{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.2167932662697598, "eval_steps": 88, "global_step": 264, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.000821186614658181, "grad_norm": 0.480191707611084, "learning_rate": 2e-05, "loss": 1.332, "step": 1 }, { "epoch": 0.001642373229316362, "grad_norm": 0.43048733472824097, "learning_rate": 4e-05, "loss": 1.2784, "step": 2 }, { "epoch": 0.002463559843974543, "grad_norm": 0.33739173412323, "learning_rate": 6e-05, "loss": 1.3286, "step": 3 }, { "epoch": 0.003284746458632724, "grad_norm": 0.432579904794693, "learning_rate": 8e-05, "loss": 1.3079, "step": 4 }, { "epoch": 0.0041059330732909054, "grad_norm": 0.3490436375141144, "learning_rate": 0.0001, "loss": 1.2182, "step": 5 }, { "epoch": 0.004927119687949086, "grad_norm": 0.20206260681152344, "learning_rate": 9.997257268239166e-05, "loss": 1.2828, "step": 6 }, { "epoch": 0.005748306302607267, "grad_norm": 0.15355628728866577, "learning_rate": 9.994514536478333e-05, "loss": 1.1616, "step": 7 }, { "epoch": 0.006569492917265448, "grad_norm": 0.16756780445575714, "learning_rate": 9.9917718047175e-05, "loss": 1.0935, "step": 8 }, { "epoch": 0.00739067953192363, "grad_norm": 0.17429664731025696, "learning_rate": 9.989029072956665e-05, "loss": 0.9694, "step": 9 }, { "epoch": 0.008211866146581811, "grad_norm": 0.22355175018310547, "learning_rate": 9.986286341195832e-05, "loss": 0.9407, "step": 10 }, { "epoch": 0.009033052761239991, "grad_norm": 0.33407703042030334, "learning_rate": 9.983543609434997e-05, "loss": 0.7717, "step": 11 }, { "epoch": 0.009854239375898173, "grad_norm": 0.47473278641700745, "learning_rate": 9.980800877674164e-05, "loss": 0.7763, "step": 12 }, { "epoch": 0.010675425990556354, "grad_norm": 0.2812059819698334, "learning_rate": 9.978058145913331e-05, "loss": 0.6258, "step": 13 }, { "epoch": 0.011496612605214535, "grad_norm": 0.23547925055027008, "learning_rate": 9.975315414152496e-05, "loss": 0.5968, "step": 14 }, { "epoch": 0.012317799219872716, "grad_norm": 0.18453630805015564, "learning_rate": 9.972572682391662e-05, "loss": 0.5368, "step": 15 }, { "epoch": 0.013138985834530896, "grad_norm": 0.16103577613830566, "learning_rate": 9.969829950630828e-05, "loss": 0.4909, "step": 16 }, { "epoch": 0.013960172449189078, "grad_norm": 0.18651455640792847, "learning_rate": 9.967087218869995e-05, "loss": 0.5135, "step": 17 }, { "epoch": 0.01478135906384726, "grad_norm": 0.11300642043352127, "learning_rate": 9.96434448710916e-05, "loss": 0.5069, "step": 18 }, { "epoch": 0.01560254567850544, "grad_norm": 0.10415703803300858, "learning_rate": 9.961601755348327e-05, "loss": 0.4851, "step": 19 }, { "epoch": 0.016423732293163622, "grad_norm": 0.11693017929792404, "learning_rate": 9.958859023587493e-05, "loss": 0.4625, "step": 20 }, { "epoch": 0.017244918907821802, "grad_norm": 0.10035043954849243, "learning_rate": 9.95611629182666e-05, "loss": 0.4822, "step": 21 }, { "epoch": 0.018066105522479982, "grad_norm": 0.10483390837907791, "learning_rate": 9.953373560065826e-05, "loss": 0.4342, "step": 22 }, { "epoch": 0.018887292137138165, "grad_norm": 2.8405802249908447, "learning_rate": 9.950630828304992e-05, "loss": 0.4664, "step": 23 }, { "epoch": 0.019708478751796345, "grad_norm": 0.13821998238563538, "learning_rate": 9.947888096544159e-05, "loss": 0.4468, "step": 24 }, { "epoch": 0.020529665366454525, "grad_norm": 0.1991378366947174, "learning_rate": 9.945145364783325e-05, "loss": 0.4605, "step": 25 }, { "epoch": 0.02135085198111271, "grad_norm": 0.07619134336709976, "learning_rate": 9.942402633022491e-05, "loss": 0.4597, "step": 26 }, { "epoch": 0.02217203859577089, "grad_norm": 0.13373583555221558, "learning_rate": 9.939659901261658e-05, "loss": 0.4626, "step": 27 }, { "epoch": 0.02299322521042907, "grad_norm": 0.09962721168994904, "learning_rate": 9.936917169500823e-05, "loss": 0.4638, "step": 28 }, { "epoch": 0.023814411825087253, "grad_norm": 0.09395964443683624, "learning_rate": 9.93417443773999e-05, "loss": 0.4569, "step": 29 }, { "epoch": 0.024635598439745433, "grad_norm": 0.09109952300786972, "learning_rate": 9.931431705979157e-05, "loss": 0.4439, "step": 30 }, { "epoch": 0.025456785054403613, "grad_norm": 0.10370515286922455, "learning_rate": 9.928688974218322e-05, "loss": 0.4425, "step": 31 }, { "epoch": 0.026277971669061793, "grad_norm": 0.2153477966785431, "learning_rate": 9.925946242457488e-05, "loss": 0.4503, "step": 32 }, { "epoch": 0.027099158283719976, "grad_norm": 0.08772841095924377, "learning_rate": 9.923203510696654e-05, "loss": 0.419, "step": 33 }, { "epoch": 0.027920344898378156, "grad_norm": 0.10951374471187592, "learning_rate": 9.920460778935821e-05, "loss": 0.4353, "step": 34 }, { "epoch": 0.028741531513036336, "grad_norm": 0.09190870076417923, "learning_rate": 9.917718047174987e-05, "loss": 0.5196, "step": 35 }, { "epoch": 0.02956271812769452, "grad_norm": 0.07667124271392822, "learning_rate": 9.914975315414153e-05, "loss": 0.4358, "step": 36 }, { "epoch": 0.0303839047423527, "grad_norm": 0.1514267474412918, "learning_rate": 9.912232583653319e-05, "loss": 0.411, "step": 37 }, { "epoch": 0.03120509135701088, "grad_norm": 0.09086549282073975, "learning_rate": 9.909489851892486e-05, "loss": 0.4003, "step": 38 }, { "epoch": 0.032026277971669063, "grad_norm": 0.2616782486438751, "learning_rate": 9.906747120131652e-05, "loss": 0.4842, "step": 39 }, { "epoch": 0.032847464586327244, "grad_norm": 0.11908283084630966, "learning_rate": 9.904004388370818e-05, "loss": 0.4143, "step": 40 }, { "epoch": 0.033668651200985424, "grad_norm": 0.07770542800426483, "learning_rate": 9.901261656609983e-05, "loss": 0.3873, "step": 41 }, { "epoch": 0.034489837815643604, "grad_norm": 0.08934606611728668, "learning_rate": 9.89851892484915e-05, "loss": 0.4235, "step": 42 }, { "epoch": 0.035311024430301784, "grad_norm": 0.09303563088178635, "learning_rate": 9.895776193088317e-05, "loss": 0.4103, "step": 43 }, { "epoch": 0.036132211044959964, "grad_norm": 0.08622181415557861, "learning_rate": 9.893033461327482e-05, "loss": 0.448, "step": 44 }, { "epoch": 0.03695339765961815, "grad_norm": 0.08822862058877945, "learning_rate": 9.890290729566649e-05, "loss": 0.3855, "step": 45 }, { "epoch": 0.03777458427427633, "grad_norm": 0.08557698875665665, "learning_rate": 9.887547997805814e-05, "loss": 0.3945, "step": 46 }, { "epoch": 0.03859577088893451, "grad_norm": 0.07540106773376465, "learning_rate": 9.884805266044981e-05, "loss": 0.4192, "step": 47 }, { "epoch": 0.03941695750359269, "grad_norm": 0.1023702397942543, "learning_rate": 9.882062534284148e-05, "loss": 0.4126, "step": 48 }, { "epoch": 0.04023814411825087, "grad_norm": 0.07779772579669952, "learning_rate": 9.879319802523313e-05, "loss": 0.4244, "step": 49 }, { "epoch": 0.04105933073290905, "grad_norm": 0.08826564252376556, "learning_rate": 9.876577070762479e-05, "loss": 0.415, "step": 50 }, { "epoch": 0.04188051734756724, "grad_norm": 0.08254576474428177, "learning_rate": 9.873834339001646e-05, "loss": 0.4346, "step": 51 }, { "epoch": 0.04270170396222542, "grad_norm": 0.08287151902914047, "learning_rate": 9.871091607240812e-05, "loss": 0.4142, "step": 52 }, { "epoch": 0.0435228905768836, "grad_norm": 0.08196476101875305, "learning_rate": 9.868348875479978e-05, "loss": 0.3822, "step": 53 }, { "epoch": 0.04434407719154178, "grad_norm": 0.08654092252254486, "learning_rate": 9.865606143719145e-05, "loss": 0.3937, "step": 54 }, { "epoch": 0.04516526380619996, "grad_norm": 0.1102684736251831, "learning_rate": 9.86286341195831e-05, "loss": 0.4508, "step": 55 }, { "epoch": 0.04598645042085814, "grad_norm": 0.08240954577922821, "learning_rate": 9.860120680197478e-05, "loss": 0.4069, "step": 56 }, { "epoch": 0.04680763703551632, "grad_norm": 0.08542217314243317, "learning_rate": 9.857377948436644e-05, "loss": 0.5002, "step": 57 }, { "epoch": 0.047628823650174505, "grad_norm": 0.08390172570943832, "learning_rate": 9.854635216675809e-05, "loss": 0.3851, "step": 58 }, { "epoch": 0.048450010264832685, "grad_norm": 0.10694168508052826, "learning_rate": 9.851892484914976e-05, "loss": 0.4026, "step": 59 }, { "epoch": 0.049271196879490865, "grad_norm": 0.0852806493639946, "learning_rate": 9.849149753154143e-05, "loss": 0.424, "step": 60 }, { "epoch": 0.050092383494149045, "grad_norm": 0.15425831079483032, "learning_rate": 9.846407021393308e-05, "loss": 0.4307, "step": 61 }, { "epoch": 0.050913570108807225, "grad_norm": 0.08546218276023865, "learning_rate": 9.843664289632475e-05, "loss": 0.4386, "step": 62 }, { "epoch": 0.051734756723465405, "grad_norm": 0.08588795363903046, "learning_rate": 9.84092155787164e-05, "loss": 0.4433, "step": 63 }, { "epoch": 0.052555943338123585, "grad_norm": 0.08901514858007431, "learning_rate": 9.838178826110807e-05, "loss": 0.4263, "step": 64 }, { "epoch": 0.05337712995278177, "grad_norm": 0.17668095231056213, "learning_rate": 9.835436094349974e-05, "loss": 0.407, "step": 65 }, { "epoch": 0.05419831656743995, "grad_norm": 0.081763856112957, "learning_rate": 9.83269336258914e-05, "loss": 0.374, "step": 66 }, { "epoch": 0.05501950318209813, "grad_norm": 0.09063572436571121, "learning_rate": 9.829950630828305e-05, "loss": 0.3839, "step": 67 }, { "epoch": 0.05584068979675631, "grad_norm": 0.08264392614364624, "learning_rate": 9.827207899067472e-05, "loss": 0.4676, "step": 68 }, { "epoch": 0.05666187641141449, "grad_norm": 0.0827123150229454, "learning_rate": 9.824465167306638e-05, "loss": 0.3801, "step": 69 }, { "epoch": 0.05748306302607267, "grad_norm": 0.07972189038991928, "learning_rate": 9.821722435545804e-05, "loss": 0.3758, "step": 70 }, { "epoch": 0.05830424964073085, "grad_norm": 0.08738942444324493, "learning_rate": 9.81897970378497e-05, "loss": 0.3634, "step": 71 }, { "epoch": 0.05912543625538904, "grad_norm": 0.08442792296409607, "learning_rate": 9.816236972024136e-05, "loss": 0.3803, "step": 72 }, { "epoch": 0.05994662287004722, "grad_norm": 0.2751137316226959, "learning_rate": 9.813494240263303e-05, "loss": 0.3869, "step": 73 }, { "epoch": 0.0607678094847054, "grad_norm": 0.19429296255111694, "learning_rate": 9.81075150850247e-05, "loss": 0.3992, "step": 74 }, { "epoch": 0.06158899609936358, "grad_norm": 0.08392605930566788, "learning_rate": 9.808008776741635e-05, "loss": 0.361, "step": 75 }, { "epoch": 0.06241018271402176, "grad_norm": 2.01043963432312, "learning_rate": 9.8052660449808e-05, "loss": 0.3787, "step": 76 }, { "epoch": 0.06323136932867994, "grad_norm": 0.0808538943529129, "learning_rate": 9.802523313219967e-05, "loss": 0.4013, "step": 77 }, { "epoch": 0.06405255594333813, "grad_norm": 0.07846518605947495, "learning_rate": 9.799780581459134e-05, "loss": 0.3808, "step": 78 }, { "epoch": 0.0648737425579963, "grad_norm": 0.08152970671653748, "learning_rate": 9.7970378496983e-05, "loss": 0.4076, "step": 79 }, { "epoch": 0.06569492917265449, "grad_norm": 0.07745791226625443, "learning_rate": 9.794295117937466e-05, "loss": 0.3981, "step": 80 }, { "epoch": 0.06651611578731266, "grad_norm": 0.0842173770070076, "learning_rate": 9.791552386176632e-05, "loss": 0.4469, "step": 81 }, { "epoch": 0.06733730240197085, "grad_norm": 0.07920663058757782, "learning_rate": 9.788809654415799e-05, "loss": 0.3988, "step": 82 }, { "epoch": 0.06815848901662903, "grad_norm": 0.07899456471204758, "learning_rate": 9.786066922654965e-05, "loss": 0.4418, "step": 83 }, { "epoch": 0.06897967563128721, "grad_norm": 0.07229727506637573, "learning_rate": 9.783324190894131e-05, "loss": 0.4016, "step": 84 }, { "epoch": 0.0698008622459454, "grad_norm": 0.08062436431646347, "learning_rate": 9.780581459133296e-05, "loss": 0.3762, "step": 85 }, { "epoch": 0.07062204886060357, "grad_norm": 0.08440960198640823, "learning_rate": 9.777838727372464e-05, "loss": 0.4362, "step": 86 }, { "epoch": 0.07144323547526175, "grad_norm": 0.07544733583927155, "learning_rate": 9.77509599561163e-05, "loss": 0.4135, "step": 87 }, { "epoch": 0.07226442208991993, "grad_norm": 0.1760656237602234, "learning_rate": 9.772353263850797e-05, "loss": 0.3887, "step": 88 }, { "epoch": 0.07226442208991993, "eval_runtime": 494.3052, "eval_samples_per_second": 0.399, "eval_steps_per_second": 0.2, "step": 88 }, { "epoch": 0.07308560870457811, "grad_norm": 0.08049104362726212, "learning_rate": 9.769610532089962e-05, "loss": 0.4272, "step": 89 }, { "epoch": 0.0739067953192363, "grad_norm": 0.07617965340614319, "learning_rate": 9.766867800329129e-05, "loss": 0.3873, "step": 90 }, { "epoch": 0.07472798193389447, "grad_norm": 0.07975462824106216, "learning_rate": 9.764125068568296e-05, "loss": 0.3903, "step": 91 }, { "epoch": 0.07554916854855266, "grad_norm": 0.08299189805984497, "learning_rate": 9.761382336807461e-05, "loss": 0.404, "step": 92 }, { "epoch": 0.07637035516321083, "grad_norm": 0.08578819036483765, "learning_rate": 9.758639605046626e-05, "loss": 0.3983, "step": 93 }, { "epoch": 0.07719154177786902, "grad_norm": 0.10872071981430054, "learning_rate": 9.755896873285793e-05, "loss": 0.3553, "step": 94 }, { "epoch": 0.07801272839252721, "grad_norm": 0.08121436834335327, "learning_rate": 9.75315414152496e-05, "loss": 0.408, "step": 95 }, { "epoch": 0.07883391500718538, "grad_norm": 0.09175996482372284, "learning_rate": 9.750411409764125e-05, "loss": 0.4008, "step": 96 }, { "epoch": 0.07965510162184357, "grad_norm": 0.08122789114713669, "learning_rate": 9.747668678003292e-05, "loss": 0.4045, "step": 97 }, { "epoch": 0.08047628823650174, "grad_norm": 0.08402436226606369, "learning_rate": 9.744925946242458e-05, "loss": 0.3814, "step": 98 }, { "epoch": 0.08129747485115993, "grad_norm": 0.08454861491918564, "learning_rate": 9.742183214481624e-05, "loss": 0.3904, "step": 99 }, { "epoch": 0.0821186614658181, "grad_norm": 0.08125888556241989, "learning_rate": 9.739440482720791e-05, "loss": 0.3681, "step": 100 }, { "epoch": 0.08293984808047629, "grad_norm": 0.08544078469276428, "learning_rate": 9.736697750959957e-05, "loss": 0.367, "step": 101 }, { "epoch": 0.08376103469513448, "grad_norm": 0.07539089769124985, "learning_rate": 9.733955019199122e-05, "loss": 0.3821, "step": 102 }, { "epoch": 0.08458222130979265, "grad_norm": 0.07529085874557495, "learning_rate": 9.731212287438289e-05, "loss": 0.4169, "step": 103 }, { "epoch": 0.08540340792445084, "grad_norm": 0.07588034868240356, "learning_rate": 9.728469555677456e-05, "loss": 0.4211, "step": 104 }, { "epoch": 0.08622459453910901, "grad_norm": 0.08019097149372101, "learning_rate": 9.725726823916621e-05, "loss": 0.4033, "step": 105 }, { "epoch": 0.0870457811537672, "grad_norm": 0.07878712564706802, "learning_rate": 9.722984092155788e-05, "loss": 0.3842, "step": 106 }, { "epoch": 0.08786696776842537, "grad_norm": 0.08083963394165039, "learning_rate": 9.720241360394953e-05, "loss": 0.3558, "step": 107 }, { "epoch": 0.08868815438308356, "grad_norm": 0.08340411633253098, "learning_rate": 9.71749862863412e-05, "loss": 0.4037, "step": 108 }, { "epoch": 0.08950934099774174, "grad_norm": 0.09317754209041595, "learning_rate": 9.714755896873287e-05, "loss": 0.3874, "step": 109 }, { "epoch": 0.09033052761239992, "grad_norm": 0.07143125683069229, "learning_rate": 9.712013165112452e-05, "loss": 0.3441, "step": 110 }, { "epoch": 0.0911517142270581, "grad_norm": 0.0909111350774765, "learning_rate": 9.709270433351618e-05, "loss": 0.3552, "step": 111 }, { "epoch": 0.09197290084171628, "grad_norm": 0.08041603863239288, "learning_rate": 9.706527701590785e-05, "loss": 0.394, "step": 112 }, { "epoch": 0.09279408745637446, "grad_norm": 0.08147992193698883, "learning_rate": 9.703784969829951e-05, "loss": 0.3916, "step": 113 }, { "epoch": 0.09361527407103264, "grad_norm": 0.08365318924188614, "learning_rate": 9.701042238069117e-05, "loss": 0.4075, "step": 114 }, { "epoch": 0.09443646068569082, "grad_norm": 0.083246149122715, "learning_rate": 9.698299506308284e-05, "loss": 0.3566, "step": 115 }, { "epoch": 0.09525764730034901, "grad_norm": 0.0942649245262146, "learning_rate": 9.695556774547449e-05, "loss": 0.3516, "step": 116 }, { "epoch": 0.09607883391500718, "grad_norm": 0.08110091835260391, "learning_rate": 9.692814042786616e-05, "loss": 0.3902, "step": 117 }, { "epoch": 0.09690002052966537, "grad_norm": 0.43124014139175415, "learning_rate": 9.690071311025783e-05, "loss": 0.3956, "step": 118 }, { "epoch": 0.09772120714432354, "grad_norm": 0.08204706013202667, "learning_rate": 9.687328579264948e-05, "loss": 0.3865, "step": 119 }, { "epoch": 0.09854239375898173, "grad_norm": 0.08262762427330017, "learning_rate": 9.684585847504113e-05, "loss": 0.3777, "step": 120 }, { "epoch": 0.0993635803736399, "grad_norm": 0.08233962953090668, "learning_rate": 9.681843115743282e-05, "loss": 0.3976, "step": 121 }, { "epoch": 0.10018476698829809, "grad_norm": 0.7428120970726013, "learning_rate": 9.679100383982447e-05, "loss": 0.377, "step": 122 }, { "epoch": 0.10100595360295628, "grad_norm": 0.07909400761127472, "learning_rate": 9.676357652221614e-05, "loss": 0.3654, "step": 123 }, { "epoch": 0.10182714021761445, "grad_norm": 0.08714035898447037, "learning_rate": 9.673614920460779e-05, "loss": 0.4181, "step": 124 }, { "epoch": 0.10264832683227264, "grad_norm": 0.08017311245203018, "learning_rate": 9.670872188699946e-05, "loss": 0.4052, "step": 125 }, { "epoch": 0.10346951344693081, "grad_norm": 0.13821078836917877, "learning_rate": 9.668129456939113e-05, "loss": 0.3733, "step": 126 }, { "epoch": 0.104290700061589, "grad_norm": 0.1609969437122345, "learning_rate": 9.665386725178278e-05, "loss": 0.3686, "step": 127 }, { "epoch": 0.10511188667624717, "grad_norm": 0.1280309557914734, "learning_rate": 9.662643993417444e-05, "loss": 0.4036, "step": 128 }, { "epoch": 0.10593307329090536, "grad_norm": 0.08747898787260056, "learning_rate": 9.65990126165661e-05, "loss": 0.3587, "step": 129 }, { "epoch": 0.10675425990556354, "grad_norm": 0.1252209097146988, "learning_rate": 9.657158529895777e-05, "loss": 0.3626, "step": 130 }, { "epoch": 0.10757544652022172, "grad_norm": 0.09374388307332993, "learning_rate": 9.654415798134943e-05, "loss": 0.3824, "step": 131 }, { "epoch": 0.1083966331348799, "grad_norm": 0.08730709552764893, "learning_rate": 9.65167306637411e-05, "loss": 0.3827, "step": 132 }, { "epoch": 0.10921781974953808, "grad_norm": 0.07719024270772934, "learning_rate": 9.648930334613275e-05, "loss": 0.3762, "step": 133 }, { "epoch": 0.11003900636419627, "grad_norm": 0.08843278139829636, "learning_rate": 9.646187602852442e-05, "loss": 0.3843, "step": 134 }, { "epoch": 0.11086019297885444, "grad_norm": 0.08568207919597626, "learning_rate": 9.643444871091608e-05, "loss": 0.3594, "step": 135 }, { "epoch": 0.11168137959351263, "grad_norm": 0.08556952327489853, "learning_rate": 9.640702139330774e-05, "loss": 0.3886, "step": 136 }, { "epoch": 0.11250256620817081, "grad_norm": 0.09016801416873932, "learning_rate": 9.63795940756994e-05, "loss": 0.4023, "step": 137 }, { "epoch": 0.11332375282282899, "grad_norm": 0.08133590966463089, "learning_rate": 9.635216675809106e-05, "loss": 0.3634, "step": 138 }, { "epoch": 0.11414493943748717, "grad_norm": 0.2211730182170868, "learning_rate": 9.632473944048273e-05, "loss": 0.3518, "step": 139 }, { "epoch": 0.11496612605214535, "grad_norm": 0.08816584199666977, "learning_rate": 9.629731212287438e-05, "loss": 0.3727, "step": 140 }, { "epoch": 0.11578731266680353, "grad_norm": 0.11618969589471817, "learning_rate": 9.626988480526605e-05, "loss": 0.4247, "step": 141 }, { "epoch": 0.1166084992814617, "grad_norm": 0.09243030101060867, "learning_rate": 9.62424574876577e-05, "loss": 0.396, "step": 142 }, { "epoch": 0.11742968589611989, "grad_norm": 0.08566376566886902, "learning_rate": 9.621503017004937e-05, "loss": 0.3852, "step": 143 }, { "epoch": 0.11825087251077808, "grad_norm": 0.08220973610877991, "learning_rate": 9.618760285244104e-05, "loss": 0.3961, "step": 144 }, { "epoch": 0.11907205912543625, "grad_norm": 0.08240345865488052, "learning_rate": 9.61601755348327e-05, "loss": 0.3518, "step": 145 }, { "epoch": 0.11989324574009444, "grad_norm": 0.08472532778978348, "learning_rate": 9.613274821722435e-05, "loss": 0.3586, "step": 146 }, { "epoch": 0.12071443235475261, "grad_norm": 0.08407485485076904, "learning_rate": 9.610532089961602e-05, "loss": 0.3797, "step": 147 }, { "epoch": 0.1215356189694108, "grad_norm": 0.09284385293722153, "learning_rate": 9.607789358200769e-05, "loss": 0.3499, "step": 148 }, { "epoch": 0.12235680558406897, "grad_norm": 0.08499818295240402, "learning_rate": 9.605046626439934e-05, "loss": 0.3722, "step": 149 }, { "epoch": 0.12317799219872716, "grad_norm": 0.080271415412426, "learning_rate": 9.602303894679101e-05, "loss": 0.3871, "step": 150 }, { "epoch": 0.12399917881338535, "grad_norm": 0.07850060611963272, "learning_rate": 9.599561162918266e-05, "loss": 0.3679, "step": 151 }, { "epoch": 0.12482036542804352, "grad_norm": 0.07685016840696335, "learning_rate": 9.596818431157433e-05, "loss": 0.4078, "step": 152 }, { "epoch": 0.1256415520427017, "grad_norm": 0.09402357786893845, "learning_rate": 9.5940756993966e-05, "loss": 0.3996, "step": 153 }, { "epoch": 0.12646273865735988, "grad_norm": 0.08445476740598679, "learning_rate": 9.591332967635765e-05, "loss": 0.3841, "step": 154 }, { "epoch": 0.12728392527201807, "grad_norm": 0.08233911544084549, "learning_rate": 9.588590235874932e-05, "loss": 0.3529, "step": 155 }, { "epoch": 0.12810511188667625, "grad_norm": 0.07896068692207336, "learning_rate": 9.585847504114099e-05, "loss": 0.3585, "step": 156 }, { "epoch": 0.12892629850133444, "grad_norm": 0.0822276696562767, "learning_rate": 9.583104772353264e-05, "loss": 0.3962, "step": 157 }, { "epoch": 0.1297474851159926, "grad_norm": 0.07977598905563354, "learning_rate": 9.580362040592431e-05, "loss": 0.454, "step": 158 }, { "epoch": 0.1305686717306508, "grad_norm": 0.0857616737484932, "learning_rate": 9.577619308831597e-05, "loss": 0.3954, "step": 159 }, { "epoch": 0.13138985834530897, "grad_norm": 0.0874355211853981, "learning_rate": 9.574876577070763e-05, "loss": 0.3591, "step": 160 }, { "epoch": 0.13221104495996716, "grad_norm": 0.07877468317747116, "learning_rate": 9.57213384530993e-05, "loss": 0.348, "step": 161 }, { "epoch": 0.13303223157462532, "grad_norm": 0.08618593961000443, "learning_rate": 9.569391113549096e-05, "loss": 0.3723, "step": 162 }, { "epoch": 0.1338534181892835, "grad_norm": 0.08144336938858032, "learning_rate": 9.566648381788261e-05, "loss": 0.4146, "step": 163 }, { "epoch": 0.1346746048039417, "grad_norm": 0.07322760671377182, "learning_rate": 9.563905650027428e-05, "loss": 0.3107, "step": 164 }, { "epoch": 0.13549579141859988, "grad_norm": 0.08007095754146576, "learning_rate": 9.561162918266595e-05, "loss": 0.4128, "step": 165 }, { "epoch": 0.13631697803325807, "grad_norm": 0.09636646509170532, "learning_rate": 9.55842018650576e-05, "loss": 0.4089, "step": 166 }, { "epoch": 0.13713816464791623, "grad_norm": 0.08381053060293198, "learning_rate": 9.555677454744927e-05, "loss": 0.3624, "step": 167 }, { "epoch": 0.13795935126257441, "grad_norm": 0.07476504147052765, "learning_rate": 9.552934722984092e-05, "loss": 0.3906, "step": 168 }, { "epoch": 0.1387805378772326, "grad_norm": 0.0901239663362503, "learning_rate": 9.550191991223259e-05, "loss": 0.3378, "step": 169 }, { "epoch": 0.1396017244918908, "grad_norm": 0.0813356265425682, "learning_rate": 9.547449259462426e-05, "loss": 0.3627, "step": 170 }, { "epoch": 0.14042291110654898, "grad_norm": 0.14319093525409698, "learning_rate": 9.544706527701591e-05, "loss": 0.3512, "step": 171 }, { "epoch": 0.14124409772120713, "grad_norm": 0.13329866528511047, "learning_rate": 9.541963795940757e-05, "loss": 0.3809, "step": 172 }, { "epoch": 0.14206528433586532, "grad_norm": 0.0815596953034401, "learning_rate": 9.539221064179923e-05, "loss": 0.3444, "step": 173 }, { "epoch": 0.1428864709505235, "grad_norm": 0.08646956831216812, "learning_rate": 9.53647833241909e-05, "loss": 0.3699, "step": 174 }, { "epoch": 0.1437076575651817, "grad_norm": 0.09374339133501053, "learning_rate": 9.533735600658256e-05, "loss": 0.3569, "step": 175 }, { "epoch": 0.14452884417983985, "grad_norm": 0.0834718644618988, "learning_rate": 9.530992868897422e-05, "loss": 0.3708, "step": 176 }, { "epoch": 0.14452884417983985, "eval_runtime": 493.9539, "eval_samples_per_second": 0.399, "eval_steps_per_second": 0.2, "step": 176 }, { "epoch": 0.14535003079449804, "grad_norm": 0.08648290485143661, "learning_rate": 9.528250137136588e-05, "loss": 0.379, "step": 177 }, { "epoch": 0.14617121740915623, "grad_norm": 0.08577203750610352, "learning_rate": 9.525507405375755e-05, "loss": 0.4368, "step": 178 }, { "epoch": 0.14699240402381442, "grad_norm": 0.1023576483130455, "learning_rate": 9.522764673614921e-05, "loss": 0.3553, "step": 179 }, { "epoch": 0.1478135906384726, "grad_norm": 0.08062634617090225, "learning_rate": 9.520021941854087e-05, "loss": 0.3616, "step": 180 }, { "epoch": 0.14863477725313076, "grad_norm": 0.07487751543521881, "learning_rate": 9.517279210093252e-05, "loss": 0.4413, "step": 181 }, { "epoch": 0.14945596386778895, "grad_norm": 0.07332492619752884, "learning_rate": 9.514536478332419e-05, "loss": 0.3166, "step": 182 }, { "epoch": 0.15027715048244714, "grad_norm": 0.08658608049154282, "learning_rate": 9.511793746571586e-05, "loss": 0.332, "step": 183 }, { "epoch": 0.15109833709710532, "grad_norm": 0.10461894422769547, "learning_rate": 9.509051014810751e-05, "loss": 0.3653, "step": 184 }, { "epoch": 0.1519195237117635, "grad_norm": 0.19619494676589966, "learning_rate": 9.506308283049918e-05, "loss": 0.4224, "step": 185 }, { "epoch": 0.15274071032642167, "grad_norm": 0.08313202857971191, "learning_rate": 9.503565551289084e-05, "loss": 0.3512, "step": 186 }, { "epoch": 0.15356189694107986, "grad_norm": 0.08686342090368271, "learning_rate": 9.50082281952825e-05, "loss": 0.3352, "step": 187 }, { "epoch": 0.15438308355573804, "grad_norm": 0.09020522236824036, "learning_rate": 9.498080087767417e-05, "loss": 0.4138, "step": 188 }, { "epoch": 0.15520427017039623, "grad_norm": 0.0799839124083519, "learning_rate": 9.495337356006583e-05, "loss": 0.3436, "step": 189 }, { "epoch": 0.15602545678505442, "grad_norm": 0.08851379156112671, "learning_rate": 9.49259462424575e-05, "loss": 0.3933, "step": 190 }, { "epoch": 0.15684664339971258, "grad_norm": 0.08521082252264023, "learning_rate": 9.489851892484916e-05, "loss": 0.3667, "step": 191 }, { "epoch": 0.15766783001437076, "grad_norm": 0.09019312262535095, "learning_rate": 9.487109160724082e-05, "loss": 0.3556, "step": 192 }, { "epoch": 0.15848901662902895, "grad_norm": 0.08162654936313629, "learning_rate": 9.484366428963248e-05, "loss": 0.3864, "step": 193 }, { "epoch": 0.15931020324368714, "grad_norm": 0.08963490277528763, "learning_rate": 9.481623697202414e-05, "loss": 0.391, "step": 194 }, { "epoch": 0.1601313898583453, "grad_norm": 0.10057719051837921, "learning_rate": 9.47888096544158e-05, "loss": 0.3599, "step": 195 }, { "epoch": 0.16095257647300348, "grad_norm": 0.1694149523973465, "learning_rate": 9.476138233680747e-05, "loss": 0.3612, "step": 196 }, { "epoch": 0.16177376308766167, "grad_norm": 0.08255323767662048, "learning_rate": 9.473395501919913e-05, "loss": 0.3441, "step": 197 }, { "epoch": 0.16259494970231986, "grad_norm": 0.09228333085775375, "learning_rate": 9.470652770159078e-05, "loss": 0.3432, "step": 198 }, { "epoch": 0.16341613631697804, "grad_norm": 0.11736617982387543, "learning_rate": 9.467910038398245e-05, "loss": 0.3408, "step": 199 }, { "epoch": 0.1642373229316362, "grad_norm": 0.16636626422405243, "learning_rate": 9.465167306637412e-05, "loss": 0.3639, "step": 200 }, { "epoch": 0.1650585095462944, "grad_norm": 0.0912085771560669, "learning_rate": 9.462424574876577e-05, "loss": 0.4055, "step": 201 }, { "epoch": 0.16587969616095258, "grad_norm": 0.08452475070953369, "learning_rate": 9.459681843115744e-05, "loss": 0.3497, "step": 202 }, { "epoch": 0.16670088277561076, "grad_norm": 0.09225429594516754, "learning_rate": 9.45693911135491e-05, "loss": 0.3765, "step": 203 }, { "epoch": 0.16752206939026895, "grad_norm": 0.08500406891107559, "learning_rate": 9.454196379594076e-05, "loss": 0.3622, "step": 204 }, { "epoch": 0.1683432560049271, "grad_norm": 0.08277002722024918, "learning_rate": 9.451453647833243e-05, "loss": 0.3596, "step": 205 }, { "epoch": 0.1691644426195853, "grad_norm": 0.0855122059583664, "learning_rate": 9.448710916072408e-05, "loss": 0.3495, "step": 206 }, { "epoch": 0.16998562923424348, "grad_norm": 0.11409081518650055, "learning_rate": 9.445968184311574e-05, "loss": 0.3123, "step": 207 }, { "epoch": 0.17080681584890167, "grad_norm": 0.08898866921663284, "learning_rate": 9.443225452550741e-05, "loss": 0.3383, "step": 208 }, { "epoch": 0.17162800246355983, "grad_norm": 0.07967101782560349, "learning_rate": 9.440482720789908e-05, "loss": 0.3927, "step": 209 }, { "epoch": 0.17244918907821802, "grad_norm": 0.08267655968666077, "learning_rate": 9.437739989029073e-05, "loss": 0.3503, "step": 210 }, { "epoch": 0.1732703756928762, "grad_norm": 0.1024966835975647, "learning_rate": 9.43499725726824e-05, "loss": 0.343, "step": 211 }, { "epoch": 0.1740915623075344, "grad_norm": 0.11825034767389297, "learning_rate": 9.432254525507405e-05, "loss": 0.3694, "step": 212 }, { "epoch": 0.17491274892219258, "grad_norm": 0.08225111663341522, "learning_rate": 9.429511793746572e-05, "loss": 0.3625, "step": 213 }, { "epoch": 0.17573393553685074, "grad_norm": 0.082975834608078, "learning_rate": 9.426769061985739e-05, "loss": 0.3692, "step": 214 }, { "epoch": 0.17655512215150893, "grad_norm": 0.09659875184297562, "learning_rate": 9.424026330224904e-05, "loss": 0.3857, "step": 215 }, { "epoch": 0.1773763087661671, "grad_norm": 0.08381886035203934, "learning_rate": 9.42128359846407e-05, "loss": 0.3658, "step": 216 }, { "epoch": 0.1781974953808253, "grad_norm": 0.08097488433122635, "learning_rate": 9.418540866703238e-05, "loss": 0.3556, "step": 217 }, { "epoch": 0.1790186819954835, "grad_norm": 0.08661879599094391, "learning_rate": 9.415798134942403e-05, "loss": 0.3495, "step": 218 }, { "epoch": 0.17983986861014165, "grad_norm": 0.09621778875589371, "learning_rate": 9.413055403181569e-05, "loss": 0.3488, "step": 219 }, { "epoch": 0.18066105522479983, "grad_norm": 0.0867924615740776, "learning_rate": 9.410312671420735e-05, "loss": 0.3496, "step": 220 }, { "epoch": 0.18148224183945802, "grad_norm": 0.09928230196237564, "learning_rate": 9.407569939659901e-05, "loss": 0.411, "step": 221 }, { "epoch": 0.1823034284541162, "grad_norm": 0.08545473217964172, "learning_rate": 9.404827207899069e-05, "loss": 0.3707, "step": 222 }, { "epoch": 0.18312461506877437, "grad_norm": 0.1317296177148819, "learning_rate": 9.402084476138234e-05, "loss": 0.4021, "step": 223 }, { "epoch": 0.18394580168343255, "grad_norm": 0.09009065479040146, "learning_rate": 9.3993417443774e-05, "loss": 0.3451, "step": 224 }, { "epoch": 0.18476698829809074, "grad_norm": 0.08332253247499466, "learning_rate": 9.396599012616567e-05, "loss": 0.3457, "step": 225 }, { "epoch": 0.18558817491274893, "grad_norm": 0.08737312257289886, "learning_rate": 9.393856280855733e-05, "loss": 0.3721, "step": 226 }, { "epoch": 0.1864093615274071, "grad_norm": 0.07472239434719086, "learning_rate": 9.391113549094899e-05, "loss": 0.3542, "step": 227 }, { "epoch": 0.18723054814206527, "grad_norm": 0.08319877088069916, "learning_rate": 9.388370817334066e-05, "loss": 0.361, "step": 228 }, { "epoch": 0.18805173475672346, "grad_norm": 0.08192326873540878, "learning_rate": 9.385628085573231e-05, "loss": 0.344, "step": 229 }, { "epoch": 0.18887292137138165, "grad_norm": 0.11502642929553986, "learning_rate": 9.382885353812398e-05, "loss": 0.363, "step": 230 }, { "epoch": 0.18969410798603983, "grad_norm": 0.08567750453948975, "learning_rate": 9.380142622051565e-05, "loss": 0.3565, "step": 231 }, { "epoch": 0.19051529460069802, "grad_norm": 0.08059141039848328, "learning_rate": 9.37739989029073e-05, "loss": 0.3722, "step": 232 }, { "epoch": 0.19133648121535618, "grad_norm": 0.07670270651578903, "learning_rate": 9.374657158529896e-05, "loss": 0.3493, "step": 233 }, { "epoch": 0.19215766783001437, "grad_norm": 0.08022642135620117, "learning_rate": 9.371914426769062e-05, "loss": 0.4142, "step": 234 }, { "epoch": 0.19297885444467255, "grad_norm": 0.08015397936105728, "learning_rate": 9.369171695008229e-05, "loss": 0.3442, "step": 235 }, { "epoch": 0.19380004105933074, "grad_norm": 0.08329442143440247, "learning_rate": 9.366428963247395e-05, "loss": 0.3587, "step": 236 }, { "epoch": 0.1946212276739889, "grad_norm": 0.08311276882886887, "learning_rate": 9.363686231486561e-05, "loss": 0.3792, "step": 237 }, { "epoch": 0.1954424142886471, "grad_norm": 0.1332862675189972, "learning_rate": 9.360943499725727e-05, "loss": 0.3451, "step": 238 }, { "epoch": 0.19626360090330527, "grad_norm": 0.081804558634758, "learning_rate": 9.358200767964894e-05, "loss": 0.4212, "step": 239 }, { "epoch": 0.19708478751796346, "grad_norm": 0.08397019654512405, "learning_rate": 9.35545803620406e-05, "loss": 0.3482, "step": 240 }, { "epoch": 0.19790597413262165, "grad_norm": 0.09495637565851212, "learning_rate": 9.352715304443226e-05, "loss": 0.3432, "step": 241 }, { "epoch": 0.1987271607472798, "grad_norm": 0.09187504649162292, "learning_rate": 9.349972572682391e-05, "loss": 0.3938, "step": 242 }, { "epoch": 0.199548347361938, "grad_norm": 0.100834921002388, "learning_rate": 9.347229840921558e-05, "loss": 0.3567, "step": 243 }, { "epoch": 0.20036953397659618, "grad_norm": 0.09292273223400116, "learning_rate": 9.344487109160725e-05, "loss": 0.3802, "step": 244 }, { "epoch": 0.20119072059125437, "grad_norm": 0.07785986363887787, "learning_rate": 9.34174437739989e-05, "loss": 0.3832, "step": 245 }, { "epoch": 0.20201190720591256, "grad_norm": 0.081189826130867, "learning_rate": 9.339001645639057e-05, "loss": 0.3448, "step": 246 }, { "epoch": 0.20283309382057071, "grad_norm": 0.08209879696369171, "learning_rate": 9.336258913878222e-05, "loss": 0.3602, "step": 247 }, { "epoch": 0.2036542804352289, "grad_norm": 0.0813421905040741, "learning_rate": 9.333516182117389e-05, "loss": 0.3477, "step": 248 }, { "epoch": 0.2044754670498871, "grad_norm": 0.08188773691654205, "learning_rate": 9.330773450356556e-05, "loss": 0.3466, "step": 249 }, { "epoch": 0.20529665366454528, "grad_norm": 0.09966633468866348, "learning_rate": 9.328030718595721e-05, "loss": 0.3776, "step": 250 }, { "epoch": 0.20611784027920346, "grad_norm": 0.08154954016208649, "learning_rate": 9.325287986834887e-05, "loss": 0.3727, "step": 251 }, { "epoch": 0.20693902689386162, "grad_norm": 0.07942931354045868, "learning_rate": 9.322545255074055e-05, "loss": 0.3468, "step": 252 }, { "epoch": 0.2077602135085198, "grad_norm": 0.07944433391094208, "learning_rate": 9.31980252331322e-05, "loss": 0.3512, "step": 253 }, { "epoch": 0.208581400123178, "grad_norm": 0.08637971431016922, "learning_rate": 9.317059791552387e-05, "loss": 0.3397, "step": 254 }, { "epoch": 0.20940258673783618, "grad_norm": 0.08189195394515991, "learning_rate": 9.314317059791553e-05, "loss": 0.4105, "step": 255 }, { "epoch": 0.21022377335249434, "grad_norm": 0.08190836012363434, "learning_rate": 9.31157432803072e-05, "loss": 0.346, "step": 256 }, { "epoch": 0.21104495996715253, "grad_norm": 0.08331865072250366, "learning_rate": 9.308831596269886e-05, "loss": 0.3655, "step": 257 }, { "epoch": 0.21186614658181072, "grad_norm": 0.0822620540857315, "learning_rate": 9.306088864509052e-05, "loss": 0.372, "step": 258 }, { "epoch": 0.2126873331964689, "grad_norm": 0.08011777698993683, "learning_rate": 9.303346132748217e-05, "loss": 0.3304, "step": 259 }, { "epoch": 0.2135085198111271, "grad_norm": 0.08395062386989594, "learning_rate": 9.300603400987384e-05, "loss": 0.3903, "step": 260 }, { "epoch": 0.21432970642578525, "grad_norm": 0.08101452887058258, "learning_rate": 9.297860669226551e-05, "loss": 0.3579, "step": 261 }, { "epoch": 0.21515089304044344, "grad_norm": 0.11504925042390823, "learning_rate": 9.295117937465716e-05, "loss": 0.3723, "step": 262 }, { "epoch": 0.21597207965510162, "grad_norm": 0.0704338401556015, "learning_rate": 9.292375205704883e-05, "loss": 0.3341, "step": 263 }, { "epoch": 0.2167932662697598, "grad_norm": 0.07944470643997192, "learning_rate": 9.289632473944048e-05, "loss": 0.3273, "step": 264 }, { "epoch": 0.2167932662697598, "eval_runtime": 507.3366, "eval_samples_per_second": 0.388, "eval_steps_per_second": 0.195, "step": 264 } ], "logging_steps": 1, "max_steps": 3651, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 88, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 1.1637146798818787e+19, "train_batch_size": 4, "trial_name": null, "trial_params": null }