| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 888888, | |
| "global_step": 308, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.012987012987012988, | |
| "grad_norm": 30.409713806756372, | |
| "learning_rate": 4.999479820473016e-07, | |
| "logits/chosen": -1.1957358121871948, | |
| "logits/rejected": -1.1989657878875732, | |
| "logps/chosen": -236.6458282470703, | |
| "logps/rejected": -274.4889221191406, | |
| "loss": 0.7214, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.9376516938209534, | |
| "rewards/margins": 0.12735898792743683, | |
| "rewards/rejected": 0.810292661190033, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.025974025974025976, | |
| "grad_norm": 32.07895252539544, | |
| "learning_rate": 4.997919498361457e-07, | |
| "logits/chosen": -1.2326594591140747, | |
| "logits/rejected": -1.2664812803268433, | |
| "logps/chosen": -257.2532958984375, | |
| "logps/rejected": -252.87025451660156, | |
| "loss": 0.727, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 1.142688512802124, | |
| "rewards/margins": 0.25847145915031433, | |
| "rewards/rejected": 0.8842170238494873, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03896103896103896, | |
| "grad_norm": 31.29868726835709, | |
| "learning_rate": 4.995319682983417e-07, | |
| "logits/chosen": -1.183756947517395, | |
| "logits/rejected": -1.1092337369918823, | |
| "logps/chosen": -191.75558471679688, | |
| "logps/rejected": -230.58950805664062, | |
| "loss": 0.7031, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": 0.7358960509300232, | |
| "rewards/margins": -0.0961524099111557, | |
| "rewards/rejected": 0.8320484757423401, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.05194805194805195, | |
| "grad_norm": 32.01008811759358, | |
| "learning_rate": 4.991681456235483e-07, | |
| "logits/chosen": -1.2155265808105469, | |
| "logits/rejected": -1.3162301778793335, | |
| "logps/chosen": -202.1005859375, | |
| "logps/rejected": -232.53140258789062, | |
| "loss": 0.6947, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 1.0601692199707031, | |
| "rewards/margins": 0.3322027921676636, | |
| "rewards/rejected": 0.72796630859375, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.06493506493506493, | |
| "grad_norm": 26.584702872904153, | |
| "learning_rate": 4.98700633214251e-07, | |
| "logits/chosen": -1.08168363571167, | |
| "logits/rejected": -1.21769118309021, | |
| "logps/chosen": -217.7674560546875, | |
| "logps/rejected": -195.05032348632812, | |
| "loss": 0.6823, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.8856821060180664, | |
| "rewards/margins": 0.06800435483455658, | |
| "rewards/rejected": 0.8176776170730591, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07792207792207792, | |
| "grad_norm": 31.831697368646147, | |
| "learning_rate": 4.981296256227569e-07, | |
| "logits/chosen": -0.912403404712677, | |
| "logits/rejected": -0.9036965370178223, | |
| "logps/chosen": -204.28919982910156, | |
| "logps/rejected": -211.12686157226562, | |
| "loss": 0.7049, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.849972665309906, | |
| "rewards/margins": 0.017677173018455505, | |
| "rewards/rejected": 0.8322954177856445, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 29.9192585773342, | |
| "learning_rate": 4.974553604702332e-07, | |
| "logits/chosen": -1.0211938619613647, | |
| "logits/rejected": -1.0898373126983643, | |
| "logps/chosen": -166.66610717773438, | |
| "logps/rejected": -154.6410675048828, | |
| "loss": 0.6707, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 1.0837771892547607, | |
| "rewards/margins": 0.4414757192134857, | |
| "rewards/rejected": 0.6423014402389526, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.1038961038961039, | |
| "grad_norm": 29.446896480009652, | |
| "learning_rate": 4.966781183478222e-07, | |
| "logits/chosen": -1.1182293891906738, | |
| "logits/rejected": -1.116457223892212, | |
| "logps/chosen": -201.10256958007812, | |
| "logps/rejected": -219.7225341796875, | |
| "loss": 0.6796, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.78659588098526, | |
| "rewards/margins": 0.23196566104888916, | |
| "rewards/rejected": 0.5546301603317261, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.11688311688311688, | |
| "grad_norm": 30.21023973321977, | |
| "learning_rate": 4.957982226998757e-07, | |
| "logits/chosen": -1.31354820728302, | |
| "logits/rejected": -1.2536346912384033, | |
| "logps/chosen": -265.8379821777344, | |
| "logps/rejected": -281.1967468261719, | |
| "loss": 0.7288, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 1.1193804740905762, | |
| "rewards/margins": 0.1973106861114502, | |
| "rewards/rejected": 0.922069787979126, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.12987012987012986, | |
| "grad_norm": 30.94898503765605, | |
| "learning_rate": 4.948160396893553e-07, | |
| "logits/chosen": -1.2565226554870605, | |
| "logits/rejected": -1.15552818775177, | |
| "logps/chosen": -174.21267700195312, | |
| "logps/rejected": -179.50820922851562, | |
| "loss": 0.6927, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.931135892868042, | |
| "rewards/margins": 0.1621209681034088, | |
| "rewards/rejected": 0.7690149545669556, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 29.667041326907114, | |
| "learning_rate": 4.937319780454559e-07, | |
| "logits/chosen": -1.3007498979568481, | |
| "logits/rejected": -1.1949548721313477, | |
| "logps/chosen": -230.96115112304688, | |
| "logps/rejected": -248.40798950195312, | |
| "loss": 0.7216, | |
| "rewards/accuracies": 0.34375, | |
| "rewards/chosen": 0.7717787027359009, | |
| "rewards/margins": -0.18773330748081207, | |
| "rewards/rejected": 0.9595120549201965, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.15584415584415584, | |
| "grad_norm": 33.222954526021944, | |
| "learning_rate": 4.925464888935161e-07, | |
| "logits/chosen": -1.1974416971206665, | |
| "logits/rejected": -1.2200247049331665, | |
| "logps/chosen": -198.02139282226562, | |
| "logps/rejected": -242.11419677734375, | |
| "loss": 0.7295, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.8105030655860901, | |
| "rewards/margins": 0.041422925889492035, | |
| "rewards/rejected": 0.7690802216529846, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.16883116883116883, | |
| "grad_norm": 31.901066989434995, | |
| "learning_rate": 4.912600655672849e-07, | |
| "logits/chosen": -1.240085482597351, | |
| "logits/rejected": -1.2631834745407104, | |
| "logps/chosen": -264.38067626953125, | |
| "logps/rejected": -317.8161926269531, | |
| "loss": 0.6988, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 1.0404579639434814, | |
| "rewards/margins": 0.05802140012383461, | |
| "rewards/rejected": 0.9824365377426147, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 34.21159593307014, | |
| "learning_rate": 4.898732434036243e-07, | |
| "logits/chosen": -1.189288854598999, | |
| "logits/rejected": -1.048648476600647, | |
| "logps/chosen": -262.818115234375, | |
| "logps/rejected": -301.31927490234375, | |
| "loss": 0.7616, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": 0.8156523108482361, | |
| "rewards/margins": -0.08142556995153427, | |
| "rewards/rejected": 0.8970779180526733, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.19480519480519481, | |
| "grad_norm": 32.29301620044956, | |
| "learning_rate": 4.883865995197318e-07, | |
| "logits/chosen": -1.1413640975952148, | |
| "logits/rejected": -1.2122159004211426, | |
| "logps/chosen": -217.71823120117188, | |
| "logps/rejected": -221.3170166015625, | |
| "loss": 0.7261, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 1.1059789657592773, | |
| "rewards/margins": 0.40076321363449097, | |
| "rewards/rejected": 0.7052158713340759, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2077922077922078, | |
| "grad_norm": 30.16619750369487, | |
| "learning_rate": 4.868007525729775e-07, | |
| "logits/chosen": -1.36598801612854, | |
| "logits/rejected": -1.3407963514328003, | |
| "logps/chosen": -239.17111206054688, | |
| "logps/rejected": -226.31658935546875, | |
| "loss": 0.6825, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.7918877005577087, | |
| "rewards/margins": 0.05119353532791138, | |
| "rewards/rejected": 0.7406941056251526, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.22077922077922077, | |
| "grad_norm": 30.605215606478595, | |
| "learning_rate": 4.851163625034529e-07, | |
| "logits/chosen": -1.1574623584747314, | |
| "logits/rejected": -1.193921685218811, | |
| "logps/chosen": -230.02108764648438, | |
| "logps/rejected": -244.59841918945312, | |
| "loss": 0.6851, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 1.080081820487976, | |
| "rewards/margins": 0.3703171908855438, | |
| "rewards/rejected": 0.7097645998001099, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.23376623376623376, | |
| "grad_norm": 31.802179808840666, | |
| "learning_rate": 4.833341302593417e-07, | |
| "logits/chosen": -1.159416913986206, | |
| "logits/rejected": -1.1420115232467651, | |
| "logps/chosen": -140.0688018798828, | |
| "logps/rejected": -152.40975952148438, | |
| "loss": 0.6732, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.8693720102310181, | |
| "rewards/margins": 0.11663462221622467, | |
| "rewards/rejected": 0.7527374625205994, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.24675324675324675, | |
| "grad_norm": 30.94540714928562, | |
| "learning_rate": 4.814547975052244e-07, | |
| "logits/chosen": -1.3302018642425537, | |
| "logits/rejected": -1.3434113264083862, | |
| "logps/chosen": -191.34698486328125, | |
| "logps/rejected": -196.69509887695312, | |
| "loss": 0.6714, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": 0.7995274662971497, | |
| "rewards/margins": 0.047811880707740784, | |
| "rewards/rejected": 0.7517155408859253, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "grad_norm": 29.865098693610847, | |
| "learning_rate": 4.794791463134399e-07, | |
| "logits/chosen": -1.2271629571914673, | |
| "logits/rejected": -1.3550831079483032, | |
| "logps/chosen": -226.42410278320312, | |
| "logps/rejected": -214.7489013671875, | |
| "loss": 0.6975, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.8646127581596375, | |
| "rewards/margins": 0.22612789273262024, | |
| "rewards/rejected": 0.6384848356246948, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 30.43675432858084, | |
| "learning_rate": 4.774079988386296e-07, | |
| "logits/chosen": -1.1719427108764648, | |
| "logits/rejected": -1.2684495449066162, | |
| "logps/chosen": -208.20394897460938, | |
| "logps/rejected": -235.50634765625, | |
| "loss": 0.6855, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 1.0848238468170166, | |
| "rewards/margins": 0.3408503830432892, | |
| "rewards/rejected": 0.7439733743667603, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 31.511971354878987, | |
| "learning_rate": 4.752422169756047e-07, | |
| "logits/chosen": -1.2040029764175415, | |
| "logits/rejected": -1.2218215465545654, | |
| "logps/chosen": -259.76959228515625, | |
| "logps/rejected": -279.75946044921875, | |
| "loss": 0.6772, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.9812463521957397, | |
| "rewards/margins": 0.2615824341773987, | |
| "rewards/rejected": 0.7196639776229858, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.2987012987012987, | |
| "grad_norm": 35.50720072355199, | |
| "learning_rate": 4.729827020006735e-07, | |
| "logits/chosen": -1.2218239307403564, | |
| "logits/rejected": -1.2937262058258057, | |
| "logps/chosen": -239.9827880859375, | |
| "logps/rejected": -250.3217315673828, | |
| "loss": 0.7242, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.8138805627822876, | |
| "rewards/margins": 0.09423254430294037, | |
| "rewards/rejected": 0.719648003578186, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3116883116883117, | |
| "grad_norm": 32.207699613934025, | |
| "learning_rate": 4.706303941965803e-07, | |
| "logits/chosen": -1.174890398979187, | |
| "logits/rejected": -1.1500227451324463, | |
| "logps/chosen": -257.21533203125, | |
| "logps/rejected": -302.0556945800781, | |
| "loss": 0.6856, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 1.0188751220703125, | |
| "rewards/margins": 0.20580947399139404, | |
| "rewards/rejected": 0.8130655884742737, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3246753246753247, | |
| "grad_norm": 31.074659437805654, | |
| "learning_rate": 4.68186272461214e-07, | |
| "logits/chosen": -1.2761106491088867, | |
| "logits/rejected": -1.307037353515625, | |
| "logps/chosen": -228.11666870117188, | |
| "logps/rejected": -254.00048828125, | |
| "loss": 0.6711, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.9066404104232788, | |
| "rewards/margins": 0.32165050506591797, | |
| "rewards/rejected": 0.5849898457527161, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.33766233766233766, | |
| "grad_norm": 31.027447722981428, | |
| "learning_rate": 4.656513539002451e-07, | |
| "logits/chosen": -1.2342908382415771, | |
| "logits/rejected": -1.2500003576278687, | |
| "logps/chosen": -182.6434783935547, | |
| "logps/rejected": -210.83444213867188, | |
| "loss": 0.6867, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.8957956433296204, | |
| "rewards/margins": 0.21347832679748535, | |
| "rewards/rejected": 0.6823172569274902, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.35064935064935066, | |
| "grad_norm": 28.860294980547383, | |
| "learning_rate": 4.6302669340386415e-07, | |
| "logits/chosen": -1.2332104444503784, | |
| "logits/rejected": -1.1100932359695435, | |
| "logps/chosen": -194.90341186523438, | |
| "logps/rejected": -220.88478088378906, | |
| "loss": 0.692, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.7950139045715332, | |
| "rewards/margins": 0.09959140419960022, | |
| "rewards/rejected": 0.6954225897789001, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 37.00612209924237, | |
| "learning_rate": 4.603133832077953e-07, | |
| "logits/chosen": -1.2518322467803955, | |
| "logits/rejected": -1.272511601448059, | |
| "logps/chosen": -219.75762939453125, | |
| "logps/rejected": -225.9881591796875, | |
| "loss": 0.7341, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.8070573806762695, | |
| "rewards/margins": 0.15917816758155823, | |
| "rewards/rejected": 0.6478793025016785, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.37662337662337664, | |
| "grad_norm": 34.181154948444195, | |
| "learning_rate": 4.575125524387701e-07, | |
| "logits/chosen": -1.0896148681640625, | |
| "logits/rejected": -1.0761992931365967, | |
| "logps/chosen": -246.28172302246094, | |
| "logps/rejected": -231.1036376953125, | |
| "loss": 0.6999, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.6937105059623718, | |
| "rewards/margins": 0.09433312714099884, | |
| "rewards/rejected": 0.5993773937225342, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.38961038961038963, | |
| "grad_norm": 33.21940371087566, | |
| "learning_rate": 4.5462536664464836e-07, | |
| "logits/chosen": -1.2278883457183838, | |
| "logits/rejected": -1.2770620584487915, | |
| "logps/chosen": -325.75714111328125, | |
| "logps/rejected": -309.88836669921875, | |
| "loss": 0.685, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.9801526069641113, | |
| "rewards/margins": 0.22209185361862183, | |
| "rewards/rejected": 0.7580606937408447, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.4025974025974026, | |
| "grad_norm": 29.951734183402802, | |
| "learning_rate": 4.516530273093825e-07, | |
| "logits/chosen": -1.3827568292617798, | |
| "logits/rejected": -1.3074092864990234, | |
| "logps/chosen": -226.56942749023438, | |
| "logps/rejected": -261.1898498535156, | |
| "loss": 0.6715, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.8671592473983765, | |
| "rewards/margins": 0.1708277314901352, | |
| "rewards/rejected": 0.6963315010070801, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4155844155844156, | |
| "grad_norm": 35.00256057592286, | |
| "learning_rate": 4.485967713530281e-07, | |
| "logits/chosen": -1.2592735290527344, | |
| "logits/rejected": -1.276845097541809, | |
| "logps/chosen": -280.13421630859375, | |
| "logps/rejected": -272.318359375, | |
| "loss": 0.6797, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.9574460983276367, | |
| "rewards/margins": 0.2645747661590576, | |
| "rewards/rejected": 0.6928713321685791, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 130.44143506253818, | |
| "learning_rate": 4.4545787061700746e-07, | |
| "logits/chosen": -1.201524019241333, | |
| "logits/rejected": -1.2254564762115479, | |
| "logps/chosen": -222.44644165039062, | |
| "logps/rejected": -215.9447021484375, | |
| "loss": 0.6865, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.8678699731826782, | |
| "rewards/margins": 0.32531124353408813, | |
| "rewards/rejected": 0.5425587296485901, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.44155844155844154, | |
| "grad_norm": 30.886647455965967, | |
| "learning_rate": 4.422376313348405e-07, | |
| "logits/chosen": -1.2102560997009277, | |
| "logits/rejected": -1.1355280876159668, | |
| "logps/chosen": -211.4716796875, | |
| "logps/rejected": -246.92666625976562, | |
| "loss": 0.6728, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.9252184629440308, | |
| "rewards/margins": 0.4986411929130554, | |
| "rewards/rejected": 0.4265773296356201, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 38.68657958814148, | |
| "learning_rate": 4.3893739358856455e-07, | |
| "logits/chosen": -1.0511149168014526, | |
| "logits/rejected": -1.1264652013778687, | |
| "logps/chosen": -193.53610229492188, | |
| "logps/rejected": -218.5620880126953, | |
| "loss": 0.701, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.7159083485603333, | |
| "rewards/margins": 0.046296779066324234, | |
| "rewards/rejected": 0.6696116328239441, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4675324675324675, | |
| "grad_norm": 30.130370931610912, | |
| "learning_rate": 4.355585307510675e-07, | |
| "logits/chosen": -1.246776819229126, | |
| "logits/rejected": -1.2227742671966553, | |
| "logps/chosen": -159.0496826171875, | |
| "logps/rejected": -173.915771484375, | |
| "loss": 0.6643, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.7583217620849609, | |
| "rewards/margins": 0.24899302423000336, | |
| "rewards/rejected": 0.5093286633491516, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.4805194805194805, | |
| "grad_norm": 33.33819375543502, | |
| "learning_rate": 4.3210244891456725e-07, | |
| "logits/chosen": -1.3857276439666748, | |
| "logits/rejected": -1.2871744632720947, | |
| "logps/chosen": -237.01812744140625, | |
| "logps/rejected": -222.86903381347656, | |
| "loss": 0.6923, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.6103696227073669, | |
| "rewards/margins": 0.10617219656705856, | |
| "rewards/rejected": 0.5041974186897278, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.4935064935064935, | |
| "grad_norm": 31.374872695083415, | |
| "learning_rate": 4.2857058630547586e-07, | |
| "logits/chosen": -1.1448109149932861, | |
| "logits/rejected": -1.0983011722564697, | |
| "logps/chosen": -251.7313690185547, | |
| "logps/rejected": -263.6651306152344, | |
| "loss": 0.6743, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": 0.733715295791626, | |
| "rewards/margins": 0.12528231739997864, | |
| "rewards/rejected": 0.608432948589325, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5064935064935064, | |
| "grad_norm": 29.77623703163317, | |
| "learning_rate": 4.2496441268589047e-07, | |
| "logits/chosen": -1.0607296228408813, | |
| "logits/rejected": -1.2361115217208862, | |
