| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.984301412872842, | |
| "eval_steps": 500, | |
| "global_step": 1590, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06279434850863422, | |
| "grad_norm": 0.7715030908584595, | |
| "learning_rate": 0.00019998048082884911, | |
| "loss": 1.3075, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.12558869701726844, | |
| "grad_norm": 0.7010838389396667, | |
| "learning_rate": 0.0001999219309353572, | |
| "loss": 1.0321, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.18838304552590268, | |
| "grad_norm": 0.5988379120826721, | |
| "learning_rate": 0.00019982437317643217, | |
| "loss": 0.9725, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.25117739403453687, | |
| "grad_norm": 0.5993463397026062, | |
| "learning_rate": 0.00019968784563700586, | |
| "loss": 0.9429, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3139717425431711, | |
| "grad_norm": 0.5684201121330261, | |
| "learning_rate": 0.0001995124016151664, | |
| "loss": 0.9221, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.37676609105180536, | |
| "grad_norm": 0.5648465752601624, | |
| "learning_rate": 0.00019929810960135172, | |
| "loss": 0.957, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.43956043956043955, | |
| "grad_norm": 0.5728427767753601, | |
| "learning_rate": 0.00019904505325161158, | |
| "loss": 0.8807, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.5023547880690737, | |
| "grad_norm": 0.5611833333969116, | |
| "learning_rate": 0.00019875333135495002, | |
| "loss": 0.8894, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.565149136577708, | |
| "grad_norm": 0.5440887808799744, | |
| "learning_rate": 0.00019842305779475968, | |
| "loss": 0.9077, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6279434850863422, | |
| "grad_norm": 0.5734382271766663, | |
| "learning_rate": 0.00019805436150436352, | |
| "loss": 0.872, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6907378335949764, | |
| "grad_norm": 0.5940066576004028, | |
| "learning_rate": 0.00019764738641668138, | |
| "loss": 0.8535, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7535321821036107, | |
| "grad_norm": 0.5004596710205078, | |
| "learning_rate": 0.0001972022914080411, | |
| "loss": 0.8571, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.8163265306122449, | |
| "grad_norm": 0.49201518297195435, | |
| "learning_rate": 0.00019671925023615573, | |
| "loss": 0.8738, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8791208791208791, | |
| "grad_norm": 0.5316975116729736, | |
| "learning_rate": 0.00019619845147229138, | |
| "loss": 0.8571, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9419152276295133, | |
| "grad_norm": 0.4743081033229828, | |
| "learning_rate": 0.00019564009842765225, | |
| "loss": 0.8552, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0047095761381475, | |
| "grad_norm": 0.488436758518219, | |
| "learning_rate": 0.0001950444090740111, | |
| "loss": 0.8686, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.0675039246467817, | |
| "grad_norm": 0.552605390548706, | |
| "learning_rate": 0.0001944116159586169, | |
| "loss": 0.7721, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.130298273155416, | |
| "grad_norm": 0.5751229524612427, | |
| "learning_rate": 0.0001937419661134121, | |
| "loss": 0.7748, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.1930926216640503, | |
| "grad_norm": 0.4798067808151245, | |
| "learning_rate": 0.00019303572095859546, | |
| "loss": 0.78, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.2558869701726845, | |
| "grad_norm": 0.5503143072128296, | |
| "learning_rate": 0.00019229315620056803, | |
| "loss": 0.7998, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3186813186813187, | |
| "grad_norm": 0.5721454620361328, | |
| "learning_rate": 0.00019151456172430183, | |
| "loss": 0.787, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.3814756671899528, | |
| "grad_norm": 0.5621303915977478, | |
| "learning_rate": 0.00019070024148017375, | |
| "loss": 0.7816, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.4442700156985873, | |
| "grad_norm": 0.5912421941757202, | |
| "learning_rate": 0.00018985051336530798, | |
| "loss": 0.7806, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.5070643642072215, | |
| "grad_norm": 0.5405421257019043, | |
| "learning_rate": 0.00018896570909947475, | |
| "loss": 0.7548, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.5698587127158556, | |
| "grad_norm": 0.47901952266693115, | |
| "learning_rate": 0.00018804617409559198, | |
| "loss": 0.7776, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.6326530612244898, | |
| "grad_norm": 0.5640778541564941, | |
| "learning_rate": 0.00018709226732488215, | |
| "loss": 0.7683, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.695447409733124, | |
| "grad_norm": 0.5678220987319946, | |
| "learning_rate": 0.00018610436117673555, | |
| "loss": 0.7773, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.7582417582417582, | |
| "grad_norm": 0.5400906801223755, | |
| "learning_rate": 0.00018508284131333602, | |
| "loss": 0.7797, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.8210361067503924, | |
| "grad_norm": 0.5115251541137695, | |
| "learning_rate": 0.00018402810651910442, | |
| "loss": 0.78, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.8838304552590266, | |
| "grad_norm": 0.5241789221763611, | |
| "learning_rate": 0.0001829405685450202, | |
| "loss": 0.7978, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.9466248037676608, | |
| "grad_norm": 0.540064811706543, | |
| "learning_rate": 0.00018182065194788025, | |
| "loss": 0.7754, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.009419152276295, | |
| "grad_norm": 0.5090541243553162, | |
| "learning_rate": 0.0001806687939245593, | |
| "loss": 0.7707, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.072213500784929, | |
| "grad_norm": 0.6285854578018188, | |
| "learning_rate": 0.00017948544414133534, | |
| "loss": 0.6989, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.1350078492935634, | |
| "grad_norm": 0.5556398630142212, | |
| "learning_rate": 0.00017827106455834733, | |
| "loss": 0.681, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.197802197802198, | |
| "grad_norm": 0.6137024760246277, | |
| "learning_rate": 0.00017702612924925376, | |
| "loss": 0.7002, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.260596546310832, | |
| "grad_norm": 0.5647903680801392, | |
| "learning_rate": 0.00017575112421616202, | |
| "loss": 0.6796, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.3233908948194664, | |
| "grad_norm": 0.5665746927261353, | |
| "learning_rate": 0.00017444654719990128, | |
| "loss": 0.7247, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.3861852433281006, | |
| "grad_norm": 0.5704404711723328, | |
| "learning_rate": 0.00017311290748571275, | |
| "loss": 0.6982, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.4489795918367347, | |
| "grad_norm": 0.6268453001976013, | |
| "learning_rate": 0.00017175072570443312, | |
| "loss": 0.6788, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.511773940345369, | |
| "grad_norm": 0.5893650054931641, | |
| "learning_rate": 0.00017036053362924896, | |
| "loss": 0.7071, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.574568288854003, | |
| "grad_norm": 0.5590313673019409, | |
| "learning_rate": 0.0001689428739681012, | |
| "loss": 0.6977, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.6373626373626373, | |
| "grad_norm": 0.5971296429634094, | |
| "learning_rate": 0.00016749830015182107, | |
| "loss": 0.7079, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.7001569858712715, | |
| "grad_norm": 0.60088050365448, | |
| "learning_rate": 0.00016602737611807976, | |
| "loss": 0.6985, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.7629513343799057, | |
| "grad_norm": 0.5298681855201721, | |
| "learning_rate": 0.00016453067609123654, | |
| "loss": 0.7271, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.82574568288854, | |
| "grad_norm": 0.6201738119125366, | |
| "learning_rate": 0.00016300878435817113, | |
| "loss": 0.6905, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.8885400313971745, | |
| "grad_norm": 0.605073869228363, | |
| "learning_rate": 0.00016146229504018775, | |
| "loss": 0.7071, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.9513343799058083, | |
| "grad_norm": 0.5479874610900879, | |
| "learning_rate": 0.00015989181186108002, | |
| "loss": 0.7233, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.014128728414443, | |
| "grad_norm": 0.5628566145896912, | |
| "learning_rate": 0.0001582979479114472, | |
| "loss": 0.6769, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.6384280920028687, | |
| "learning_rate": 0.0001566813254093538, | |
| "loss": 0.6165, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 3.1397174254317113, | |
| "grad_norm": 0.6049516797065735, | |
| "learning_rate": 0.00015504257545742584, | |
| "loss": 0.6414, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.2025117739403455, | |
| "grad_norm": 0.6833707690238953, | |
| "learning_rate": 0.0001533823377964791, | |
| "loss": 0.6559, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 3.2653061224489797, | |
| "grad_norm": 0.6257244944572449, | |
| "learning_rate": 0.00015170126055577462, | |
| "loss": 0.6465, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 3.328100470957614, | |
| "grad_norm": 0.6548301577568054, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.6332, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 3.390894819466248, | |
| "grad_norm": 0.6910219788551331, | |
| "learning_rate": 0.00014827922027307451, | |
| "loss": 0.6331, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 3.4536891679748822, | |
| "grad_norm": 0.5907390713691711, | |
| "learning_rate": 0.00014653959313887813, | |
| "loss": 0.6106, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.5164835164835164, | |
| "grad_norm": 0.6401733756065369, | |
| "learning_rate": 0.00014478179771900632, | |
| "loss": 0.642, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 3.5792778649921506, | |
| "grad_norm": 0.5502592325210571, | |
| "learning_rate": 0.00014300652022765207, | |
| "loss": 0.6392, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 3.642072213500785, | |
| "grad_norm": 0.6283037066459656, | |
| "learning_rate": 0.00014121445370371923, | |
| "loss": 0.612, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 3.704866562009419, | |
| "grad_norm": 0.6005645990371704, | |
| "learning_rate": 0.0001394062977402717, | |
| "loss": 0.6289, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 3.767660910518053, | |
| "grad_norm": 0.6629378199577332, | |
| "learning_rate": 0.00013758275821142382, | |
| "loss": 0.6281, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.830455259026688, | |
| "grad_norm": 0.6070135235786438, | |
| "learning_rate": 0.00013574454699677893, | |
| "loss": 0.6308, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 3.8932496075353216, | |
| "grad_norm": 0.6706437468528748, | |
| "learning_rate": 0.00013389238170352318, | |
| "loss": 0.6474, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 3.956043956043956, | |
| "grad_norm": 0.591901957988739, | |
| "learning_rate": 0.00013202698538628376, | |
| "loss": 0.6418, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 4.01883830455259, | |
| "grad_norm": 0.5693519711494446, | |
| "learning_rate": 0.0001301490862648603, | |
| "loss": 0.6092, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 4.081632653061225, | |
| "grad_norm": 0.6287245154380798, | |
| "learning_rate": 0.0001282594174399399, | |
| "loss": 0.5511, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 4.144427001569858, | |
| "grad_norm": 0.6552153825759888, | |
| "learning_rate": 0.00012635871660690676, | |
| "loss": 0.5828, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 4.207221350078493, | |
| "grad_norm": 0.664696216583252, | |
| "learning_rate": 0.00012444772576785827, | |
| "loss": 0.5643, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 4.270015698587127, | |
| "grad_norm": 0.6447893381118774, | |
| "learning_rate": 0.0001225271909419395, | |
| "loss": 0.5883, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 4.332810047095761, | |
| "grad_norm": 0.6885414123535156, | |
| "learning_rate": 0.00012059786187410984, | |
| "loss": 0.5826, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 4.395604395604396, | |
| "grad_norm": 0.608504593372345, | |
| "learning_rate": 0.00011866049174245491, | |
| "loss": 0.5499, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.45839874411303, | |
| "grad_norm": 0.6774938702583313, | |
| "learning_rate": 0.00011671583686415832, | |
| "loss": 0.5785, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 4.521193092621664, | |
| "grad_norm": 0.6806432604789734, | |
| "learning_rate": 0.00011476465640024814, | |
| "loss": 0.5678, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 4.583987441130298, | |
| "grad_norm": 0.7246624827384949, | |
| "learning_rate": 0.00011280771205923268, | |
| "loss": 0.5797, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 4.646781789638933, | |
| "grad_norm": 0.6730565428733826, | |
| "learning_rate": 0.00011084576779974257, | |
| "loss": 0.565, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 4.7095761381475665, | |
| "grad_norm": 0.6672644019126892, | |
| "learning_rate": 0.00010887958953229349, | |
| "loss": 0.5624, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.772370486656201, | |
| "grad_norm": 0.7056964039802551, | |
| "learning_rate": 0.0001069099448202878, | |
| "loss": 0.587, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 4.835164835164835, | |
| "grad_norm": 0.7322887182235718, | |
| "learning_rate": 0.00010493760258037031, | |
| "loss": 0.5712, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 4.8979591836734695, | |
| "grad_norm": 0.7399300336837769, | |
| "learning_rate": 0.00010296333278225599, | |
| "loss": 0.5842, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 4.960753532182103, | |
| "grad_norm": 0.6447321176528931, | |
| "learning_rate": 0.00010098790614814658, | |
| "loss": 0.559, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 5.023547880690738, | |
| "grad_norm": 0.7017489075660706, | |
| "learning_rate": 9.901209385185345e-05, | |
| "loss": 0.5439, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.086342229199372, | |
| "grad_norm": 0.7099691033363342, | |
| "learning_rate": 9.703666721774402e-05, | |
| "loss": 0.5144, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 5.149136577708006, | |
| "grad_norm": 0.6762498617172241, | |
| "learning_rate": 9.506239741962971e-05, | |
| "loss": 0.5159, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 5.211930926216641, | |
| "grad_norm": 0.722254753112793, | |
| "learning_rate": 9.309005517971222e-05, | |
| "loss": 0.5174, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 5.274725274725275, | |
| "grad_norm": 0.7049859166145325, | |
| "learning_rate": 9.112041046770653e-05, | |
| "loss": 0.5067, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 5.337519623233909, | |
| "grad_norm": 0.6614891886711121, | |
| "learning_rate": 8.915423220025747e-05, | |
| "loss": 0.5178, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 5.400313971742543, | |
| "grad_norm": 0.7367340326309204, | |
| "learning_rate": 8.719228794076733e-05, | |
| "loss": 0.5158, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 5.463108320251178, | |
| "grad_norm": 0.7544198632240295, | |
| "learning_rate": 8.523534359975189e-05, | |
| "loss": 0.5181, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 5.525902668759811, | |
| "grad_norm": 0.7748332023620605, | |
| "learning_rate": 8.328416313584169e-05, | |
| "loss": 0.5211, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 5.588697017268446, | |
| "grad_norm": 0.7265179753303528, | |
| "learning_rate": 8.13395082575451e-05, | |
| "loss": 0.5005, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 5.65149136577708, | |
| "grad_norm": 0.7204436659812927, | |
| "learning_rate": 7.940213812589018e-05, | |
| "loss": 0.5034, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 0.7397501468658447, | |
| "learning_rate": 7.747280905806052e-05, | |
| "loss": 0.5204, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 5.777080062794348, | |
| "grad_norm": 0.7541511654853821, | |
| "learning_rate": 7.555227423214174e-05, | |
| "loss": 0.4947, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 5.839874411302983, | |
| "grad_norm": 0.7153120636940002, | |
| "learning_rate": 7.364128339309326e-05, | |
| "loss": 0.5189, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 5.9026687598116165, | |
| "grad_norm": 0.7152422070503235, | |
| "learning_rate": 7.174058256006012e-05, | |
| "loss": 0.5254, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 5.965463108320251, | |
| "grad_norm": 0.7143374681472778, | |
| "learning_rate": 6.985091373513972e-05, | |
| "loss": 0.5044, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 6.028257456828886, | |
| "grad_norm": 0.6690304279327393, | |
| "learning_rate": 6.797301461371625e-05, | |
| "loss": 0.482, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 6.0910518053375196, | |
| "grad_norm": 0.7355020642280579, | |
| "learning_rate": 6.610761829647685e-05, | |
| "loss": 0.4561, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 6.153846153846154, | |
| "grad_norm": 0.7926068902015686, | |
| "learning_rate": 6.425545300322112e-05, | |
| "loss": 0.4708, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 6.216640502354788, | |
| "grad_norm": 0.7191044688224792, | |
| "learning_rate": 6.24172417885762e-05, | |
| "loss": 0.4586, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 6.279434850863423, | |
| "grad_norm": 0.7463113069534302, | |
| "learning_rate": 6.0593702259728336e-05, | |
| "loss": 0.4674, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.342229199372056, | |
| "grad_norm": 0.743307888507843, | |
| "learning_rate": 5.8785546296280816e-05, | |
| "loss": 0.471, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 6.405023547880691, | |
| "grad_norm": 0.7751266956329346, | |
| "learning_rate": 5.699347977234799e-05, | |
| "loss": 0.4655, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 6.467817896389325, | |
| "grad_norm": 0.8425696492195129, | |
| "learning_rate": 5.5218202280993725e-05, | |
| "loss": 0.476, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 6.530612244897959, | |
| "grad_norm": 0.778333842754364, | |
| "learning_rate": 5.3460406861121894e-05, | |
| "loss": 0.4667, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 6.593406593406593, | |
| "grad_norm": 0.7623078227043152, | |
| "learning_rate": 5.172077972692553e-05, | |
| "loss": 0.4799, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 6.656200941915228, | |
| "grad_norm": 0.7711922526359558, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 0.4621, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 6.718995290423862, | |
| "grad_norm": 0.812324583530426, | |
| "learning_rate": 4.829873944422544e-05, | |
| "loss": 0.4612, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 6.781789638932496, | |
| "grad_norm": 0.7873876094818115, | |
| "learning_rate": 4.661766220352097e-05, | |
| "loss": 0.4794, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 6.84458398744113, | |
| "grad_norm": 0.8011544942855835, | |
| "learning_rate": 4.495742454257418e-05, | |
| "loss": 0.4608, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 6.9073783359497645, | |
| "grad_norm": 0.77611243724823, | |
| "learning_rate": 4.3318674590646237e-05, | |
| "loss": 0.471, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 6.970172684458399, | |
| "grad_norm": 0.8160433769226074, | |
| "learning_rate": 4.170205208855281e-05, | |
| "loss": 0.4717, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 7.032967032967033, | |
| "grad_norm": 0.7271689176559448, | |
| "learning_rate": 4.010818813892e-05, | |
| "loss": 0.4322, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 7.0957613814756675, | |
| "grad_norm": 0.8266540765762329, | |
| "learning_rate": 3.8537704959812294e-05, | |
| "loss": 0.4184, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 7.158555729984301, | |
| "grad_norm": 0.8125666975975037, | |
| "learning_rate": 3.69912156418289e-05, | |
| "loss": 0.4284, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 7.221350078492936, | |
| "grad_norm": 0.839213490486145, | |
| "learning_rate": 3.546932390876351e-05, | |
| "loss": 0.4389, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 7.28414442700157, | |
| "grad_norm": 0.89272141456604, | |
| "learning_rate": 3.397262388192029e-05, | |
| "loss": 0.4477, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 7.346938775510204, | |
| "grad_norm": 0.7432073950767517, | |
| "learning_rate": 3.250169984817897e-05, | |
| "loss": 0.43, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 7.409733124018838, | |
| "grad_norm": 0.7860047817230225, | |
| "learning_rate": 3.105712603189884e-05, | |
| "loss": 0.4269, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 7.472527472527473, | |
| "grad_norm": 0.832013726234436, | |
| "learning_rate": 2.9639466370751068e-05, | |
| "loss": 0.4256, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 7.535321821036106, | |
| "grad_norm": 0.7955825924873352, | |
| "learning_rate": 2.8249274295566864e-05, | |
| "loss": 0.4282, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 7.598116169544741, | |
| "grad_norm": 0.7972998023033142, | |
| "learning_rate": 2.688709251428725e-05, | |
| "loss": 0.4388, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 7.660910518053376, | |
| "grad_norm": 0.8105794191360474, | |
| "learning_rate": 2.555345280009872e-05, | |
| "loss": 0.4381, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 7.723704866562009, | |
| "grad_norm": 0.7821062207221985, | |
| "learning_rate": 2.4248875783837987e-05, | |
| "loss": 0.4276, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 7.786499215070644, | |
| "grad_norm": 0.805131733417511, | |
| "learning_rate": 2.297387075074625e-05, | |
| "loss": 0.4307, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 7.849293563579278, | |
| "grad_norm": 0.797322690486908, | |
| "learning_rate": 2.1728935441652686e-05, | |
| "loss": 0.4295, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 7.912087912087912, | |
| "grad_norm": 0.7943819761276245, | |
| "learning_rate": 2.0514555858664663e-05, | |
| "loss": 0.4346, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 7.974882260596546, | |
| "grad_norm": 0.814616322517395, | |
| "learning_rate": 1.93312060754407e-05, | |
| "loss": 0.4331, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 8.03767660910518, | |
| "grad_norm": 0.79474937915802, | |
| "learning_rate": 1.817934805211976e-05, | |
| "loss": 0.4153, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 8.100470957613815, | |
| "grad_norm": 0.8072588443756104, | |
| "learning_rate": 1.7059431454979824e-05, | |
| "loss": 0.4072, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 8.16326530612245, | |
| "grad_norm": 0.8833736777305603, | |
| "learning_rate": 1.5971893480895583e-05, | |
| "loss": 0.3996, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 8.226059654631083, | |
| "grad_norm": 0.8510453104972839, | |
| "learning_rate": 1.491715868666399e-05, | |
| "loss": 0.4132, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 8.288854003139717, | |
| "grad_norm": 0.8392331004142761, | |
| "learning_rate": 1.3895638823264446e-05, | |
| "loss": 0.4181, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 8.351648351648352, | |
| "grad_norm": 0.7935534119606018, | |
| "learning_rate": 1.290773267511788e-05, | |
| "loss": 0.3973, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 8.414442700156986, | |
| "grad_norm": 0.821882963180542, | |
| "learning_rate": 1.1953825904408034e-05, | |
| "loss": 0.4069, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 8.47723704866562, | |
| "grad_norm": 0.8700546026229858, | |
| "learning_rate": 1.103429090052528e-05, | |
| "loss": 0.4144, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 8.540031397174253, | |
| "grad_norm": 0.822404146194458, | |
| "learning_rate": 1.0149486634692018e-05, | |
| "loss": 0.4052, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 8.602825745682889, | |
| "grad_norm": 0.8342509269714355, | |
| "learning_rate": 9.299758519826273e-06, | |
| "loss": 0.4076, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 8.665620094191523, | |
| "grad_norm": 0.8416010141372681, | |
| "learning_rate": 8.485438275698154e-06, | |
| "loss": 0.4161, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 8.728414442700156, | |
| "grad_norm": 0.7646155953407288, | |
| "learning_rate": 7.706843799431984e-06, | |
| "loss": 0.4153, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 8.791208791208792, | |
| "grad_norm": 0.7787067294120789, | |
| "learning_rate": 6.964279041404553e-06, | |
| "loss": 0.3995, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 8.854003139717426, | |
| "grad_norm": 0.9011712074279785, | |
| "learning_rate": 6.258033886587911e-06, | |
| "loss": 0.4144, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 8.91679748822606, | |
| "grad_norm": 0.8679221272468567, | |
| "learning_rate": 5.588384041383088e-06, | |
| "loss": 0.4147, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 8.979591836734693, | |
| "grad_norm": 0.8739622831344604, | |
| "learning_rate": 4.955590925988895e-06, | |
| "loss": 0.4011, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 9.042386185243329, | |
| "grad_norm": 0.824476420879364, | |
| "learning_rate": 4.359901572347758e-06, | |
| "loss": 0.4138, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 9.105180533751962, | |
| "grad_norm": 0.8952732682228088, | |
| "learning_rate": 3.8015485277086205e-06, | |
| "loss": 0.4088, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 9.167974882260596, | |
| "grad_norm": 0.7407020330429077, | |
| "learning_rate": 3.280749763844293e-06, | |
| "loss": 0.4018, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 9.23076923076923, | |
| "grad_norm": 0.8719793558120728, | |
| "learning_rate": 2.7977085919589254e-06, | |
| "loss": 0.3925, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 9.293563579277865, | |
| "grad_norm": 0.8144113421440125, | |
| "learning_rate": 2.3526135833186525e-06, | |
| "loss": 0.4012, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 9.3563579277865, | |
| "grad_norm": 0.8647878170013428, | |
| "learning_rate": 1.945638495636515e-06, | |
| "loss": 0.4006, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 9.419152276295133, | |
| "grad_norm": 0.827995240688324, | |
| "learning_rate": 1.576942205240317e-06, | |
| "loss": 0.3984, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 9.481946624803768, | |
| "grad_norm": 0.8608419895172119, | |
| "learning_rate": 1.2466686450499865e-06, | |
| "loss": 0.3956, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 9.544740973312402, | |
| "grad_norm": 0.8144489526748657, | |
| "learning_rate": 9.549467483884412e-07, | |
| "loss": 0.3967, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 9.607535321821036, | |
| "grad_norm": 0.906897783279419, | |
| "learning_rate": 7.018903986483083e-07, | |
| "loss": 0.4039, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 9.67032967032967, | |
| "grad_norm": 0.8140434622764587, | |
| "learning_rate": 4.875983848335874e-07, | |
| "loss": 0.3924, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 9.733124018838305, | |
| "grad_norm": 0.8026575446128845, | |
| "learning_rate": 3.12154362994177e-07, | |
| "loss": 0.4028, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 9.795918367346939, | |
| "grad_norm": 0.8668120503425598, | |
| "learning_rate": 1.7562682356786487e-07, | |
| "loss": 0.404, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 9.858712715855573, | |
| "grad_norm": 0.8131958842277527, | |
| "learning_rate": 7.806906464281615e-08, | |
| "loss": 0.4016, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 9.921507064364206, | |
| "grad_norm": 0.820686399936676, | |
| "learning_rate": 1.951917115091684e-08, | |
| "loss": 0.3993, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 9.984301412872842, | |
| "grad_norm": 0.8288590312004089, | |
| "learning_rate": 0.0, | |
| "loss": 0.389, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 9.984301412872842, | |
| "step": 1590, | |
| "total_flos": 1.04418647212032e+17, | |
| "train_loss": 0.5835858842861728, | |
| "train_runtime": 7264.6263, | |
| "train_samples_per_second": 0.877, | |
| "train_steps_per_second": 0.219 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1590, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 1.04418647212032e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |