 3.8566009806911032,1057 "learning_rate": 6.982242995619941e-07,1058 "loss": 0.3418,1059 "step": 1501060 },1061 {1062 "epoch": 0.05774378585086042,1063 "grad_norm": 9.883895151296194,1064 "learning_rate": 6.991502063409444e-07,1065 "loss": 0.4184,1066 "step": 1511067 },1068 {1069 "epoch": 0.05812619502868069,1070 "grad_norm": 3.9589817252241994,1071 "learning_rate": 7.000700014794673e-07,1072 "loss": 0.9268,1073 "step": 1521074 },1075 {1076 "epoch": 0.058508604206500954,1077 "grad_norm": 3.6920673081202664,1078 "learning_rate": 7.009837651310913e-07,1079 "loss": 0.7311,1080 "step": 1531081 },1082 {1083 "epoch": 0.05889101338432123,1084 "grad_norm": 2.8818112834107232,1085 "learning_rate": 7.018915758828043e-07,1086 "loss": 0.507,1087 "step": 1541088 },1089 {1090 "epoch": 0.05927342256214149,1091 "grad_norm": 4.095078580332929,1092 "learning_rate": 7.027935107956114e-07,1093 "loss": 0.6913,1094 "step": 1551095 },1096 {1097 "epoch": 0.05965583173996176,1098 "grad_norm": 4.097583635557744,1099 "learning_rate": 7.036896454437902e-07,1100 "loss": 0.7102,1101 "step": 1561102 },1103 {1104 "epoch": 0.060038240917782024,1105 "grad_norm": 2.6535932830808338,1106 "learning_rate": 7.045800539528896e-07,1107 "loss": 0.2475,1108 "step": 1571109 },1110 {1111 "epoch": 0.060420650095602296,1112 "grad_norm": 2.724522743078367,1113 "learning_rate": 7.054648090365243e-07,1114 "loss": 0.4209,1115 "step": 1581116 },1117 {1118 "epoch": 0.06080305927342256,1119 "grad_norm": 3.0125584007079067,1120 "learning_rate": 7.063439820320076e-07,1121 "loss": 0.4966,1122 "step": 1591123 },1124 {1125 "epoch": 0.06118546845124283,1126 "grad_norm": 3.511451787422904,1127 "learning_rate": 7.072176429348676e-07,1128 "loss": 0.4125,1129 "step": 1601130 },1131 {1132 "epoch": 0.0615678776290631,1133 "grad_norm": 4.837683638209349,1134 "learning_rate": 7.080858604322876e-07,1135 "loss": 0.6034,1136 "step": 1611137 },1138 {1139 "epoch": 0.061950286806883366,1140 "grad_norm": 3.7945186383087077,1141 "learning_rate": 7.089487019355111e-07,1142 "loss": 0.3376,1143 "step": 1621144 },1145 {1146 "epoch": 0.06233269598470363,1147 "grad_norm": 7.647899074910034,1148 "learning_rate": 7.098062336112514e-07,1149 "loss": 0.6145,1150 "step": 1631151 },1152 {1153 "epoch": 0.0627151051625239,1154 "grad_norm": 6.282933251606341,1155 "learning_rate": 7.106585204121373e-07,1156 "loss": 1.0463,1157 "step": 1641158 },1159 {1160 "epoch": 0.06309751434034416,1161 "grad_norm": 3.282497350571109,1162 "learning_rate": 7.115056261062372e-07,1163 "loss": 0.5178,1164 "step": 1651165 },1166 {1167 "epoch": 0.06347992351816444,1168 "grad_norm": 2.1352796433069017,1169 "learning_rate": 7.123476133056884e-07,1170 "loss": 0.6417,1171 "step": 1661172 },1173 {1174 "epoch": 0.06386233269598471,1175 "grad_norm": 2.5930365994625575,1176 "learning_rate": 7.131845434944682e-07,1177 "loss": 0.4984,1178 "step": 1671179 },1180 {1181 "epoch": 0.06424474187380497,1182 "grad_norm": 3.0331021971899714,1183 "learning_rate": 7.140164770553345e-07,1184 "loss": 0.454,1185 "step": 1681186 },1187 {1188 "epoch": 0.06462715105162524,1189 "grad_norm": 3.7143123017921016,1190 "learning_rate": 7.148434732959689e-07,1191 "loss": 0.5334,1192 "step": 1691193 },1194 {1195 "epoch": 0.06500956022944551,1196 "grad_norm": 2.653427329658854,1197 "learning_rate": 7.156655904743474e-07,1198 "loss": 0.3391,1199 "step": 1701200 },