| "logps/chosen": -208.2403564453125, | |
| "logps/rejected": -205.75625610351562, | |
| "loss": 0.6756, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.9102213382720947, | |
| "rewards/margins": 0.41290420293807983, | |
| "rewards/rejected": 0.4973170757293701, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "grad_norm": 29.02788697480319, | |
| "learning_rate": 4.2128542874196107e-07, | |
| "logits/chosen": -1.0309326648712158, | |
| "logits/rejected": -1.143478274345398, | |
| "logps/chosen": -173.32151794433594, | |
| "logps/rejected": -161.68482971191406, | |
| "loss": 0.6691, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.7697413563728333, | |
| "rewards/margins": 0.14579711854457855, | |
| "rewards/rejected": 0.6239442229270935, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5324675324675324, | |
| "grad_norm": 29.784869306609654, | |
| "learning_rate": 4.1753516545938986e-07, | |
| "logits/chosen": -1.2103991508483887, | |
| "logits/rejected": -1.3010125160217285, | |
| "logps/chosen": -190.08010864257812, | |
| "logps/rejected": -182.82606506347656, | |
| "loss": 0.6618, | |
| "rewards/accuracies": 0.40625, | |
| "rewards/chosen": 0.6524017453193665, | |
| "rewards/margins": 0.07943220436573029, | |
| "rewards/rejected": 0.5729695558547974, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 32.96758632787842, | |
| "learning_rate": 4.137151834863213e-07, | |
| "logits/chosen": -1.1179546117782593, | |
| "logits/rejected": -1.1667120456695557, | |
| "logps/chosen": -252.30087280273438, | |
| "logps/rejected": -258.8082275390625, | |
| "loss": 0.6678, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.7305130958557129, | |
| "rewards/margins": 0.33587121963500977, | |
| "rewards/rejected": 0.39464184641838074, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.5584415584415584, | |
| "grad_norm": 29.473194289500984, | |
| "learning_rate": 4.098270724838879e-07, | |
| "logits/chosen": -1.2531524896621704, | |
| "logits/rejected": -1.2107303142547607, | |
| "logps/chosen": -257.7474365234375, | |
| "logps/rejected": -297.6868591308594, | |
| "loss": 0.6752, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.9224438667297363, | |
| "rewards/margins": 0.45967769622802734, | |
| "rewards/rejected": 0.4627661406993866, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 25.375543282285655, | |
| "learning_rate": 4.058724504646834e-07, | |
| "logits/chosen": -1.212619662284851, | |
| "logits/rejected": -1.218203067779541, | |
| "logps/chosen": -242.98947143554688, | |
| "logps/rejected": -272.5478515625, | |
| "loss": 0.629, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.7270645499229431, | |
| "rewards/margins": 0.3879895508289337, | |
| "rewards/rejected": 0.33907490968704224, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5844155844155844, | |
| "grad_norm": 33.30211893834397, | |
| "learning_rate": 4.018529631194369e-07, | |
| "logits/chosen": -1.3774781227111816, | |
| "logits/rejected": -1.4289326667785645, | |
| "logps/chosen": -341.9387512207031, | |
| "logps/rejected": -348.09283447265625, | |
| "loss": 0.6743, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.8079380989074707, | |
| "rewards/margins": 0.1343669593334198, | |
| "rewards/rejected": 0.6735711097717285, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5974025974025974, | |
| "grad_norm": 33.04398574915242, | |
| "learning_rate": 3.9777028313216913e-07, | |
| "logits/chosen": -1.1324537992477417, | |
| "logits/rejected": -1.2160053253173828, | |
| "logps/chosen": -216.34042358398438, | |
| "logps/rejected": -239.47853088378906, | |
| "loss": 0.6886, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.5493214726448059, | |
| "rewards/margins": 0.09607386589050293, | |
| "rewards/rejected": 0.4532475769519806, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6103896103896104, | |
| "grad_norm": 27.7423356377416, | |
| "learning_rate": 3.9362610948411584e-07, | |
| "logits/chosen": -1.273106336593628, | |
| "logits/rejected": -1.278911828994751, | |
| "logps/chosen": -298.0914306640625, | |
| "logps/rejected": -303.9120788574219, | |
| "loss": 0.6527, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.9156701564788818, | |
| "rewards/margins": 0.36599063873291016, | |
| "rewards/rejected": 0.5496795177459717, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6233766233766234, | |
| "grad_norm": 28.666974084256797, | |
| "learning_rate": 3.8942216674670737e-07, | |
| "logits/chosen": -1.1729968786239624, | |
| "logits/rejected": -1.2594560384750366, | |
| "logps/chosen": -201.3909454345703, | |
| "logps/rejected": -202.02003479003906, | |
| "loss": 0.6969, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": 0.5368354320526123, | |
| "rewards/margins": 0.11352111399173737, | |
| "rewards/rejected": 0.4233143925666809, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6363636363636364, | |
| "grad_norm": 31.936323967340297, | |
| "learning_rate": 3.851602043638994e-07, | |
| "logits/chosen": -1.2385543584823608, | |
| "logits/rejected": -1.3384761810302734, | |
| "logps/chosen": -247.0762939453125, | |
| "logps/rejected": -246.0072021484375, | |
| "loss": 0.7103, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.6583046317100525, | |
| "rewards/margins": 0.3179153501987457, | |
| "rewards/rejected": 0.3403893709182739, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6493506493506493, | |
| "grad_norm": 32.627149716732674, | |
| "learning_rate": 3.80841995924153e-07, | |
| "logits/chosen": -1.2925565242767334, | |
| "logits/rejected": -1.22373628616333, | |
| "logps/chosen": -183.79818725585938, | |
| "logps/rejected": -207.68206787109375, | |
| "loss": 0.6958, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.48615562915802, | |
| "rewards/margins": 0.16806553304195404, | |
| "rewards/rejected": 0.3180900812149048, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6623376623376623, | |
| "grad_norm": 35.550165524685106, | |
| "learning_rate": 3.7646933842236707e-07, | |
| "logits/chosen": -1.3273155689239502, | |
| "logits/rejected": -1.2905724048614502, | |
| "logps/chosen": -210.8981475830078, | |
| "logps/rejected": -253.29486083984375, | |
| "loss": 0.674, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.7722782492637634, | |
| "rewards/margins": 0.2569873034954071, | |
| "rewards/rejected": 0.5152909755706787, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6753246753246753, | |
| "grad_norm": 37.08342788520365, | |
| "learning_rate": 3.720440515120703e-07, | |
| "logits/chosen": -1.0195050239562988, | |
| "logits/rejected": -1.0432817935943604, | |
| "logps/chosen": -190.58543395996094, | |
| "logps/rejected": -212.59231567382812, | |
| "loss": 0.6802, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.7294161915779114, | |
| "rewards/margins": 0.22578200697898865, | |
| "rewards/rejected": 0.5036342144012451, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.6883116883116883, | |
| "grad_norm": 31.103871645303624, | |
| "learning_rate": 3.6756797674818417e-07, | |
| "logits/chosen": -1.1702743768692017, | |
| "logits/rejected": -1.2568817138671875, | |
| "logps/chosen": -288.2113037109375, | |
| "logps/rejected": -276.70550537109375, | |
| "loss": 0.667, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.7896406650543213, | |
| "rewards/margins": 0.35757139325141907, | |
| "rewards/rejected": 0.43206924200057983, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.7012987012987013, | |
| "grad_norm": 33.27213712499375, | |
| "learning_rate": 3.630429768206714e-07, | |
| "logits/chosen": -1.1950292587280273, | |
| "logits/rejected": -1.2156962156295776, | |
| "logps/chosen": -260.03192138671875, | |
| "logps/rejected": -280.8499450683594, | |
| "loss": 0.6698, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.6720072627067566, | |
| "rewards/margins": 0.1990586519241333, | |
| "rewards/rejected": 0.4729485809803009, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 34.347475898145326, | |
| "learning_rate": 3.584709347793895e-07, | |
| "logits/chosen": -1.2583609819412231, | |
| "logits/rejected": -1.2271919250488281, | |
| "logps/chosen": -218.08056640625, | |
| "logps/rejected": -265.71612548828125, | |
| "loss": 0.6771, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.7517715692520142, | |
| "rewards/margins": 0.31436362862586975, | |
| "rewards/rejected": 0.4374079704284668, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 28.787797168356807, | |
| "learning_rate": 3.5385375325047163e-07, | |
| "logits/chosen": -1.301234245300293, | |
| "logits/rejected": -1.3219377994537354, | |
| "logps/chosen": -198.1061553955078, | |
| "logps/rejected": -215.04071044921875, | |
| "loss": 0.6702, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.6467806696891785, | |
| "rewards/margins": 0.201984703540802, | |
| "rewards/rejected": 0.44479596614837646, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7402597402597403, | |
| "grad_norm": 31.164777041001017, | |
| "learning_rate": 3.491933536445606e-07, | |
| "logits/chosen": -1.3274401426315308, | |
| "logits/rejected": -1.2342091798782349, | |
| "logps/chosen": -153.36151123046875, | |
| "logps/rejected": -195.70001220703125, | |
| "loss": 0.6501, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.6609562635421753, | |
| "rewards/margins": 0.36767348647117615, | |
| "rewards/rejected": 0.2932826578617096, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7532467532467533, | |
| "grad_norm": 28.597437073725573, | |
| "learning_rate": 3.4449167535722664e-07, | |
| "logits/chosen": -1.0330864191055298, | |
| "logits/rejected": -1.1435695886611938, | |
| "logps/chosen": -273.2784423828125, | |
| "logps/rejected": -281.6010437011719, | |
| "loss": 0.6587, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.6794790029525757, | |
| "rewards/margins": 0.4547131061553955, | |
| "rewards/rejected": 0.22476595640182495, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7662337662337663, | |
| "grad_norm": 32.62513636994299, | |
| "learning_rate": 3.3975067496189963e-07, | |
| "logits/chosen": -1.2125685214996338, | |
| "logits/rejected": -1.3594530820846558, | |
| "logps/chosen": -317.5354919433594, | |
| "logps/rejected": -315.6915588378906, | |
| "loss": 0.65, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.8727169036865234, | |
| "rewards/margins": 0.7781965136528015, | |
| "rewards/rejected": 0.09452031552791595, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 32.86814732954617, | |
| "learning_rate": 3.349723253956541e-07, | |
| "logits/chosen": -1.0453016757965088, | |
| "logits/rejected": -1.1661120653152466, | |
| "logps/chosen": -145.11253356933594, | |
| "logps/rejected": -178.77328491210938, | |
| "loss": 0.6635, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.7022942900657654, | |
| "rewards/margins": 0.42999541759490967, | |
| "rewards/rejected": 0.2722988426685333, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7922077922077922, | |
| "grad_norm": 31.986096882532095, | |
| "learning_rate": 3.3015861513818383e-07, | |
| "logits/chosen": -1.1393004655838013, | |
| "logits/rejected": -1.071953296661377, | |
| "logps/chosen": -257.8083801269531, | |
| "logps/rejected": -292.0323791503906, | |
| "loss": 0.66, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.6557089686393738, | |
| "rewards/margins": 0.2878965139389038, | |
| "rewards/rejected": 0.3678124248981476, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.8051948051948052, | |
| "grad_norm": 32.75329554864317, | |
| "learning_rate": 3.2531154738430853e-07, | |
| "logits/chosen": -1.2495208978652954, | |
| "logits/rejected": -1.1901376247406006, | |
| "logps/chosen": -277.14208984375, | |
| "logps/rejected": -308.6678161621094, | |
| "loss": 0.6484, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.6612053513526917, | |
| "rewards/margins": 0.33286944031715393, | |
| "rewards/rejected": 0.32833582162857056, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8181818181818182, | |
| "grad_norm": 35.313227849390415, | |
| "learning_rate": 3.204331392103574e-07, | |
| "logits/chosen": -1.0496852397918701, | |
| "logits/rejected": -1.0938959121704102, | |
| "logps/chosen": -183.056884765625, | |
| "logps/rejected": -218.71853637695312, | |
| "loss": 0.6607, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.7791882157325745, | |
| "rewards/margins": 0.2896581292152405, | |
| "rewards/rejected": 0.489530086517334, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8311688311688312, | |
| "grad_norm": 32.66165401072115, | |
| "learning_rate": 3.155254207347755e-07, | |
| "logits/chosen": -1.154829740524292, | |
| "logits/rejected": -1.0259711742401123, | |
| "logps/chosen": -193.2096710205078, | |
| "logps/rejected": -234.06759643554688, | |
| "loss": 0.6539, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.7302087545394897, | |
| "rewards/margins": 0.37150081992149353, | |
| "rewards/rejected": 0.3587079346179962, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8441558441558441, | |
| "grad_norm": 34.704917232369, | |
| "learning_rate": 3.1059043427330314e-07, | |
| "logits/chosen": -1.179085373878479, | |
| "logits/rejected": -1.1466186046600342, | |
| "logps/chosen": -290.7761535644531, | |
| "logps/rejected": -247.35595703125, | |
| "loss": 0.689, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": 0.7207962274551392, | |
| "rewards/margins": 0.21811310946941376, | |
| "rewards/rejected": 0.5026831030845642, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 31.672800837060226, | |
| "learning_rate": 3.056302334890786e-07, | |
| "logits/chosen": -1.0660974979400635, | |
| "logits/rejected": -1.1430083513259888, | |
| "logps/chosen": -186.87071228027344, | |
| "logps/rejected": -191.1336669921875, | |
| "loss": 0.6728, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.577528178691864, | |
| "rewards/margins": 0.338672012090683, | |
| "rewards/rejected": 0.23885619640350342, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8701298701298701, | |
| "grad_norm": 31.658738700878413, | |
| "learning_rate": 3.0064688253802024e-07, | |
| "logits/chosen": -1.0814266204833984, | |
| "logits/rejected": -1.141867756843567, | |
| "logps/chosen": -219.94424438476562, | |
| "logps/rejected": -249.24383544921875, | |
| "loss": 0.6633, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.6220312118530273, | |
| "rewards/margins": 0.4069012999534607, | |
| "rewards/rejected": 0.21512994170188904, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.8831168831168831, | |
| "grad_norm": 33.16970393800331, | |
| "learning_rate": 2.956424552098404e-07, | |
| "logits/chosen": -1.313258171081543, | |
| "logits/rejected": -1.3151941299438477, | |
| "logps/chosen": -300.03961181640625, | |
| "logps/rejected": -340.99566650390625, | |
| "loss": 0.6998, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.7123932242393494, | |
| "rewards/margins": 0.30548784136772156, | |
| "rewards/rejected": 0.4069053828716278, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.8961038961038961, | |
| "grad_norm": 27.93166457525636, | |
| "learning_rate": 2.9061903406505153e-07, | |
| "logits/chosen": -1.141364574432373, | |
| "logits/rejected": -1.0247657299041748, | |
| "logps/chosen": -200.64974975585938, | |
| "logps/rejected": -197.97593688964844, | |
| "loss": 0.6424, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.630258321762085, | |
| "rewards/margins": 0.3348933160305023, | |
| "rewards/rejected": 0.29536497592926025, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 28.870063032351457, | |
| "learning_rate": 2.8557870956832133e-07, | |
| "logits/chosen": -1.274742603302002, | |
| "logits/rejected": -1.2089191675186157, | |
| "logps/chosen": -197.0199737548828, | |
| "logps/rejected": -220.10044860839844, | |
| "loss": 0.642, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.6193188428878784, | |
| "rewards/margins": 0.4841911494731903, | |
| "rewards/rejected": 0.1351277232170105, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.922077922077922, | |
| "grad_norm": 34.83723284802987, | |
| "learning_rate": 2.8052357921854e-07, | |
| "logits/chosen": -1.0989036560058594, | |
| "logits/rejected": -1.2587521076202393, | |
| "logps/chosen": -167.8390350341797, | |
| "logps/rejected": -161.0415496826172, | |
| "loss": 0.6461, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.7375519275665283, | |
| "rewards/margins": 0.4270685017108917, | |
| "rewards/rejected": 0.3104833960533142, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.935064935064935, | |
| "grad_norm": 32.08059076203408, | |
| "learning_rate": 2.754557466759589e-07, | |
| "logits/chosen": -1.296328067779541, | |
| "logits/rejected": -1.3102202415466309, | |
| "logps/chosen": -209.74093627929688, | |
| "logps/rejected": -220.9139862060547, | |
| "loss": 0.6487, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.611607551574707, | |
| "rewards/margins": 0.38875120878219604, | |
| "rewards/rejected": 0.22285637259483337, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.948051948051948, | |
| "grad_norm": 30.710641933445263, | |
| "learning_rate": 2.703773208867658e-07, | |
| "logits/chosen": -1.2812106609344482, | |
| "logits/rejected": -1.2686681747436523, | |
| "logps/chosen": -289.8981018066406, | |
| "logps/rejected": -298.4099426269531, | |
| "loss": 0.6562, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.6197431087493896, | |
| "rewards/margins": 0.27826988697052, | |
| "rewards/rejected": 0.341473251581192, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.961038961038961, | |
| "grad_norm": 32.51693326041137, | |
| "learning_rate": 2.652904152054607e-07, | |
| "logits/chosen": -1.2882963418960571, | |
| "logits/rejected": -1.234848141670227, | |
| "logps/chosen": -212.38536071777344, | |
| "logps/rejected": -235.0468292236328, | |
| "loss": 0.657, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.8348413109779358, | |
| "rewards/margins": 0.48668912053108215, | |
| "rewards/rejected": 0.34815219044685364, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.974025974025974, | |
| "grad_norm": 33.13371475089395, | |
| "learning_rate": 2.6019714651539645e-07, | |
| "logits/chosen": -1.3393000364303589, | |
| "logits/rejected": -1.3905009031295776, | |
| "logps/chosen": -223.33636474609375, | |
| "logps/rejected": -243.6895294189453, | |
| "loss": 0.6381, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.5773805379867554, | |
| "rewards/margins": 0.328535795211792, | |
| "rewards/rejected": 0.24884477257728577, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.987012987012987, | |
| "grad_norm": 29.25513766134415, | |
| "learning_rate": 2.550996343478514e-07, | |
| "logits/chosen": -1.210328459739685, | |
| "logits/rejected": -1.1693918704986572, | |
| "logps/chosen": -266.2218933105469, | |
| "logps/rejected": -318.95904541015625, | |
| "loss": 0.6367, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.5575440526008606, | |
| "rewards/margins": 0.3903830647468567, | |
| "rewards/rejected": 0.1671610325574875, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 33.4683934031025, | |
| "learning_rate": 2.5e-07, | |
| "logits/chosen": -1.2191190719604492, | |
| "logits/rejected": -1.215990424156189, | |
| "logps/chosen": -267.1896667480469, | |
| "logps/rejected": -297.6923828125, | |
| "loss": 0.6461, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.5746477246284485, | |
| "rewards/margins": 0.4244813323020935, | |
| "rewards/rejected": 0.15016639232635498, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.0129870129870129, | |
| "grad_norm": 22.52614785258628, | |
| "learning_rate": 2.449003656521487e-07, | |
| "logits/chosen": -1.052141785621643, | |
| "logits/rejected": -1.2538810968399048, | |
| "logps/chosen": -262.6394958496094, | |
| "logps/rejected": -265.5911560058594, | |
| "loss": 0.5416, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.68734210729599, | |
| "rewards/margins": 0.7715808749198914, | |
| "rewards/rejected": -0.08423884212970734, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.025974025974026, | |
| "grad_norm": 17.71525419241131, | |
| "learning_rate": 2.3980285348460363e-07, | |
| "logits/chosen": -1.2752505540847778, | |
| "logits/rejected": -1.297239065170288, | |
| "logps/chosen": -180.13455200195312, | |
| "logps/rejected": -194.86289978027344, | |
| "loss": 0.5164, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.637496292591095, | |
| "rewards/margins": 0.591891884803772, | |
| "rewards/rejected": 0.045604437589645386, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0389610389610389, | |
| "grad_norm": 17.624016800680035, | |
| "learning_rate": 2.3470958479453937e-07, | |
| "logits/chosen": -1.3482556343078613, | |
| "logits/rejected": -1.2880206108093262, | |
| "logps/chosen": -183.36322021484375, | |
| "logps/rejected": -208.04705810546875, | |
| "loss": 0.524, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.6870389580726624, | |
| "rewards/margins": 0.8608755469322205, | |
| "rewards/rejected": -0.17383655905723572, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.051948051948052, | |
| "grad_norm": 16.65756304879653, | |
| "learning_rate": 2.296226791132342e-07, | |
| "logits/chosen": -1.240028977394104, | |
| "logits/rejected": -1.2343027591705322, | |
| "logps/chosen": -233.16773986816406, | |
| "logps/rejected": -250.71063232421875, | |
| "loss": 0.5179, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.796766996383667, | |
| "rewards/margins": 0.684018611907959, | |
| "rewards/rejected": 0.11274835467338562, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.0649350649350648, | |
| "grad_norm": 18.048904117373002, | |
| "learning_rate": 2.245442533240412e-07, | |
| "logits/chosen": -1.2781399488449097, | |
| "logits/rejected": -1.2129652500152588, | |
| "logps/chosen": -244.93289184570312, | |
| "logps/rejected": -284.34075927734375, | |
| "loss": 0.5295, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": 0.6863990426063538, | |
| "rewards/margins": 0.871242344379425, | |
| "rewards/rejected": -0.1848432570695877, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.077922077922078, | |
| "grad_norm": 20.34696734972406, | |
| "learning_rate": 2.1947642078146e-07, | |
| "logits/chosen": -1.322028398513794, | |
| "logits/rejected": -1.3042186498641968, | |
| "logps/chosen": -271.44500732421875, | |
| "logps/rejected": -270.650146484375, | |
| "loss": 0.5102, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.853551983833313, | |
| "rewards/margins": 0.9862384796142578, | |
| "rewards/rejected": -0.1326863318681717, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 17.55986753162257, | |
| "learning_rate": 2.1442129043167873e-07, | |
| "logits/chosen": -1.328125, | |
| "logits/rejected": -1.3073420524597168, | |
| "logps/chosen": -206.6400909423828, | |
| "logps/rejected": -220.58033752441406, | |
| "loss": 0.5149, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 0.7431928515434265, | |
| "rewards/margins": 0.8430217504501343, | |
| "rewards/rejected": -0.09982895851135254, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.103896103896104, | |
| "grad_norm": 17.856674110170477, | |
| "learning_rate": 2.0938096593494853e-07, | |
| "logits/chosen": -0.9651075005531311, | |
| "logits/rejected": -1.0858749151229858, | |
| "logps/chosen": -185.41217041015625, | |
| "logps/rejected": -206.42196655273438, | |
| "loss": 0.5435, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.7554150819778442, | |
| "rewards/margins": 0.6504155397415161, | |
| "rewards/rejected": 0.10499954968690872, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.1168831168831168, | |
| "grad_norm": 19.78444270286199, | |
| "learning_rate": 2.043575447901596e-07, | |
| "logits/chosen": -1.22752845287323, | |
| "logits/rejected": -1.306286334991455, | |
| "logps/chosen": -281.8180236816406, | |
| "logps/rejected": -299.3645935058594, | |
| "loss": 0.5278, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.7911855578422546, | |
| "rewards/margins": 0.6221764087677002, | |
| "rewards/rejected": 0.16900911927223206, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.12987012987013, | |
| "grad_norm": 20.24750613090253, | |
| "learning_rate": 1.9935311746197976e-07, | |
| "logits/chosen": -1.0992745161056519, | |
| "logits/rejected": -1.0037593841552734, | |
| "logps/chosen": -219.87008666992188, | |
| "logps/rejected": -246.12489318847656, | |
| "loss": 0.5273, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 0.8513480424880981, | |
| "rewards/margins": 0.9342952370643616, | |
| "rewards/rejected": -0.08294717967510223, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 19.866822049463956, | |
| "learning_rate": 1.9436976651092142e-07, | |
| "logits/chosen": -1.3867590427398682, | |
| "logits/rejected": -1.2858481407165527, | |
| "logps/chosen": -258.7950439453125, | |
| "logps/rejected": -296.06842041015625, | |
| "loss": 0.5338, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": 0.7717873454093933, | |
| "rewards/margins": 0.8829994201660156, | |
| "rewards/rejected": -0.11121205985546112, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.155844155844156, | |
| "grad_norm": 18.093341348498807, | |
| "learning_rate": 1.8940956572669692e-07, | |
| "logits/chosen": -1.0915521383285522, | |
| "logits/rejected": -1.1871168613433838, | |
| "logps/chosen": -247.3667755126953, | |
| "logps/rejected": -254.37728881835938, | |
| "loss": 0.4999, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": 0.7344910502433777, | |
| "rewards/margins": 0.7577481865882874, | |
| "rewards/rejected": -0.023257076740264893, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.1688311688311688, | |
| "grad_norm": 19.525075850508752, | |
| "learning_rate": 1.8447457926522452e-07, | |
| "logits/chosen": -0.9950531721115112, | |
| "logits/rejected": -0.994107186794281, | |
| "logps/chosen": -259.6279296875, | |
| "logps/rejected": -308.6029052734375, | |
| "loss": 0.5492, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.7692892551422119, | |
| "rewards/margins": 0.9317154884338379, | |
| "rewards/rejected": -0.16242632269859314, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.1818181818181819, | |
| "grad_norm": 19.472127981634152, | |
| "learning_rate": 1.7956686078964255e-07, | |
| "logits/chosen": -1.3088796138763428, | |
| "logits/rejected": -1.4440947771072388, | |
| "logps/chosen": -304.2322082519531, | |
| "logps/rejected": -302.7728271484375, | |
| "loss": 0.5274, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": 0.815383791923523, | |
| "rewards/margins": 0.9992111921310425, | |
| "rewards/rejected": -0.18382729589939117, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.1948051948051948, | |
| "grad_norm": 17.961643519105607, | |
| "learning_rate": 1.7468845261569147e-07, | |
| "logits/chosen": -1.3604259490966797, | |
| "logits/rejected": -1.1473455429077148, | |
| "logps/chosen": -278.00860595703125, | |
| "logps/rejected": -333.7747802734375, | |
| "loss": 0.5072, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": 0.8540161848068237, | |
| "rewards/margins": 0.9145561456680298, | |
| "rewards/rejected": -0.060539960861206055, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.2077922077922079, | |
| "grad_norm": 17.34162227747004, | |
| "learning_rate": 1.698413848618161e-07, | |
| "logits/chosen": -1.2001099586486816, | |
| "logits/rejected": -1.1247773170471191, | |
| "logps/chosen": -219.12808227539062, | |
| "logps/rejected": -240.87974548339844, | |
| "loss": 0.5124, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.792206883430481, | |
| "rewards/margins": 0.9953896403312683, | |
| "rewards/rejected": -0.20318272709846497, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.2207792207792207, | |
| "grad_norm": 18.899208870167122, | |
| "learning_rate": 1.6502767460434585e-07, | |
| "logits/chosen": -1.1938526630401611, | |
| "logits/rejected": -1.2088103294372559, | |
| "logps/chosen": -343.2425537109375, | |
| "logps/rejected": -352.4480895996094, | |
| "loss": 0.5276, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.7310408353805542, | |
| "rewards/margins": 0.948591947555542, | |
| "rewards/rejected": -0.21755114197731018, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2337662337662338, | |
| "grad_norm": 21.1592051257712, | |
| "learning_rate": 1.602493250381003e-07, | |
| "logits/chosen": -1.2564481496810913, | |
| "logits/rejected": -1.3384077548980713, | |
| "logps/chosen": -211.84091186523438, | |
| "logps/rejected": -175.0691375732422, | |
| "loss": 0.5227, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.6941075325012207, | |
| "rewards/margins": 0.8241455554962158, | |
| "rewards/rejected": -0.13003802299499512, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.2467532467532467, | |
| "grad_norm": 17.37996372066179, | |
| "learning_rate": 1.555083246427734e-07, | |
| "logits/chosen": -1.1990139484405518, | |
| "logits/rejected": -1.3094348907470703, | |
| "logps/chosen": -227.32859802246094, | |
| "logps/rejected": -240.32440185546875, | |
| "loss": 0.5063, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.8420521020889282, | |
| "rewards/margins": 0.7185032367706299, | |
| "rewards/rejected": 0.12354880571365356, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.2597402597402598, | |
| "grad_norm": 19.348154887947697, | |
| "learning_rate": 1.5080664635543932e-07, | |
| "logits/chosen": -1.2713487148284912, | |
| "logits/rejected": -1.1336112022399902, | |
| "logps/chosen": -217.08712768554688, | |
| "logps/rejected": -286.9374084472656, | |
| "loss": 0.5347, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.9916350245475769, | |
| "rewards/margins": 0.9090840220451355, | |
| "rewards/rejected": 0.08255089819431305, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 18.399160019731372, | |
| "learning_rate": 1.461462467495284e-07, | |
| "logits/chosen": -1.1585338115692139, | |
| "logits/rejected": -1.1779534816741943, | |
| "logps/chosen": -184.70875549316406, | |
| "logps/rejected": -199.60733032226562, | |
| "loss": 0.5194, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.6531062722206116, | |
| "rewards/margins": 0.5796053409576416, | |
| "rewards/rejected": 0.07350096851587296, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 17.142926149258614, | |
| "learning_rate": 1.4152906522061047e-07, | |
| "logits/chosen": -1.145853877067566, | |
| "logits/rejected": -1.1914734840393066, | |
| "logps/chosen": -244.29345703125, | |
| "logps/rejected": -252.35826110839844, | |
| "loss": 0.5101, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.8408298492431641, | |
| "rewards/margins": 0.9886577725410461, | |
| "rewards/rejected": -0.14782798290252686, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.2987012987012987, | |
| "grad_norm": 19.542604645193908, | |
| "learning_rate": 1.369570231793286e-07, | |
| "logits/chosen": -1.2806802988052368, | |
| "logits/rejected": -1.2916889190673828, | |
| "logps/chosen": -233.01324462890625, | |
| "logps/rejected": -256.65887451171875, | |
| "loss": 0.515, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.8423460721969604, | |
| "rewards/margins": 0.8902460336685181, | |
| "rewards/rejected": -0.04789996147155762, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3116883116883118, | |
| "grad_norm": 18.81859489288187, | |
| "learning_rate": 1.3243202325181578e-07, | |
| "logits/chosen": -1.0917601585388184, | |
| "logits/rejected": -1.1492509841918945, | |
| "logps/chosen": -241.70211791992188, | |
| "logps/rejected": -247.6362762451172, | |
| "loss": 0.5173, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.8184989094734192, | |
| "rewards/margins": 0.8419989347457886, | |
| "rewards/rejected": -0.023499924689531326, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.3246753246753247, | |
| "grad_norm": 21.98362763457428, | |
| "learning_rate": 1.2795594848792974e-07, | |
| "logits/chosen": -1.2968580722808838, | |
| "logits/rejected": -1.3090989589691162, | |
| "logps/chosen": -209.9777374267578, | |
| "logps/rejected": -249.19898986816406, | |
| "loss": 0.5144, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 0.6729121208190918, | |
| "rewards/margins": 0.8331139087677002, | |
| "rewards/rejected": -0.16020171344280243, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.3376623376623376, | |
| "grad_norm": 19.163927573547067, | |
| "learning_rate": 1.2353066157763304e-07, | |
| "logits/chosen": -1.165097713470459, | |
| "logits/rejected": -1.0794737339019775, | |
| "logps/chosen": -243.25994873046875, | |
| "logps/rejected": -296.20458984375, | |
| "loss": 0.5074, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.8912444710731506, | |
| "rewards/margins": 1.0708703994750977, | |
| "rewards/rejected": -0.17962592840194702, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.3506493506493507, | |
| "grad_norm": 19.327223964730898, | |
| "learning_rate": 1.1915800407584702e-07, | |
| "logits/chosen": -1.2650034427642822, | |
| "logits/rejected": -1.260176181793213, | |
| "logps/chosen": -297.4647216796875, | |
| "logps/rejected": -283.3460998535156, | |
| "loss": 0.5184, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.607912540435791, | |
| "rewards/margins": 0.5631698369979858, | |
| "rewards/rejected": 0.04474277421832085, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 17.1533818598118, | |
| "learning_rate": 1.1483979563610069e-07, | |
| "logits/chosen": -1.3990474939346313, | |
| "logits/rejected": -1.1992498636245728, | |
| "logps/chosen": -179.99732971191406, | |
| "logps/rejected": -254.8299560546875, | |
| "loss": 0.5, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 0.6216685771942139, | |
| "rewards/margins": 0.651849627494812, | |
| "rewards/rejected": -0.03018118068575859, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.3766233766233766, | |
| "grad_norm": 18.960132927151744, | |
| "learning_rate": 1.1057783325329267e-07, | |
| "logits/chosen": -1.1669042110443115, | |
| "logits/rejected": -1.221852421760559, | |
| "logps/chosen": -292.6320495605469, | |
| "logps/rejected": -333.9742736816406, | |
| "loss": 0.505, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": 0.997809648513794, | |
| "rewards/margins": 1.133697748184204, | |
| "rewards/rejected": -0.13588815927505493, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.3896103896103895, | |
| "grad_norm": 16.02074293153386, | |
| "learning_rate": 1.0637389051588425e-07, | |
| "logits/chosen": -1.1035025119781494, | |
| "logits/rejected": -1.113924503326416, | |
| "logps/chosen": -217.35040283203125, | |
| "logps/rejected": -243.62997436523438, | |
| "loss": 0.5012, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": 0.7329657673835754, | |
| "rewards/margins": 0.827349066734314, | |
| "rewards/rejected": -0.09438329190015793, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.4025974025974026, | |
| "grad_norm": 16.742187965244536, | |
| "learning_rate": 1.0222971686783089e-07, | |
| "logits/chosen": -1.392643690109253, | |
| "logits/rejected": -1.4035465717315674, | |
| "logps/chosen": -252.4545440673828, | |
| "logps/rejected": -249.23336791992188, | |
| "loss": 0.5069, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 0.7851105332374573, | |
| "rewards/margins": 0.8927165865898132, | |
| "rewards/rejected": -0.10760608315467834, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.4155844155844157, | |
| "grad_norm": 20.851477095001684, | |
| "learning_rate": 9.814703688056319e-08, | |
| "logits/chosen": -1.1926751136779785, | |
| "logits/rejected": -1.2301442623138428, | |
| "logps/chosen": -165.44703674316406, | |
| "logps/rejected": -194.3583221435547, | |
| "loss": 0.5326, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": 0.8975037336349487, | |
| "rewards/margins": 0.9634904265403748, | |
| "rewards/rejected": -0.065986767411232, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 16.315287900746362, | |
| "learning_rate": 9.412754953531663e-08, | |
| "logits/chosen": -1.3106112480163574, | |
| "logits/rejected": -1.3371086120605469, | |
| "logps/chosen": -255.00607299804688, | |
| "logps/rejected": -258.26361083984375, | |
| "loss": 0.5037, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.6847482919692993, | |
| "rewards/margins": 0.697127103805542, | |
| "rewards/rejected": -0.012378760613501072, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.4415584415584415, | |
| "grad_norm": 15.292540532901763, | |
| "learning_rate": 9.017292751611219e-08, | |
| "logits/chosen": -0.9778515100479126, | |
| "logits/rejected": -0.9790754318237305, | |
| "logps/chosen": -189.63674926757812, | |
| "logps/rejected": -222.9846649169922, | |
| "loss": 0.5021, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 0.6293007135391235, | |
| "rewards/margins": 0.9083631038665771, | |
| "rewards/rejected": -0.27906233072280884, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 17.14845228574236, | |
| "learning_rate": 8.628481651367875e-08, | |
| "logits/chosen": -1.227216362953186, | |
| "logits/rejected": -1.306528091430664, | |
| "logps/chosen": -212.0201873779297, | |
| "logps/rejected": -204.0389404296875, | |
| "loss": 0.5137, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": 0.9179770946502686, | |
| "rewards/margins": 0.7093281745910645, | |
| "rewards/rejected": 0.20864906907081604, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.4675324675324675, | |
| "grad_norm": 18.007638315726304, | |
| "learning_rate": 8.246483454061015e-08, | |
| "logits/chosen": -1.3051916360855103, | |
| "logits/rejected": -1.121058702468872, | |
| "logps/chosen": -251.83587646484375, | |
| "logps/rejected": -286.392822265625, | |
| "loss": 0.5128, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": 0.7694352269172668, | |
| "rewards/margins": 1.2012625932693481, | |
| "rewards/rejected": -0.43182751536369324, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.4805194805194806, | |
| "grad_norm": 16.89635894695866, | |
| "learning_rate": 7.871457125803896e-08, | |
| "logits/chosen": -1.3657732009887695, | |
| "logits/rejected": -1.4764338731765747, | |
| "logps/chosen": -242.94061279296875, | |
| "logps/rejected": -254.3301239013672, | |
| "loss": 0.4956, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.9610437750816345, | |
| "rewards/margins": 1.0195040702819824, | |
| "rewards/rejected": -0.058460310101509094, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.4935064935064934, | |
| "grad_norm": 18.122088719985634, | |
| "learning_rate": 7.503558731410958e-08, | |
| "logits/chosen": -1.1523171663284302, | |
| "logits/rejected": -1.183262586593628, | |
| "logps/chosen": -255.1221466064453, | |
| "logps/rejected": -274.8419494628906, | |
| "loss": 0.5169, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.6205594539642334, | |
| "rewards/margins": 0.730150580406189, | |
| "rewards/rejected": -0.10959106683731079, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.5064935064935066, | |
| "grad_norm": 17.706521703256886, | |
| "learning_rate": 7.14294136945241e-08, | |
| "logits/chosen": -1.226336121559143, | |
| "logits/rejected": -1.068296194076538, | |
| "logps/chosen": -210.2167510986328, | |
| "logps/rejected": -264.81085205078125, | |
| "loss": 0.5089, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.7191243171691895, | |
| "rewards/margins": 1.1839776039123535, | |
| "rewards/rejected": -0.46485334634780884, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.5194805194805194, | |
| "grad_norm": 16.66406313160021, | |
| "learning_rate": 6.789755108543274e-08, | |
| "logits/chosen": -1.228325605392456, | |
| "logits/rejected": -1.3681098222732544, | |
| "logps/chosen": -172.350830078125, | |
| "logps/rejected": -175.53936767578125, | |
| "loss": 0.5051, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.6207537651062012, | |
| "rewards/margins": 0.7215192317962646, | |
| "rewards/rejected": -0.10076545178890228, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.5324675324675323, | |
| "grad_norm": 18.51693060533818, | |
| "learning_rate": 6.444146924893251e-08, | |
| "logits/chosen": -1.101006269454956, | |
| "logits/rejected": -1.0755804777145386, | |
| "logps/chosen": -233.41761779785156, | |
| "logps/rejected": -264.3063659667969, | |
| "loss": 0.5173, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": 0.8273410797119141, | |
| "rewards/margins": 1.0633716583251953, | |
| "rewards/rejected": -0.23603063821792603, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.5454545454545454, | |
| "grad_norm": 18.052624843204704, | |
| "learning_rate": 6.106260641143546e-08, | |
| "logits/chosen": -1.1633551120758057, | |
| "logits/rejected": -1.1994410753250122, | |
| "logps/chosen": -269.3063659667969, | |
| "logps/rejected": -271.7884216308594, | |
| "loss": 0.4853, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": 0.6259198188781738, | |
| "rewards/margins": 0.9741422533988953, | |
| "rewards/rejected": -0.34822240471839905, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.5584415584415585, | |
| "grad_norm": 24.12818089831554, | |
| "learning_rate": 5.776236866515946e-08, | |
| "logits/chosen": -1.0952610969543457, | |
| "logits/rejected": -1.2616946697235107, | |
| "logps/chosen": -209.5765380859375, | |
| "logps/rejected": -185.8994140625, | |
| "loss": 0.5069, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.6148033738136292, | |
| "rewards/margins": 0.7846094369888306, | |
| "rewards/rejected": -0.16980606317520142, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 17.25585157151766, | |
| "learning_rate": 5.454212938299255e-08, | |
| "logits/chosen": -1.220323920249939, | |
| "logits/rejected": -1.1233211755752563, | |
| "logps/chosen": -210.1920623779297, | |
| "logps/rejected": -257.0649719238281, | |
| "loss": 0.4983, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.8331396579742432, | |
| "rewards/margins": 0.8172454833984375, | |
| "rewards/rejected": 0.01589421182870865, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.5844155844155843, | |
| "grad_norm": 18.694264506215937, | |
| "learning_rate": 5.140322864697183e-08, | |
| "logits/chosen": -1.2713161706924438, | |
| "logits/rejected": -1.1681023836135864, | |
| "logps/chosen": -250.68112182617188, | |
| "logps/rejected": -248.07406616210938, | |
| "loss": 0.5152, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": 0.8290749192237854, | |
| "rewards/margins": 0.7929165363311768, | |
| "rewards/rejected": 0.036158446222543716, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.5974025974025974, | |
| "grad_norm": 18.951231695832167, | |
| "learning_rate": 4.8346972690617494e-08, | |
| "logits/chosen": -1.2086520195007324, | |
| "logits/rejected": -1.2194980382919312, | |
| "logps/chosen": -237.5225067138672, | |
| "logps/rejected": -294.5182800292969, | |
| "loss": 0.4988, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.823986291885376, | |
| "rewards/margins": 1.3896714448928833, | |
| "rewards/rejected": -0.5656850934028625, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.6103896103896105, | |
| "grad_norm": 17.838639398070846, | |
| "learning_rate": 4.53746333553516e-08, | |
| "logits/chosen": -1.3609139919281006, | |
| "logits/rejected": -1.3907623291015625, | |
| "logps/chosen": -254.64718627929688, | |
| "logps/rejected": -267.22265625, | |
| "loss": 0.499, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.7264419794082642, | |
| "rewards/margins": 0.9065679311752319, | |
| "rewards/rejected": -0.18012598156929016, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.6233766233766234, | |
| "grad_norm": 15.282492364953745, | |
| "learning_rate": 4.248744756122985e-08, | |
| "logits/chosen": -1.4239810705184937, | |
| "logits/rejected": -1.3679332733154297, | |
| "logps/chosen": -231.71435546875, | |
| "logps/rejected": -287.30572509765625, | |
| "loss": 0.5073, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 0.5901214480400085, | |
| "rewards/margins": 0.7662074565887451, | |
| "rewards/rejected": -0.17608599364757538, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 18.90237932042577, | |
| "learning_rate": 3.968661679220467e-08, | |
| "logits/chosen": -1.0735975503921509, | |
| "logits/rejected": -1.167972207069397, | |
| "logps/chosen": -188.3458709716797, | |
| "logps/rejected": -179.90293884277344, | |
| "loss": 0.5256, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 0.6942843794822693, | |
| "rewards/margins": 0.7579864859580994, | |
| "rewards/rejected": -0.06370209902524948, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.6493506493506493, | |
| "grad_norm": 17.45584095837454, | |
| "learning_rate": 3.6973306596135873e-08, | |
| "logits/chosen": -1.3392971754074097, | |
| "logits/rejected": -1.358009696006775, | |
| "logps/chosen": -254.31295776367188, | |
| "logps/rejected": -254.20733642578125, | |
| "loss": 0.502, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": 0.7509973645210266, | |
| "rewards/margins": 0.8210946917533875, | |
| "rewards/rejected": -0.07009733468294144, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.6623376623376624, | |
| "grad_norm": 19.754012599719967, | |
| "learning_rate": 3.43486460997548e-08, | |
| "logits/chosen": -1.2106963396072388, | |
| "logits/rejected": -1.2819626331329346, | |
| "logps/chosen": -253.831787109375, | |
| "logps/rejected": -303.4576721191406, | |
| "loss": 0.5023, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.732483446598053, | |
| "rewards/margins": 1.113276720046997, | |
| "rewards/rejected": -0.38079336285591125, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.6753246753246753, | |
| "grad_norm": 18.3258569948234, | |
| "learning_rate": 3.1813727538785943e-08, | |
| "logits/chosen": -1.3981492519378662, | |
| "logits/rejected": -1.4056671857833862, | |
| "logps/chosen": -356.7002868652344, | |
| "logps/rejected": -362.1513366699219, | |
| "loss": 0.494, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 0.6453557014465332, | |
| "rewards/margins": 1.1145414113998413, | |
| "rewards/rejected": -0.46918565034866333, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.6883116883116882, | |
| "grad_norm": 19.670963229677074, | |
| "learning_rate": 2.936960580341971e-08, | |
| "logits/chosen": -1.1725897789001465, | |
| "logits/rejected": -1.261063575744629, | |
| "logps/chosen": -216.58990478515625, | |
| "logps/rejected": -220.63986206054688, | |
| "loss": 0.5231, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.7476328015327454, | |
| "rewards/margins": 0.7463090419769287, | |
| "rewards/rejected": 0.0013237213715910912, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.7012987012987013, | |
| "grad_norm": 17.983233221990766, | |
| "learning_rate": 2.701729799932653e-08, | |
| "logits/chosen": -1.3052499294281006, | |
| "logits/rejected": -1.333134412765503, | |
| "logps/chosen": -277.6193542480469, | |
| "logps/rejected": -287.4527893066406, | |
| "loss": 0.4954, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.726482629776001, | |
| "rewards/margins": 0.9591711759567261, | |
| "rewards/rejected": -0.23268845677375793, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 17.467592052173433, | |
| "learning_rate": 2.475778302439524e-08, | |
| "logits/chosen": -1.2939479351043701, | |
| "logits/rejected": -1.3279248476028442, | |
| "logps/chosen": -207.76170349121094, | |
| "logps/rejected": -233.1382598876953, | |
| "loss": 0.472, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.8571805953979492, | |
| "rewards/margins": 0.9823651313781738, | |
| "rewards/rejected": -0.12518461048603058, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.7272727272727273, | |
| "grad_norm": 16.46335190300962, | |
| "learning_rate": 2.259200116137039e-08, | |
| "logits/chosen": -1.1292874813079834, | |
| "logits/rejected": -1.2274911403656006, | |
| "logps/chosen": -277.028564453125, | |
| "logps/rejected": -308.0982971191406, | |
| "loss": 0.4976, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.9631680846214294, | |
| "rewards/margins": 1.1252551078796387, | |
| "rewards/rejected": -0.16208705306053162, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.7402597402597402, | |
| "grad_norm": 18.56546944179356, | |
| "learning_rate": 2.0520853686560175e-08, | |
| "logits/chosen": -1.2864351272583008, | |
| "logits/rejected": -1.2748289108276367, | |
| "logps/chosen": -203.02589416503906, | |
| "logps/rejected": -217.65682983398438, | |
| "loss": 0.5072, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.5683714747428894, | |
| "rewards/margins": 0.9107145667076111, | |
| "rewards/rejected": -0.3423430323600769, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.7532467532467533, | |
| "grad_norm": 16.59250739344183, | |
| "learning_rate": 1.854520249477551e-08, | |
| "logits/chosen": -1.3317527770996094, | |
| "logits/rejected": -1.3174333572387695, | |
| "logps/chosen": -205.539306640625, | |
| "logps/rejected": -236.86062622070312, | |
| "loss": 0.487, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.8909409642219543, | |
| "rewards/margins": 1.0206615924835205, | |
| "rewards/rejected": -0.1297205686569214, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.7662337662337664, | |
| "grad_norm": 18.312480737747663, | |
| "learning_rate": 1.666586974065831e-08, | |
| "logits/chosen": -1.1755964756011963, | |
| "logits/rejected": -1.2511848211288452, | |
| "logps/chosen": -244.85443115234375, | |
| "logps/rejected": -236.78126525878906, | |
| "loss": 0.5334, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.5452998876571655, | |
| "rewards/margins": 0.6891751885414124, | |
| "rewards/rejected": -0.14387527108192444, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.7792207792207793, | |
| "grad_norm": 18.481484735169893, | |
| "learning_rate": 1.4883637496547141e-08, | |
| "logits/chosen": -1.2515636682510376, | |
| "logits/rejected": -1.2902557849884033, | |
| "logps/chosen": -254.2708740234375, | |
| "logps/rejected": -258.9367980957031, | |
| "loss": 0.5094, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": 0.7975621223449707, | |
| "rewards/margins": 1.1254695653915405, | |
| "rewards/rejected": -0.327907532453537, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.7922077922077921, | |
| "grad_norm": 18.685754884224313, | |
| "learning_rate": 1.3199247427022526e-08, | |
| "logits/chosen": -1.2295678853988647, | |
| "logits/rejected": -1.2680904865264893, | |
| "logps/chosen": -196.48304748535156, | |
| "logps/rejected": -207.93121337890625, | |
| "loss": 0.5151, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.8263696432113647, | |
| "rewards/margins": 0.9237637519836426, | |
| "rewards/rejected": -0.09739402681589127, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.8051948051948052, | |
| "grad_norm": 18.498386316775253, | |
| "learning_rate": 1.16134004802681e-08, | |
| "logits/chosen": -1.0846458673477173, | |
| "logits/rejected": -0.9547269940376282, | |
| "logps/chosen": -231.32818603515625, | |
| "logps/rejected": -263.1034851074219, | |
| "loss": 0.5058, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": 0.6925534605979919, | |
| "rewards/margins": 1.1582170724868774, | |
| "rewards/rejected": -0.4656636714935303, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 17.340516668358195, | |
| "learning_rate": 1.0126756596375685e-08, | |
| "logits/chosen": -1.3053959608078003, | |
| "logits/rejected": -1.2687203884124756, | |
| "logps/chosen": -206.0025177001953, | |
| "logps/rejected": -235.99029541015625, | |
| "loss": 0.499, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.8264460563659668, | |
| "rewards/margins": 0.9402331113815308, | |
| "rewards/rejected": -0.11378702521324158, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.8311688311688312, | |
| "grad_norm": 20.777883031718197, | |
| "learning_rate": 8.739934432715034e-09, | |
| "logits/chosen": -1.2384371757507324, | |
| "logits/rejected": -1.3831499814987183, | |
| "logps/chosen": -226.1611785888672, | |
| "logps/rejected": -219.2269287109375, | |
| "loss": 0.5138, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 0.514397382736206, | |
| "rewards/margins": 0.7629111409187317, | |
| "rewards/rejected": -0.24851378798484802, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.844155844155844, | |
| "grad_norm": 18.170821830096596, | |
| "learning_rate": 7.453511106483901e-09, | |
| "logits/chosen": -1.4039347171783447, | |
| "logits/rejected": -1.4059028625488281, | |
| "logps/chosen": -269.5606689453125, | |
| "logps/rejected": -254.4373779296875, | |
| "loss": 0.5001, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": 0.7399085164070129, | |
| "rewards/margins": 1.1216527223587036, | |
| "rewards/rejected": -0.3817441165447235, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 21.929296495653098, | |
| "learning_rate": 6.268021954544095e-09, | |
| "logits/chosen": -1.2451683282852173, | |
| "logits/rejected": -1.2341443300247192, | |
| "logps/chosen": -249.87173461914062, | |
| "logps/rejected": -256.73297119140625, | |
| "loss": 0.5052, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.665867805480957, | |
| "rewards/margins": 0.8860350251197815, | |
| "rewards/rejected": -0.22016723453998566, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.87012987012987, | |
| "grad_norm": 17.66627473591946, | |
| "learning_rate": 5.1839603106447475e-09, | |
| "logits/chosen": -1.1848328113555908, | |
| "logits/rejected": -1.2601234912872314, | |
| "logps/chosen": -274.2748718261719, | |
| "logps/rejected": -276.2279968261719, | |
| "loss": 0.5143, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.40582898259162903, | |
| "rewards/margins": 0.7143059372901917, | |
| "rewards/rejected": -0.3084769546985626, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.883116883116883, | |
| "grad_norm": 17.62107555046559, | |
| "learning_rate": 4.201777300124249e-09, | |
| "logits/chosen": -1.3278290033340454, | |
| "logits/rejected": -1.412712574005127, | |
| "logps/chosen": -232.41680908203125, | |
| "logps/rejected": -261.15911865234375, | |
| "loss": 0.4936, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.6634901165962219, | |
| "rewards/margins": 1.093658685684204, | |
| "rewards/rejected": -0.43016865849494934, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.896103896103896, | |
| "grad_norm": 15.858708451261625, | |
| "learning_rate": 3.3218816521777827e-09, | |
| "logits/chosen": -1.1245518922805786, | |
| "logits/rejected": -1.1566898822784424, | |
| "logps/chosen": -204.53695678710938, | |
| "logps/rejected": -195.2686309814453, | |
| "loss": 0.5017, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.6790772080421448, | |
| "rewards/margins": 0.7251225709915161, | |
| "rewards/rejected": -0.04604536294937134, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.9090909090909092, | |
| "grad_norm": 18.1779116105005, | |
| "learning_rate": 2.5446395297668287e-09, | |
| "logits/chosen": -1.2490640878677368, | |
| "logits/rejected": -1.3185060024261475, | |
| "logps/chosen": -197.9644775390625, | |
| "logps/rejected": -211.1698455810547, | |
| "loss": 0.503, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.6239749789237976, | |
| "rewards/margins": 1.010934591293335, | |
| "rewards/rejected": -0.38695967197418213, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.922077922077922, | |
| "grad_norm": 18.8639806757367, | |
| "learning_rate": 1.870374377243078e-09, | |
| "logits/chosen": -1.2287623882293701, | |
| "logits/rejected": -1.214994192123413, | |
| "logps/chosen": -183.43939208984375, | |
| "logps/rejected": -200.90785217285156, | |
| "loss": 0.5078, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.3988313674926758, | |
| "rewards/margins": 0.920490026473999, | |
| "rewards/rejected": -0.5216587781906128, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.935064935064935, | |
| "grad_norm": 19.016439869677555, | |
| "learning_rate": 1.2993667857489898e-09, | |
| "logits/chosen": -1.365774393081665, | |
| "logits/rejected": -1.4144957065582275, | |
| "logps/chosen": -299.92120361328125, | |
| "logps/rejected": -313.76287841796875, | |
| "loss": 0.492, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.790663480758667, | |
| "rewards/margins": 1.3730117082595825, | |
| "rewards/rejected": -0.5823482275009155, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.948051948051948, | |
| "grad_norm": 14.859752452370639, | |
| "learning_rate": 8.318543764516961e-10, | |
| "logits/chosen": -1.3680145740509033, | |
| "logits/rejected": -1.3740514516830444, | |
| "logps/chosen": -264.17486572265625, | |
| "logps/rejected": -253.31149291992188, | |
| "loss": 0.5, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": 0.784439206123352, | |
| "rewards/margins": 1.083054542541504, | |
| "rewards/rejected": -0.29861530661582947, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.9610389610389611, | |
| "grad_norm": 18.397492445789904, | |
| "learning_rate": 4.680317016582669e-10, | |
| "logits/chosen": -1.3651466369628906, | |
| "logits/rejected": -1.2224037647247314, | |
| "logps/chosen": -203.9071502685547, | |
| "logps/rejected": -270.9750061035156, | |
| "loss": 0.4975, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 0.7834948301315308, | |
| "rewards/margins": 0.9753363132476807, | |
| "rewards/rejected": -0.1918414682149887, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.974025974025974, | |
| "grad_norm": 17.010218612769055, | |
| "learning_rate": 2.0805016385427865e-10, | |
| "logits/chosen": -1.2519904375076294, | |
| "logits/rejected": -1.3218713998794556, | |
| "logps/chosen": -171.00653076171875, | |
| "logps/rejected": -186.12115478515625, | |
| "loss": 0.5156, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.7086281776428223, | |
| "rewards/margins": 0.9142274856567383, | |
| "rewards/rejected": -0.20559939742088318, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.987012987012987, | |
| "grad_norm": 20.168300406467708, | |
| "learning_rate": 5.2017952698379943e-11, | |
| "logits/chosen": -1.107275128364563, | |
| "logits/rejected": -1.237258791923523, | |
| "logps/chosen": -269.6531982421875, | |
| "logps/rejected": -244.50665283203125, | |
| "loss": 0.5078, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": 0.7939411997795105, | |
| "rewards/margins": 1.0421514511108398, | |
| "rewards/rejected": -0.24821028113365173, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 19.11260041379628, | |
| "learning_rate": 0.0, | |
| "logits/chosen": -1.115605115890503, | |
| "logits/rejected": -1.0857875347137451, | |
| "logps/chosen": -228.82167053222656, | |
| "logps/rejected": -256.9181823730469, | |
| "loss": 0.5006, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.7351489067077637, | |
| "rewards/margins": 0.7865853905677795, | |
| "rewards/rejected": -0.051436468958854675, | |
| "step": 308 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 308, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |