{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.903708523096942, "eval_steps": 500, "global_step": 3000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0013012361743656475, "grad_norm": 17.517343521118164, "learning_rate": 5e-06, "loss": 5.0519, "step": 1 }, { "epoch": 0.002602472348731295, "grad_norm": 18.685779571533203, "learning_rate": 1e-05, "loss": 5.0202, "step": 2 }, { "epoch": 0.003903708523096942, "grad_norm": 16.044301986694336, "learning_rate": 1.5e-05, "loss": 4.8842, "step": 3 }, { "epoch": 0.00520494469746259, "grad_norm": 10.560856819152832, "learning_rate": 2e-05, "loss": 4.2169, "step": 4 }, { "epoch": 0.006506180871828237, "grad_norm": 8.9611177444458, "learning_rate": 2.5e-05, "loss": 3.6082, "step": 5 }, { "epoch": 0.007807417046193884, "grad_norm": 8.003009796142578, "learning_rate": 3e-05, "loss": 3.9438, "step": 6 }, { "epoch": 0.009108653220559532, "grad_norm": 6.665566444396973, "learning_rate": 3.5e-05, "loss": 3.1167, "step": 7 }, { "epoch": 0.01040988939492518, "grad_norm": 6.121012210845947, "learning_rate": 4e-05, "loss": 2.9282, "step": 8 }, { "epoch": 0.011711125569290826, "grad_norm": 5.4443793296813965, "learning_rate": 4.5e-05, "loss": 2.8338, "step": 9 }, { "epoch": 0.013012361743656473, "grad_norm": 4.207590103149414, "learning_rate": 5e-05, "loss": 2.7779, "step": 10 }, { "epoch": 0.014313597918022121, "grad_norm": 3.760772705078125, "learning_rate": 5.500000000000001e-05, "loss": 2.7435, "step": 11 }, { "epoch": 0.015614834092387769, "grad_norm": 3.7939486503601074, "learning_rate": 6e-05, "loss": 2.8161, "step": 12 }, { "epoch": 0.016916070266753416, "grad_norm": 2.754817485809326, "learning_rate": 6.500000000000001e-05, "loss": 2.6686, "step": 13 }, { "epoch": 0.018217306441119064, "grad_norm": 2.911052703857422, "learning_rate": 7e-05, "loss": 2.71, "step": 14 }, { "epoch": 0.01951854261548471, "grad_norm": 3.6347625255584717, "learning_rate": 7.500000000000001e-05, "loss": 2.3657, "step": 15 }, { "epoch": 0.02081977878985036, "grad_norm": 3.8363871574401855, "learning_rate": 8e-05, "loss": 2.5253, "step": 16 }, { "epoch": 0.022121014964216004, "grad_norm": 2.4171555042266846, "learning_rate": 8.5e-05, "loss": 2.6132, "step": 17 }, { "epoch": 0.02342225113858165, "grad_norm": 1.639997959136963, "learning_rate": 9e-05, "loss": 2.3991, "step": 18 }, { "epoch": 0.0247234873129473, "grad_norm": 1.5733214616775513, "learning_rate": 9.5e-05, "loss": 2.2322, "step": 19 }, { "epoch": 0.026024723487312947, "grad_norm": 1.9828444719314575, "learning_rate": 0.0001, "loss": 2.5436, "step": 20 }, { "epoch": 0.027325959661678594, "grad_norm": 1.9971580505371094, "learning_rate": 9.996723460026212e-05, "loss": 2.5574, "step": 21 }, { "epoch": 0.028627195836044242, "grad_norm": 1.7128098011016846, "learning_rate": 9.993446920052425e-05, "loss": 2.3003, "step": 22 }, { "epoch": 0.02992843201040989, "grad_norm": 1.710244059562683, "learning_rate": 9.990170380078637e-05, "loss": 2.3337, "step": 23 }, { "epoch": 0.031229668184775537, "grad_norm": 1.7096962928771973, "learning_rate": 9.986893840104849e-05, "loss": 2.049, "step": 24 }, { "epoch": 0.03253090435914118, "grad_norm": 1.4927719831466675, "learning_rate": 9.983617300131062e-05, "loss": 2.0908, "step": 25 }, { "epoch": 0.03383214053350683, "grad_norm": 3.1123430728912354, "learning_rate": 9.980340760157274e-05, "loss": 1.985, "step": 26 }, { "epoch": 0.03513337670787248, "grad_norm": 1.1555583477020264, "learning_rate": 9.977064220183486e-05, "loss": 2.2549, "step": 27 }, { "epoch": 0.03643461288223813, "grad_norm": 1.1390047073364258, "learning_rate": 9.973787680209699e-05, "loss": 2.3801, "step": 28 }, { "epoch": 0.03773584905660377, "grad_norm": 1.011423945426941, "learning_rate": 9.970511140235911e-05, "loss": 2.2661, "step": 29 }, { "epoch": 0.03903708523096942, "grad_norm": 1.5627050399780273, "learning_rate": 9.967234600262124e-05, "loss": 1.7911, "step": 30 }, { "epoch": 0.04033832140533507, "grad_norm": 1.0536160469055176, "learning_rate": 9.963958060288337e-05, "loss": 2.3433, "step": 31 }, { "epoch": 0.04163955757970072, "grad_norm": 1.0806959867477417, "learning_rate": 9.960681520314549e-05, "loss": 2.2874, "step": 32 }, { "epoch": 0.04294079375406636, "grad_norm": 1.3830180168151855, "learning_rate": 9.95740498034076e-05, "loss": 2.2553, "step": 33 }, { "epoch": 0.04424202992843201, "grad_norm": 1.773996114730835, "learning_rate": 9.954128440366974e-05, "loss": 2.2893, "step": 34 }, { "epoch": 0.04554326610279766, "grad_norm": 1.5383678674697876, "learning_rate": 9.950851900393186e-05, "loss": 2.4157, "step": 35 }, { "epoch": 0.0468445022771633, "grad_norm": 1.4748953580856323, "learning_rate": 9.947575360419397e-05, "loss": 2.2163, "step": 36 }, { "epoch": 0.048145738451528954, "grad_norm": 1.0089681148529053, "learning_rate": 9.94429882044561e-05, "loss": 2.3735, "step": 37 }, { "epoch": 0.0494469746258946, "grad_norm": 1.5883677005767822, "learning_rate": 9.941022280471822e-05, "loss": 1.9901, "step": 38 }, { "epoch": 0.05074821080026025, "grad_norm": 1.1488697528839111, "learning_rate": 9.937745740498034e-05, "loss": 2.2269, "step": 39 }, { "epoch": 0.05204944697462589, "grad_norm": 1.1208925247192383, "learning_rate": 9.934469200524247e-05, "loss": 2.0623, "step": 40 }, { "epoch": 0.053350683148991544, "grad_norm": 0.8709912300109863, "learning_rate": 9.931192660550459e-05, "loss": 1.9676, "step": 41 }, { "epoch": 0.05465191932335719, "grad_norm": 1.2127772569656372, "learning_rate": 9.927916120576671e-05, "loss": 2.1904, "step": 42 }, { "epoch": 0.05595315549772284, "grad_norm": 0.8526835441589355, "learning_rate": 9.924639580602884e-05, "loss": 2.2763, "step": 43 }, { "epoch": 0.057254391672088484, "grad_norm": 1.016025424003601, "learning_rate": 9.921363040629096e-05, "loss": 1.7844, "step": 44 }, { "epoch": 0.05855562784645413, "grad_norm": 1.1103867292404175, "learning_rate": 9.918086500655308e-05, "loss": 2.0452, "step": 45 }, { "epoch": 0.05985686402081978, "grad_norm": 2.3868448734283447, "learning_rate": 9.914809960681521e-05, "loss": 2.2632, "step": 46 }, { "epoch": 0.06115810019518542, "grad_norm": 1.1302167177200317, "learning_rate": 9.911533420707733e-05, "loss": 2.1762, "step": 47 }, { "epoch": 0.062459336369551074, "grad_norm": 1.224557638168335, "learning_rate": 9.908256880733946e-05, "loss": 2.3639, "step": 48 }, { "epoch": 0.06376057254391672, "grad_norm": 1.4134968519210815, "learning_rate": 9.904980340760159e-05, "loss": 1.8443, "step": 49 }, { "epoch": 0.06506180871828236, "grad_norm": 1.0528838634490967, "learning_rate": 9.90170380078637e-05, "loss": 2.3078, "step": 50 }, { "epoch": 0.06636304489264802, "grad_norm": 0.8730337023735046, "learning_rate": 9.898427260812582e-05, "loss": 2.1224, "step": 51 }, { "epoch": 0.06766428106701367, "grad_norm": 0.9286751747131348, "learning_rate": 9.895150720838796e-05, "loss": 2.0625, "step": 52 }, { "epoch": 0.06896551724137931, "grad_norm": 0.7579008936882019, "learning_rate": 9.891874180865007e-05, "loss": 2.1307, "step": 53 }, { "epoch": 0.07026675341574495, "grad_norm": 1.0247118473052979, "learning_rate": 9.888597640891219e-05, "loss": 2.577, "step": 54 }, { "epoch": 0.07156798959011061, "grad_norm": 0.8918821811676025, "learning_rate": 9.885321100917432e-05, "loss": 2.2683, "step": 55 }, { "epoch": 0.07286922576447626, "grad_norm": 0.8160094022750854, "learning_rate": 9.882044560943644e-05, "loss": 2.2259, "step": 56 }, { "epoch": 0.0741704619388419, "grad_norm": 0.8922228813171387, "learning_rate": 9.878768020969856e-05, "loss": 2.1132, "step": 57 }, { "epoch": 0.07547169811320754, "grad_norm": 0.9422688484191895, "learning_rate": 9.875491480996069e-05, "loss": 2.2051, "step": 58 }, { "epoch": 0.07677293428757319, "grad_norm": 1.3098868131637573, "learning_rate": 9.872214941022281e-05, "loss": 2.0608, "step": 59 }, { "epoch": 0.07807417046193885, "grad_norm": 0.9947183728218079, "learning_rate": 9.868938401048493e-05, "loss": 1.8803, "step": 60 }, { "epoch": 0.07937540663630449, "grad_norm": 0.7552891969680786, "learning_rate": 9.865661861074706e-05, "loss": 1.7667, "step": 61 }, { "epoch": 0.08067664281067013, "grad_norm": 0.8031612634658813, "learning_rate": 9.862385321100918e-05, "loss": 1.6942, "step": 62 }, { "epoch": 0.08197787898503578, "grad_norm": 2.6570839881896973, "learning_rate": 9.85910878112713e-05, "loss": 2.1892, "step": 63 }, { "epoch": 0.08327911515940144, "grad_norm": 2.479191303253174, "learning_rate": 9.855832241153343e-05, "loss": 2.155, "step": 64 }, { "epoch": 0.08458035133376708, "grad_norm": 1.8286980390548706, "learning_rate": 9.852555701179554e-05, "loss": 2.3632, "step": 65 }, { "epoch": 0.08588158750813273, "grad_norm": 1.4002997875213623, "learning_rate": 9.849279161205766e-05, "loss": 1.7372, "step": 66 }, { "epoch": 0.08718282368249837, "grad_norm": 0.8476343750953674, "learning_rate": 9.84600262123198e-05, "loss": 1.9668, "step": 67 }, { "epoch": 0.08848405985686401, "grad_norm": 0.8929819464683533, "learning_rate": 9.842726081258192e-05, "loss": 1.6921, "step": 68 }, { "epoch": 0.08978529603122967, "grad_norm": 0.8034639358520508, "learning_rate": 9.839449541284404e-05, "loss": 1.9668, "step": 69 }, { "epoch": 0.09108653220559532, "grad_norm": 2.7729508876800537, "learning_rate": 9.836173001310617e-05, "loss": 2.3142, "step": 70 }, { "epoch": 0.09238776837996096, "grad_norm": 0.8077794909477234, "learning_rate": 9.832896461336829e-05, "loss": 2.3338, "step": 71 }, { "epoch": 0.0936890045543266, "grad_norm": 0.8827224373817444, "learning_rate": 9.829619921363041e-05, "loss": 2.3245, "step": 72 }, { "epoch": 0.09499024072869226, "grad_norm": 0.7031412124633789, "learning_rate": 9.826343381389254e-05, "loss": 2.0289, "step": 73 }, { "epoch": 0.09629147690305791, "grad_norm": 0.7389160394668579, "learning_rate": 9.823066841415466e-05, "loss": 2.0043, "step": 74 }, { "epoch": 0.09759271307742355, "grad_norm": 1.0655488967895508, "learning_rate": 9.819790301441678e-05, "loss": 2.3617, "step": 75 }, { "epoch": 0.0988939492517892, "grad_norm": 0.8107008934020996, "learning_rate": 9.816513761467891e-05, "loss": 2.2833, "step": 76 }, { "epoch": 0.10019518542615485, "grad_norm": 0.7775471210479736, "learning_rate": 9.813237221494103e-05, "loss": 2.0068, "step": 77 }, { "epoch": 0.1014964216005205, "grad_norm": 0.9807071685791016, "learning_rate": 9.809960681520314e-05, "loss": 2.0661, "step": 78 }, { "epoch": 0.10279765777488614, "grad_norm": 1.0619250535964966, "learning_rate": 9.806684141546528e-05, "loss": 2.4224, "step": 79 }, { "epoch": 0.10409889394925179, "grad_norm": 0.8782541751861572, "learning_rate": 9.80340760157274e-05, "loss": 1.9562, "step": 80 }, { "epoch": 0.10540013012361743, "grad_norm": 4.051916122436523, "learning_rate": 9.800131061598951e-05, "loss": 2.1436, "step": 81 }, { "epoch": 0.10670136629798309, "grad_norm": 0.8316817879676819, "learning_rate": 9.796854521625164e-05, "loss": 2.1561, "step": 82 }, { "epoch": 0.10800260247234873, "grad_norm": 1.0690973997116089, "learning_rate": 9.793577981651376e-05, "loss": 2.151, "step": 83 }, { "epoch": 0.10930383864671438, "grad_norm": 0.9902814626693726, "learning_rate": 9.790301441677588e-05, "loss": 2.0915, "step": 84 }, { "epoch": 0.11060507482108002, "grad_norm": 0.8110764026641846, "learning_rate": 9.787024901703801e-05, "loss": 2.1923, "step": 85 }, { "epoch": 0.11190631099544568, "grad_norm": 2.869603157043457, "learning_rate": 9.783748361730014e-05, "loss": 2.2611, "step": 86 }, { "epoch": 0.11320754716981132, "grad_norm": 3.8598458766937256, "learning_rate": 9.780471821756226e-05, "loss": 2.1222, "step": 87 }, { "epoch": 0.11450878334417697, "grad_norm": 0.7530633807182312, "learning_rate": 9.777195281782439e-05, "loss": 2.0086, "step": 88 }, { "epoch": 0.11581001951854261, "grad_norm": 1.5288785696029663, "learning_rate": 9.773918741808651e-05, "loss": 2.3896, "step": 89 }, { "epoch": 0.11711125569290826, "grad_norm": 1.950155258178711, "learning_rate": 9.770642201834863e-05, "loss": 2.252, "step": 90 }, { "epoch": 0.11841249186727391, "grad_norm": 0.8893597722053528, "learning_rate": 9.767365661861076e-05, "loss": 2.085, "step": 91 }, { "epoch": 0.11971372804163956, "grad_norm": 1.4703004360198975, "learning_rate": 9.764089121887288e-05, "loss": 1.8911, "step": 92 }, { "epoch": 0.1210149642160052, "grad_norm": 0.7792260050773621, "learning_rate": 9.7608125819135e-05, "loss": 1.9686, "step": 93 }, { "epoch": 0.12231620039037085, "grad_norm": 0.6849901080131531, "learning_rate": 9.757536041939713e-05, "loss": 2.029, "step": 94 }, { "epoch": 0.1236174365647365, "grad_norm": 0.994543731212616, "learning_rate": 9.754259501965924e-05, "loss": 2.0933, "step": 95 }, { "epoch": 0.12491867273910215, "grad_norm": 1.1104122400283813, "learning_rate": 9.750982961992136e-05, "loss": 2.2133, "step": 96 }, { "epoch": 0.1262199089134678, "grad_norm": 0.8785324096679688, "learning_rate": 9.74770642201835e-05, "loss": 2.1046, "step": 97 }, { "epoch": 0.12752114508783344, "grad_norm": 1.6333246231079102, "learning_rate": 9.744429882044561e-05, "loss": 2.2746, "step": 98 }, { "epoch": 0.12882238126219908, "grad_norm": 1.1150643825531006, "learning_rate": 9.741153342070773e-05, "loss": 2.2105, "step": 99 }, { "epoch": 0.13012361743656473, "grad_norm": 0.6662530899047852, "learning_rate": 9.737876802096986e-05, "loss": 2.1218, "step": 100 }, { "epoch": 0.13142485361093037, "grad_norm": 0.7546570301055908, "learning_rate": 9.734600262123198e-05, "loss": 1.8761, "step": 101 }, { "epoch": 0.13272608978529604, "grad_norm": 1.2329723834991455, "learning_rate": 9.73132372214941e-05, "loss": 1.1596, "step": 102 }, { "epoch": 0.1340273259596617, "grad_norm": 1.5249181985855103, "learning_rate": 9.728047182175623e-05, "loss": 1.9452, "step": 103 }, { "epoch": 0.13532856213402733, "grad_norm": 0.7947364449501038, "learning_rate": 9.724770642201836e-05, "loss": 1.9494, "step": 104 }, { "epoch": 0.13662979830839297, "grad_norm": 0.7596271634101868, "learning_rate": 9.721494102228048e-05, "loss": 2.3542, "step": 105 }, { "epoch": 0.13793103448275862, "grad_norm": 1.9511692523956299, "learning_rate": 9.718217562254261e-05, "loss": 2.2319, "step": 106 }, { "epoch": 0.13923227065712426, "grad_norm": 0.8536252379417419, "learning_rate": 9.714941022280473e-05, "loss": 2.2582, "step": 107 }, { "epoch": 0.1405335068314899, "grad_norm": 0.7407233119010925, "learning_rate": 9.711664482306685e-05, "loss": 1.8296, "step": 108 }, { "epoch": 0.14183474300585555, "grad_norm": 0.7386873364448547, "learning_rate": 9.708387942332898e-05, "loss": 2.2545, "step": 109 }, { "epoch": 0.14313597918022122, "grad_norm": 1.4318211078643799, "learning_rate": 9.70511140235911e-05, "loss": 2.0058, "step": 110 }, { "epoch": 0.14443721535458687, "grad_norm": 0.8044078946113586, "learning_rate": 9.701834862385321e-05, "loss": 2.4069, "step": 111 }, { "epoch": 0.1457384515289525, "grad_norm": 0.7449604868888855, "learning_rate": 9.698558322411535e-05, "loss": 2.0807, "step": 112 }, { "epoch": 0.14703968770331816, "grad_norm": 1.4405558109283447, "learning_rate": 9.695281782437746e-05, "loss": 1.9744, "step": 113 }, { "epoch": 0.1483409238776838, "grad_norm": 1.1529614925384521, "learning_rate": 9.692005242463958e-05, "loss": 2.1042, "step": 114 }, { "epoch": 0.14964216005204944, "grad_norm": 3.0407700538635254, "learning_rate": 9.688728702490171e-05, "loss": 2.238, "step": 115 }, { "epoch": 0.1509433962264151, "grad_norm": 0.6912652850151062, "learning_rate": 9.685452162516383e-05, "loss": 2.0615, "step": 116 }, { "epoch": 0.15224463240078073, "grad_norm": 1.0139275789260864, "learning_rate": 9.682175622542595e-05, "loss": 2.0184, "step": 117 }, { "epoch": 0.15354586857514638, "grad_norm": 1.0847564935684204, "learning_rate": 9.678899082568808e-05, "loss": 2.0705, "step": 118 }, { "epoch": 0.15484710474951205, "grad_norm": 1.3264814615249634, "learning_rate": 9.67562254259502e-05, "loss": 2.2593, "step": 119 }, { "epoch": 0.1561483409238777, "grad_norm": 0.7898878455162048, "learning_rate": 9.672346002621232e-05, "loss": 2.2445, "step": 120 }, { "epoch": 0.15744957709824334, "grad_norm": 0.8438037037849426, "learning_rate": 9.669069462647445e-05, "loss": 1.7945, "step": 121 }, { "epoch": 0.15875081327260898, "grad_norm": 1.286659598350525, "learning_rate": 9.665792922673657e-05, "loss": 2.0522, "step": 122 }, { "epoch": 0.16005204944697463, "grad_norm": 1.4849517345428467, "learning_rate": 9.66251638269987e-05, "loss": 2.2513, "step": 123 }, { "epoch": 0.16135328562134027, "grad_norm": 0.7801520824432373, "learning_rate": 9.659239842726083e-05, "loss": 2.22, "step": 124 }, { "epoch": 0.16265452179570591, "grad_norm": 0.8621413111686707, "learning_rate": 9.655963302752295e-05, "loss": 2.2404, "step": 125 }, { "epoch": 0.16395575797007156, "grad_norm": 0.7480095624923706, "learning_rate": 9.652686762778506e-05, "loss": 1.9914, "step": 126 }, { "epoch": 0.1652569941444372, "grad_norm": 1.5666066408157349, "learning_rate": 9.64941022280472e-05, "loss": 2.2681, "step": 127 }, { "epoch": 0.16655823031880287, "grad_norm": 0.6715552806854248, "learning_rate": 9.646133682830931e-05, "loss": 2.5731, "step": 128 }, { "epoch": 0.16785946649316852, "grad_norm": 1.2249606847763062, "learning_rate": 9.642857142857143e-05, "loss": 2.1593, "step": 129 }, { "epoch": 0.16916070266753416, "grad_norm": 0.7186049818992615, "learning_rate": 9.639580602883356e-05, "loss": 2.207, "step": 130 }, { "epoch": 0.1704619388418998, "grad_norm": 0.760864794254303, "learning_rate": 9.636304062909568e-05, "loss": 2.1475, "step": 131 }, { "epoch": 0.17176317501626545, "grad_norm": 1.0262935161590576, "learning_rate": 9.63302752293578e-05, "loss": 1.8129, "step": 132 }, { "epoch": 0.1730644111906311, "grad_norm": 1.3907544612884521, "learning_rate": 9.629750982961993e-05, "loss": 1.6596, "step": 133 }, { "epoch": 0.17436564736499674, "grad_norm": 0.7645061016082764, "learning_rate": 9.626474442988205e-05, "loss": 1.9515, "step": 134 }, { "epoch": 0.17566688353936238, "grad_norm": 0.7364439368247986, "learning_rate": 9.623197903014417e-05, "loss": 1.6136, "step": 135 }, { "epoch": 0.17696811971372803, "grad_norm": 1.7316800355911255, "learning_rate": 9.61992136304063e-05, "loss": 2.233, "step": 136 }, { "epoch": 0.1782693558880937, "grad_norm": 0.6953511834144592, "learning_rate": 9.616644823066842e-05, "loss": 1.9901, "step": 137 }, { "epoch": 0.17957059206245934, "grad_norm": 2.367100238800049, "learning_rate": 9.613368283093053e-05, "loss": 2.0875, "step": 138 }, { "epoch": 0.180871828236825, "grad_norm": 0.7945904731750488, "learning_rate": 9.610091743119267e-05, "loss": 2.07, "step": 139 }, { "epoch": 0.18217306441119063, "grad_norm": 1.1325865983963013, "learning_rate": 9.606815203145478e-05, "loss": 2.3136, "step": 140 }, { "epoch": 0.18347430058555628, "grad_norm": 0.7356148362159729, "learning_rate": 9.603538663171692e-05, "loss": 2.1773, "step": 141 }, { "epoch": 0.18477553675992192, "grad_norm": 0.8309590816497803, "learning_rate": 9.600262123197903e-05, "loss": 2.3874, "step": 142 }, { "epoch": 0.18607677293428757, "grad_norm": 0.8953571915626526, "learning_rate": 9.596985583224116e-05, "loss": 2.1606, "step": 143 }, { "epoch": 0.1873780091086532, "grad_norm": 0.6322633028030396, "learning_rate": 9.593709043250328e-05, "loss": 2.009, "step": 144 }, { "epoch": 0.18867924528301888, "grad_norm": 1.7357701063156128, "learning_rate": 9.590432503276541e-05, "loss": 1.9772, "step": 145 }, { "epoch": 0.18998048145738453, "grad_norm": 1.1223201751708984, "learning_rate": 9.587155963302753e-05, "loss": 2.3266, "step": 146 }, { "epoch": 0.19128171763175017, "grad_norm": 0.5505300164222717, "learning_rate": 9.583879423328965e-05, "loss": 1.7733, "step": 147 }, { "epoch": 0.19258295380611581, "grad_norm": 0.6640098094940186, "learning_rate": 9.580602883355178e-05, "loss": 2.07, "step": 148 }, { "epoch": 0.19388418998048146, "grad_norm": 0.66763836145401, "learning_rate": 9.57732634338139e-05, "loss": 2.0165, "step": 149 }, { "epoch": 0.1951854261548471, "grad_norm": 0.7837227582931519, "learning_rate": 9.574049803407602e-05, "loss": 2.211, "step": 150 }, { "epoch": 0.19648666232921275, "grad_norm": 0.598918080329895, "learning_rate": 9.570773263433815e-05, "loss": 2.1438, "step": 151 }, { "epoch": 0.1977878985035784, "grad_norm": 0.6743712425231934, "learning_rate": 9.567496723460027e-05, "loss": 2.1171, "step": 152 }, { "epoch": 0.19908913467794404, "grad_norm": 0.7008499503135681, "learning_rate": 9.564220183486238e-05, "loss": 1.7588, "step": 153 }, { "epoch": 0.2003903708523097, "grad_norm": 0.986320436000824, "learning_rate": 9.560943643512452e-05, "loss": 2.2904, "step": 154 }, { "epoch": 0.20169160702667535, "grad_norm": 0.6219415068626404, "learning_rate": 9.557667103538663e-05, "loss": 2.1517, "step": 155 }, { "epoch": 0.202992843201041, "grad_norm": 0.981801450252533, "learning_rate": 9.554390563564875e-05, "loss": 2.2713, "step": 156 }, { "epoch": 0.20429407937540664, "grad_norm": 1.1217375993728638, "learning_rate": 9.551114023591088e-05, "loss": 2.3594, "step": 157 }, { "epoch": 0.20559531554977228, "grad_norm": 2.6344940662384033, "learning_rate": 9.5478374836173e-05, "loss": 2.1689, "step": 158 }, { "epoch": 0.20689655172413793, "grad_norm": 0.8961989283561707, "learning_rate": 9.544560943643512e-05, "loss": 2.3509, "step": 159 }, { "epoch": 0.20819778789850357, "grad_norm": 0.804413914680481, "learning_rate": 9.541284403669725e-05, "loss": 1.8953, "step": 160 }, { "epoch": 0.20949902407286922, "grad_norm": 0.7609094381332397, "learning_rate": 9.538007863695938e-05, "loss": 2.2635, "step": 161 }, { "epoch": 0.21080026024723486, "grad_norm": 0.6744243502616882, "learning_rate": 9.53473132372215e-05, "loss": 2.2937, "step": 162 }, { "epoch": 0.21210149642160053, "grad_norm": 0.7265585660934448, "learning_rate": 9.531454783748363e-05, "loss": 1.9234, "step": 163 }, { "epoch": 0.21340273259596618, "grad_norm": 0.900113582611084, "learning_rate": 9.528178243774575e-05, "loss": 2.3306, "step": 164 }, { "epoch": 0.21470396877033182, "grad_norm": 0.8969595432281494, "learning_rate": 9.524901703800787e-05, "loss": 1.6144, "step": 165 }, { "epoch": 0.21600520494469747, "grad_norm": 0.7095807790756226, "learning_rate": 9.521625163827e-05, "loss": 1.5004, "step": 166 }, { "epoch": 0.2173064411190631, "grad_norm": 0.98898845911026, "learning_rate": 9.518348623853212e-05, "loss": 1.8337, "step": 167 }, { "epoch": 0.21860767729342875, "grad_norm": 0.6653530597686768, "learning_rate": 9.515072083879424e-05, "loss": 1.8877, "step": 168 }, { "epoch": 0.2199089134677944, "grad_norm": 1.0299891233444214, "learning_rate": 9.511795543905637e-05, "loss": 2.0675, "step": 169 }, { "epoch": 0.22121014964216004, "grad_norm": 0.7869843244552612, "learning_rate": 9.508519003931848e-05, "loss": 1.8768, "step": 170 }, { "epoch": 0.2225113858165257, "grad_norm": 0.7304084300994873, "learning_rate": 9.50524246395806e-05, "loss": 2.1968, "step": 171 }, { "epoch": 0.22381262199089136, "grad_norm": 1.3576040267944336, "learning_rate": 9.501965923984273e-05, "loss": 1.8993, "step": 172 }, { "epoch": 0.225113858165257, "grad_norm": 0.7704970836639404, "learning_rate": 9.498689384010485e-05, "loss": 1.8359, "step": 173 }, { "epoch": 0.22641509433962265, "grad_norm": 0.7894173860549927, "learning_rate": 9.495412844036697e-05, "loss": 2.0996, "step": 174 }, { "epoch": 0.2277163305139883, "grad_norm": 0.6739584803581238, "learning_rate": 9.49213630406291e-05, "loss": 2.0422, "step": 175 }, { "epoch": 0.22901756668835394, "grad_norm": 0.9023209810256958, "learning_rate": 9.488859764089122e-05, "loss": 2.3911, "step": 176 }, { "epoch": 0.23031880286271958, "grad_norm": 1.102971076965332, "learning_rate": 9.485583224115334e-05, "loss": 1.8928, "step": 177 }, { "epoch": 0.23162003903708522, "grad_norm": 0.8042721748352051, "learning_rate": 9.482306684141547e-05, "loss": 1.9615, "step": 178 }, { "epoch": 0.23292127521145087, "grad_norm": 1.0606317520141602, "learning_rate": 9.47903014416776e-05, "loss": 1.7585, "step": 179 }, { "epoch": 0.2342225113858165, "grad_norm": 0.7630911469459534, "learning_rate": 9.475753604193972e-05, "loss": 1.8687, "step": 180 }, { "epoch": 0.23552374756018218, "grad_norm": 0.7714146971702576, "learning_rate": 9.472477064220184e-05, "loss": 1.8172, "step": 181 }, { "epoch": 0.23682498373454783, "grad_norm": 0.7254714369773865, "learning_rate": 9.469200524246397e-05, "loss": 2.3588, "step": 182 }, { "epoch": 0.23812621990891347, "grad_norm": 0.6917139887809753, "learning_rate": 9.465923984272609e-05, "loss": 2.1141, "step": 183 }, { "epoch": 0.23942745608327912, "grad_norm": 1.2046674489974976, "learning_rate": 9.462647444298822e-05, "loss": 2.0556, "step": 184 }, { "epoch": 0.24072869225764476, "grad_norm": 1.0688233375549316, "learning_rate": 9.459370904325034e-05, "loss": 1.5707, "step": 185 }, { "epoch": 0.2420299284320104, "grad_norm": 0.6971908807754517, "learning_rate": 9.456094364351245e-05, "loss": 2.0598, "step": 186 }, { "epoch": 0.24333116460637605, "grad_norm": 1.3693625926971436, "learning_rate": 9.452817824377459e-05, "loss": 1.989, "step": 187 }, { "epoch": 0.2446324007807417, "grad_norm": 1.053463339805603, "learning_rate": 9.44954128440367e-05, "loss": 2.0792, "step": 188 }, { "epoch": 0.24593363695510737, "grad_norm": 0.6398875117301941, "learning_rate": 9.446264744429882e-05, "loss": 1.9216, "step": 189 }, { "epoch": 0.247234873129473, "grad_norm": 0.7208705544471741, "learning_rate": 9.442988204456095e-05, "loss": 2.3596, "step": 190 }, { "epoch": 0.24853610930383865, "grad_norm": 1.0241165161132812, "learning_rate": 9.439711664482307e-05, "loss": 2.2075, "step": 191 }, { "epoch": 0.2498373454782043, "grad_norm": 0.7135631442070007, "learning_rate": 9.436435124508519e-05, "loss": 2.2433, "step": 192 }, { "epoch": 0.2511385816525699, "grad_norm": 1.113846778869629, "learning_rate": 9.433158584534732e-05, "loss": 1.792, "step": 193 }, { "epoch": 0.2524398178269356, "grad_norm": 0.764241099357605, "learning_rate": 9.429882044560944e-05, "loss": 1.7384, "step": 194 }, { "epoch": 0.25374105400130126, "grad_norm": 0.7354151606559753, "learning_rate": 9.426605504587156e-05, "loss": 1.9651, "step": 195 }, { "epoch": 0.2550422901756669, "grad_norm": 0.5430366396903992, "learning_rate": 9.423328964613369e-05, "loss": 2.0621, "step": 196 }, { "epoch": 0.25634352635003255, "grad_norm": 0.8997671008110046, "learning_rate": 9.420052424639582e-05, "loss": 1.761, "step": 197 }, { "epoch": 0.25764476252439816, "grad_norm": 0.9637977480888367, "learning_rate": 9.416775884665794e-05, "loss": 1.9355, "step": 198 }, { "epoch": 0.25894599869876384, "grad_norm": 0.8447169661521912, "learning_rate": 9.413499344692005e-05, "loss": 2.1779, "step": 199 }, { "epoch": 0.26024723487312945, "grad_norm": 0.7721760272979736, "learning_rate": 9.410222804718219e-05, "loss": 1.6571, "step": 200 }, { "epoch": 0.2615484710474951, "grad_norm": 0.7018890976905823, "learning_rate": 9.40694626474443e-05, "loss": 2.0269, "step": 201 }, { "epoch": 0.26284970722186074, "grad_norm": 0.779867947101593, "learning_rate": 9.403669724770642e-05, "loss": 2.3586, "step": 202 }, { "epoch": 0.2641509433962264, "grad_norm": 0.6686696410179138, "learning_rate": 9.400393184796855e-05, "loss": 2.1471, "step": 203 }, { "epoch": 0.2654521795705921, "grad_norm": 1.5210613012313843, "learning_rate": 9.397116644823067e-05, "loss": 1.5798, "step": 204 }, { "epoch": 0.2667534157449577, "grad_norm": 0.8553242683410645, "learning_rate": 9.39384010484928e-05, "loss": 2.2458, "step": 205 }, { "epoch": 0.2680546519193234, "grad_norm": 0.6799885034561157, "learning_rate": 9.390563564875492e-05, "loss": 2.0975, "step": 206 }, { "epoch": 0.269355888093689, "grad_norm": 1.182754397392273, "learning_rate": 9.387287024901704e-05, "loss": 1.7813, "step": 207 }, { "epoch": 0.27065712426805466, "grad_norm": 0.7255248427391052, "learning_rate": 9.384010484927917e-05, "loss": 2.0315, "step": 208 }, { "epoch": 0.2719583604424203, "grad_norm": 0.8037382960319519, "learning_rate": 9.380733944954129e-05, "loss": 2.0904, "step": 209 }, { "epoch": 0.27325959661678595, "grad_norm": 0.7194886803627014, "learning_rate": 9.37745740498034e-05, "loss": 2.1891, "step": 210 }, { "epoch": 0.2745608327911516, "grad_norm": 3.858039617538452, "learning_rate": 9.374180865006554e-05, "loss": 1.8871, "step": 211 }, { "epoch": 0.27586206896551724, "grad_norm": 0.7340083122253418, "learning_rate": 9.370904325032766e-05, "loss": 1.875, "step": 212 }, { "epoch": 0.2771633051398829, "grad_norm": 0.8930732011795044, "learning_rate": 9.367627785058977e-05, "loss": 1.9219, "step": 213 }, { "epoch": 0.2784645413142485, "grad_norm": 0.7505930066108704, "learning_rate": 9.36435124508519e-05, "loss": 1.6215, "step": 214 }, { "epoch": 0.2797657774886142, "grad_norm": 1.3782440423965454, "learning_rate": 9.361074705111402e-05, "loss": 1.9636, "step": 215 }, { "epoch": 0.2810670136629798, "grad_norm": 1.2225100994110107, "learning_rate": 9.357798165137616e-05, "loss": 2.1157, "step": 216 }, { "epoch": 0.2823682498373455, "grad_norm": 0.8257637619972229, "learning_rate": 9.354521625163827e-05, "loss": 2.2174, "step": 217 }, { "epoch": 0.2836694860117111, "grad_norm": 0.6706610321998596, "learning_rate": 9.35124508519004e-05, "loss": 2.1739, "step": 218 }, { "epoch": 0.2849707221860768, "grad_norm": 0.5655174851417542, "learning_rate": 9.347968545216252e-05, "loss": 1.9041, "step": 219 }, { "epoch": 0.28627195836044245, "grad_norm": 0.6872392296791077, "learning_rate": 9.344692005242464e-05, "loss": 2.1031, "step": 220 }, { "epoch": 0.28757319453480806, "grad_norm": 0.4288671910762787, "learning_rate": 9.341415465268677e-05, "loss": 1.4674, "step": 221 }, { "epoch": 0.28887443070917374, "grad_norm": 0.7409505844116211, "learning_rate": 9.338138925294889e-05, "loss": 2.1262, "step": 222 }, { "epoch": 0.29017566688353935, "grad_norm": 0.814349353313446, "learning_rate": 9.334862385321101e-05, "loss": 2.2173, "step": 223 }, { "epoch": 0.291476903057905, "grad_norm": 0.6999825835227966, "learning_rate": 9.331585845347314e-05, "loss": 2.2245, "step": 224 }, { "epoch": 0.29277813923227064, "grad_norm": 1.4298471212387085, "learning_rate": 9.328309305373526e-05, "loss": 1.5078, "step": 225 }, { "epoch": 0.2940793754066363, "grad_norm": 0.8152809739112854, "learning_rate": 9.325032765399739e-05, "loss": 2.1479, "step": 226 }, { "epoch": 0.29538061158100193, "grad_norm": 0.634086012840271, "learning_rate": 9.321756225425951e-05, "loss": 1.9709, "step": 227 }, { "epoch": 0.2966818477553676, "grad_norm": 0.5806602835655212, "learning_rate": 9.318479685452162e-05, "loss": 1.8589, "step": 228 }, { "epoch": 0.2979830839297333, "grad_norm": 0.6519486308097839, "learning_rate": 9.315203145478376e-05, "loss": 1.213, "step": 229 }, { "epoch": 0.2992843201040989, "grad_norm": 1.0760489702224731, "learning_rate": 9.311926605504587e-05, "loss": 2.0669, "step": 230 }, { "epoch": 0.30058555627846456, "grad_norm": 0.7079692482948303, "learning_rate": 9.308650065530799e-05, "loss": 1.9443, "step": 231 }, { "epoch": 0.3018867924528302, "grad_norm": 0.7697675824165344, "learning_rate": 9.305373525557012e-05, "loss": 2.1778, "step": 232 }, { "epoch": 0.30318802862719585, "grad_norm": 0.7142135500907898, "learning_rate": 9.302096985583224e-05, "loss": 2.0919, "step": 233 }, { "epoch": 0.30448926480156147, "grad_norm": 0.7445077896118164, "learning_rate": 9.298820445609436e-05, "loss": 2.0158, "step": 234 }, { "epoch": 0.30579050097592714, "grad_norm": 0.6997343301773071, "learning_rate": 9.295543905635649e-05, "loss": 2.1736, "step": 235 }, { "epoch": 0.30709173715029275, "grad_norm": 0.6820374727249146, "learning_rate": 9.292267365661862e-05, "loss": 2.0966, "step": 236 }, { "epoch": 0.3083929733246584, "grad_norm": 0.9942825436592102, "learning_rate": 9.288990825688074e-05, "loss": 1.6209, "step": 237 }, { "epoch": 0.3096942094990241, "grad_norm": 1.7016510963439941, "learning_rate": 9.285714285714286e-05, "loss": 2.2946, "step": 238 }, { "epoch": 0.3109954456733897, "grad_norm": 0.7558296322822571, "learning_rate": 9.282437745740499e-05, "loss": 1.8074, "step": 239 }, { "epoch": 0.3122966818477554, "grad_norm": 0.707486093044281, "learning_rate": 9.279161205766711e-05, "loss": 1.8608, "step": 240 }, { "epoch": 0.313597918022121, "grad_norm": 1.2906628847122192, "learning_rate": 9.275884665792923e-05, "loss": 2.1403, "step": 241 }, { "epoch": 0.3148991541964867, "grad_norm": 0.6364115476608276, "learning_rate": 9.272608125819136e-05, "loss": 2.295, "step": 242 }, { "epoch": 0.3162003903708523, "grad_norm": 0.7926555275917053, "learning_rate": 9.269331585845348e-05, "loss": 2.2656, "step": 243 }, { "epoch": 0.31750162654521796, "grad_norm": 0.7089528441429138, "learning_rate": 9.266055045871561e-05, "loss": 2.2851, "step": 244 }, { "epoch": 0.3188028627195836, "grad_norm": 41.0919075012207, "learning_rate": 9.262778505897773e-05, "loss": 2.1502, "step": 245 }, { "epoch": 0.32010409889394925, "grad_norm": 15.70914077758789, "learning_rate": 9.259501965923984e-05, "loss": 2.2084, "step": 246 }, { "epoch": 0.3214053350683149, "grad_norm": 0.7123581767082214, "learning_rate": 9.256225425950197e-05, "loss": 2.2712, "step": 247 }, { "epoch": 0.32270657124268054, "grad_norm": 3.663393020629883, "learning_rate": 9.252948885976409e-05, "loss": 2.2096, "step": 248 }, { "epoch": 0.3240078074170462, "grad_norm": 0.7280385494232178, "learning_rate": 9.249672346002621e-05, "loss": 2.1999, "step": 249 }, { "epoch": 0.32530904359141183, "grad_norm": 0.6268571615219116, "learning_rate": 9.246395806028834e-05, "loss": 2.0408, "step": 250 }, { "epoch": 0.3266102797657775, "grad_norm": 0.7782486081123352, "learning_rate": 9.243119266055046e-05, "loss": 2.182, "step": 251 }, { "epoch": 0.3279115159401431, "grad_norm": 4.297134876251221, "learning_rate": 9.239842726081258e-05, "loss": 2.3012, "step": 252 }, { "epoch": 0.3292127521145088, "grad_norm": 0.6242889165878296, "learning_rate": 9.236566186107471e-05, "loss": 2.057, "step": 253 }, { "epoch": 0.3305139882888744, "grad_norm": 0.9788745045661926, "learning_rate": 9.233289646133684e-05, "loss": 1.8535, "step": 254 }, { "epoch": 0.3318152244632401, "grad_norm": 0.6852668523788452, "learning_rate": 9.230013106159896e-05, "loss": 2.0967, "step": 255 }, { "epoch": 0.33311646063760575, "grad_norm": 0.6787961721420288, "learning_rate": 9.226736566186108e-05, "loss": 2.1541, "step": 256 }, { "epoch": 0.33441769681197137, "grad_norm": 0.72751384973526, "learning_rate": 9.223460026212321e-05, "loss": 2.3555, "step": 257 }, { "epoch": 0.33571893298633704, "grad_norm": 0.7236286997795105, "learning_rate": 9.220183486238533e-05, "loss": 1.6994, "step": 258 }, { "epoch": 0.33702016916070265, "grad_norm": 1.2532989978790283, "learning_rate": 9.216906946264744e-05, "loss": 2.2207, "step": 259 }, { "epoch": 0.3383214053350683, "grad_norm": 0.7751891016960144, "learning_rate": 9.213630406290958e-05, "loss": 1.658, "step": 260 }, { "epoch": 0.33962264150943394, "grad_norm": 0.6352059841156006, "learning_rate": 9.21035386631717e-05, "loss": 2.129, "step": 261 }, { "epoch": 0.3409238776837996, "grad_norm": 0.7744314670562744, "learning_rate": 9.207077326343381e-05, "loss": 2.144, "step": 262 }, { "epoch": 0.34222511385816523, "grad_norm": 0.7391191124916077, "learning_rate": 9.203800786369594e-05, "loss": 2.1773, "step": 263 }, { "epoch": 0.3435263500325309, "grad_norm": 1.0224254131317139, "learning_rate": 9.200524246395806e-05, "loss": 2.1114, "step": 264 }, { "epoch": 0.3448275862068966, "grad_norm": 0.6617637276649475, "learning_rate": 9.197247706422019e-05, "loss": 2.0069, "step": 265 }, { "epoch": 0.3461288223812622, "grad_norm": 0.9309566020965576, "learning_rate": 9.193971166448231e-05, "loss": 0.9166, "step": 266 }, { "epoch": 0.34743005855562786, "grad_norm": 0.852107048034668, "learning_rate": 9.190694626474443e-05, "loss": 2.1029, "step": 267 }, { "epoch": 0.3487312947299935, "grad_norm": 1.0176396369934082, "learning_rate": 9.187418086500656e-05, "loss": 2.1532, "step": 268 }, { "epoch": 0.35003253090435915, "grad_norm": 0.7276338338851929, "learning_rate": 9.184141546526868e-05, "loss": 2.1806, "step": 269 }, { "epoch": 0.35133376707872477, "grad_norm": 0.9919301271438599, "learning_rate": 9.18086500655308e-05, "loss": 2.1517, "step": 270 }, { "epoch": 0.35263500325309044, "grad_norm": 0.8287739157676697, "learning_rate": 9.177588466579293e-05, "loss": 2.4058, "step": 271 }, { "epoch": 0.35393623942745606, "grad_norm": 0.8146002292633057, "learning_rate": 9.174311926605506e-05, "loss": 2.3355, "step": 272 }, { "epoch": 0.35523747560182173, "grad_norm": 2.9626729488372803, "learning_rate": 9.171035386631718e-05, "loss": 2.0144, "step": 273 }, { "epoch": 0.3565387117761874, "grad_norm": 0.8142338395118713, "learning_rate": 9.16775884665793e-05, "loss": 2.2458, "step": 274 }, { "epoch": 0.357839947950553, "grad_norm": 0.6895964741706848, "learning_rate": 9.164482306684143e-05, "loss": 1.9132, "step": 275 }, { "epoch": 0.3591411841249187, "grad_norm": 0.6626771688461304, "learning_rate": 9.161205766710354e-05, "loss": 1.6244, "step": 276 }, { "epoch": 0.3604424202992843, "grad_norm": 0.6539360284805298, "learning_rate": 9.157929226736566e-05, "loss": 2.1859, "step": 277 }, { "epoch": 0.36174365647365, "grad_norm": 1.0775957107543945, "learning_rate": 9.15465268676278e-05, "loss": 1.9108, "step": 278 }, { "epoch": 0.3630448926480156, "grad_norm": 0.8867959976196289, "learning_rate": 9.151376146788991e-05, "loss": 1.9537, "step": 279 }, { "epoch": 0.36434612882238127, "grad_norm": 1.2823213338851929, "learning_rate": 9.148099606815203e-05, "loss": 2.2128, "step": 280 }, { "epoch": 0.3656473649967469, "grad_norm": 0.7368249297142029, "learning_rate": 9.144823066841416e-05, "loss": 2.1538, "step": 281 }, { "epoch": 0.36694860117111255, "grad_norm": 1.277786374092102, "learning_rate": 9.141546526867628e-05, "loss": 2.1334, "step": 282 }, { "epoch": 0.3682498373454782, "grad_norm": 0.7459700703620911, "learning_rate": 9.13826998689384e-05, "loss": 2.1591, "step": 283 }, { "epoch": 0.36955107351984384, "grad_norm": 0.7938099503517151, "learning_rate": 9.134993446920053e-05, "loss": 2.3955, "step": 284 }, { "epoch": 0.3708523096942095, "grad_norm": 0.739590048789978, "learning_rate": 9.131716906946265e-05, "loss": 2.1415, "step": 285 }, { "epoch": 0.37215354586857513, "grad_norm": 0.6934852600097656, "learning_rate": 9.128440366972478e-05, "loss": 2.1209, "step": 286 }, { "epoch": 0.3734547820429408, "grad_norm": 0.6983306407928467, "learning_rate": 9.12516382699869e-05, "loss": 2.1712, "step": 287 }, { "epoch": 0.3747560182173064, "grad_norm": 0.6534546613693237, "learning_rate": 9.121887287024901e-05, "loss": 2.3622, "step": 288 }, { "epoch": 0.3760572543916721, "grad_norm": 0.7787095308303833, "learning_rate": 9.118610747051115e-05, "loss": 1.8637, "step": 289 }, { "epoch": 0.37735849056603776, "grad_norm": 0.7561502456665039, "learning_rate": 9.115334207077328e-05, "loss": 2.0546, "step": 290 }, { "epoch": 0.3786597267404034, "grad_norm": 0.610038161277771, "learning_rate": 9.11205766710354e-05, "loss": 1.9994, "step": 291 }, { "epoch": 0.37996096291476905, "grad_norm": 0.6360822319984436, "learning_rate": 9.108781127129751e-05, "loss": 1.8957, "step": 292 }, { "epoch": 0.38126219908913467, "grad_norm": 1.0899056196212769, "learning_rate": 9.105504587155964e-05, "loss": 1.6823, "step": 293 }, { "epoch": 0.38256343526350034, "grad_norm": 0.6136271953582764, "learning_rate": 9.102228047182176e-05, "loss": 1.981, "step": 294 }, { "epoch": 0.38386467143786596, "grad_norm": 0.7351566553115845, "learning_rate": 9.098951507208388e-05, "loss": 2.5255, "step": 295 }, { "epoch": 0.38516590761223163, "grad_norm": 0.725439727306366, "learning_rate": 9.095674967234601e-05, "loss": 1.5367, "step": 296 }, { "epoch": 0.38646714378659724, "grad_norm": 2.39194917678833, "learning_rate": 9.092398427260813e-05, "loss": 1.3339, "step": 297 }, { "epoch": 0.3877683799609629, "grad_norm": 0.7107850909233093, "learning_rate": 9.089121887287025e-05, "loss": 2.1604, "step": 298 }, { "epoch": 0.3890696161353286, "grad_norm": 0.705926775932312, "learning_rate": 9.085845347313238e-05, "loss": 1.8897, "step": 299 }, { "epoch": 0.3903708523096942, "grad_norm": 0.6793768405914307, "learning_rate": 9.08256880733945e-05, "loss": 1.5943, "step": 300 }, { "epoch": 0.3916720884840599, "grad_norm": 2.0021564960479736, "learning_rate": 9.079292267365662e-05, "loss": 2.183, "step": 301 }, { "epoch": 0.3929733246584255, "grad_norm": 0.6673298478126526, "learning_rate": 9.076015727391875e-05, "loss": 2.0515, "step": 302 }, { "epoch": 0.39427456083279117, "grad_norm": 1.2079287767410278, "learning_rate": 9.072739187418086e-05, "loss": 2.031, "step": 303 }, { "epoch": 0.3955757970071568, "grad_norm": 0.7156415581703186, "learning_rate": 9.069462647444298e-05, "loss": 2.3156, "step": 304 }, { "epoch": 0.39687703318152245, "grad_norm": 0.642803966999054, "learning_rate": 9.066186107470511e-05, "loss": 2.3982, "step": 305 }, { "epoch": 0.39817826935588807, "grad_norm": 0.7380147576332092, "learning_rate": 9.062909567496723e-05, "loss": 1.7944, "step": 306 }, { "epoch": 0.39947950553025374, "grad_norm": 0.8421279788017273, "learning_rate": 9.059633027522936e-05, "loss": 2.2881, "step": 307 }, { "epoch": 0.4007807417046194, "grad_norm": 0.7420101761817932, "learning_rate": 9.056356487549148e-05, "loss": 2.1111, "step": 308 }, { "epoch": 0.40208197787898503, "grad_norm": 0.7158164978027344, "learning_rate": 9.053079947575361e-05, "loss": 2.1432, "step": 309 }, { "epoch": 0.4033832140533507, "grad_norm": 0.574983537197113, "learning_rate": 9.049803407601573e-05, "loss": 2.4254, "step": 310 }, { "epoch": 0.4046844502277163, "grad_norm": 0.593608021736145, "learning_rate": 9.046526867627786e-05, "loss": 2.0312, "step": 311 }, { "epoch": 0.405985686402082, "grad_norm": 0.7281979322433472, "learning_rate": 9.043250327653998e-05, "loss": 2.1313, "step": 312 }, { "epoch": 0.4072869225764476, "grad_norm": 0.6187828183174133, "learning_rate": 9.03997378768021e-05, "loss": 1.7342, "step": 313 }, { "epoch": 0.4085881587508133, "grad_norm": 0.5926035046577454, "learning_rate": 9.036697247706423e-05, "loss": 2.223, "step": 314 }, { "epoch": 0.4098893949251789, "grad_norm": 0.584149956703186, "learning_rate": 9.033420707732635e-05, "loss": 2.0782, "step": 315 }, { "epoch": 0.41119063109954457, "grad_norm": 0.6237155795097351, "learning_rate": 9.030144167758847e-05, "loss": 2.3002, "step": 316 }, { "epoch": 0.41249186727391024, "grad_norm": 0.7975071668624878, "learning_rate": 9.02686762778506e-05, "loss": 2.1712, "step": 317 }, { "epoch": 0.41379310344827586, "grad_norm": 0.5999512076377869, "learning_rate": 9.023591087811272e-05, "loss": 2.1215, "step": 318 }, { "epoch": 0.41509433962264153, "grad_norm": 1.1658486127853394, "learning_rate": 9.020314547837483e-05, "loss": 1.6549, "step": 319 }, { "epoch": 0.41639557579700714, "grad_norm": 0.8601174354553223, "learning_rate": 9.017038007863697e-05, "loss": 2.1838, "step": 320 }, { "epoch": 0.4176968119713728, "grad_norm": 0.7607709169387817, "learning_rate": 9.013761467889908e-05, "loss": 2.0199, "step": 321 }, { "epoch": 0.41899804814573843, "grad_norm": 0.6651380658149719, "learning_rate": 9.01048492791612e-05, "loss": 2.0216, "step": 322 }, { "epoch": 0.4202992843201041, "grad_norm": 0.7463403344154358, "learning_rate": 9.007208387942333e-05, "loss": 2.0604, "step": 323 }, { "epoch": 0.4216005204944697, "grad_norm": 0.6778780221939087, "learning_rate": 9.003931847968545e-05, "loss": 2.232, "step": 324 }, { "epoch": 0.4229017566688354, "grad_norm": 0.6796633005142212, "learning_rate": 9.000655307994758e-05, "loss": 2.315, "step": 325 }, { "epoch": 0.42420299284320107, "grad_norm": 0.6565291881561279, "learning_rate": 8.99737876802097e-05, "loss": 2.1574, "step": 326 }, { "epoch": 0.4255042290175667, "grad_norm": 0.8654565215110779, "learning_rate": 8.994102228047182e-05, "loss": 2.2732, "step": 327 }, { "epoch": 0.42680546519193235, "grad_norm": 2.064732074737549, "learning_rate": 8.990825688073395e-05, "loss": 1.8561, "step": 328 }, { "epoch": 0.42810670136629797, "grad_norm": 0.5791717767715454, "learning_rate": 8.987549148099608e-05, "loss": 2.0795, "step": 329 }, { "epoch": 0.42940793754066364, "grad_norm": 0.5610654354095459, "learning_rate": 8.98427260812582e-05, "loss": 2.2568, "step": 330 }, { "epoch": 0.43070917371502926, "grad_norm": 0.5965012907981873, "learning_rate": 8.980996068152032e-05, "loss": 2.1776, "step": 331 }, { "epoch": 0.43201040988939493, "grad_norm": 0.6665313243865967, "learning_rate": 8.977719528178245e-05, "loss": 2.0479, "step": 332 }, { "epoch": 0.43331164606376055, "grad_norm": 0.7921652793884277, "learning_rate": 8.974442988204457e-05, "loss": 1.9188, "step": 333 }, { "epoch": 0.4346128822381262, "grad_norm": 0.8238445520401001, "learning_rate": 8.971166448230668e-05, "loss": 2.3111, "step": 334 }, { "epoch": 0.4359141184124919, "grad_norm": 1.8357375860214233, "learning_rate": 8.967889908256882e-05, "loss": 1.9003, "step": 335 }, { "epoch": 0.4372153545868575, "grad_norm": 0.9394525289535522, "learning_rate": 8.964613368283093e-05, "loss": 1.4174, "step": 336 }, { "epoch": 0.4385165907612232, "grad_norm": 0.7256464958190918, "learning_rate": 8.961336828309305e-05, "loss": 2.092, "step": 337 }, { "epoch": 0.4398178269355888, "grad_norm": 0.48097097873687744, "learning_rate": 8.958060288335518e-05, "loss": 1.6353, "step": 338 }, { "epoch": 0.44111906310995447, "grad_norm": 0.754593551158905, "learning_rate": 8.95478374836173e-05, "loss": 2.1595, "step": 339 }, { "epoch": 0.4424202992843201, "grad_norm": 0.6024689674377441, "learning_rate": 8.951507208387942e-05, "loss": 1.6777, "step": 340 }, { "epoch": 0.44372153545868576, "grad_norm": 0.5494019985198975, "learning_rate": 8.948230668414155e-05, "loss": 1.9141, "step": 341 }, { "epoch": 0.4450227716330514, "grad_norm": 1.0905380249023438, "learning_rate": 8.944954128440367e-05, "loss": 1.6304, "step": 342 }, { "epoch": 0.44632400780741704, "grad_norm": 1.2289693355560303, "learning_rate": 8.941677588466579e-05, "loss": 1.9869, "step": 343 }, { "epoch": 0.4476252439817827, "grad_norm": 9.169536590576172, "learning_rate": 8.938401048492792e-05, "loss": 2.3972, "step": 344 }, { "epoch": 0.44892648015614833, "grad_norm": 0.7125851511955261, "learning_rate": 8.935124508519004e-05, "loss": 2.1931, "step": 345 }, { "epoch": 0.450227716330514, "grad_norm": 0.6280768513679504, "learning_rate": 8.931847968545217e-05, "loss": 2.3341, "step": 346 }, { "epoch": 0.4515289525048796, "grad_norm": 0.6747077107429504, "learning_rate": 8.92857142857143e-05, "loss": 2.1558, "step": 347 }, { "epoch": 0.4528301886792453, "grad_norm": 0.7397748827934265, "learning_rate": 8.925294888597642e-05, "loss": 2.2183, "step": 348 }, { "epoch": 0.4541314248536109, "grad_norm": 0.668656587600708, "learning_rate": 8.922018348623854e-05, "loss": 2.2473, "step": 349 }, { "epoch": 0.4554326610279766, "grad_norm": 0.7301071286201477, "learning_rate": 8.918741808650067e-05, "loss": 1.9006, "step": 350 }, { "epoch": 0.4567338972023422, "grad_norm": 0.6628713011741638, "learning_rate": 8.915465268676278e-05, "loss": 1.9162, "step": 351 }, { "epoch": 0.45803513337670787, "grad_norm": 0.5624001622200012, "learning_rate": 8.91218872870249e-05, "loss": 2.1728, "step": 352 }, { "epoch": 0.45933636955107354, "grad_norm": 0.5206666588783264, "learning_rate": 8.908912188728703e-05, "loss": 2.2862, "step": 353 }, { "epoch": 0.46063760572543916, "grad_norm": 0.5872126817703247, "learning_rate": 8.905635648754915e-05, "loss": 2.0451, "step": 354 }, { "epoch": 0.46193884189980483, "grad_norm": 0.758165717124939, "learning_rate": 8.902359108781127e-05, "loss": 2.0025, "step": 355 }, { "epoch": 0.46324007807417045, "grad_norm": 0.6995052099227905, "learning_rate": 8.89908256880734e-05, "loss": 1.986, "step": 356 }, { "epoch": 0.4645413142485361, "grad_norm": 0.6720854640007019, "learning_rate": 8.895806028833552e-05, "loss": 2.0589, "step": 357 }, { "epoch": 0.46584255042290174, "grad_norm": 0.9026831984519958, "learning_rate": 8.892529488859764e-05, "loss": 1.9846, "step": 358 }, { "epoch": 0.4671437865972674, "grad_norm": 0.698437511920929, "learning_rate": 8.889252948885977e-05, "loss": 2.3646, "step": 359 }, { "epoch": 0.468445022771633, "grad_norm": 1.6582105159759521, "learning_rate": 8.885976408912189e-05, "loss": 1.4566, "step": 360 }, { "epoch": 0.4697462589459987, "grad_norm": 0.5398248434066772, "learning_rate": 8.8826998689384e-05, "loss": 2.2753, "step": 361 }, { "epoch": 0.47104749512036437, "grad_norm": 0.7266470789909363, "learning_rate": 8.879423328964614e-05, "loss": 2.3123, "step": 362 }, { "epoch": 0.47234873129473, "grad_norm": 3.3679184913635254, "learning_rate": 8.876146788990825e-05, "loss": 1.4608, "step": 363 }, { "epoch": 0.47364996746909566, "grad_norm": 0.4939165711402893, "learning_rate": 8.872870249017037e-05, "loss": 1.5993, "step": 364 }, { "epoch": 0.4749512036434613, "grad_norm": 2.1580731868743896, "learning_rate": 8.869593709043252e-05, "loss": 2.4425, "step": 365 }, { "epoch": 0.47625243981782694, "grad_norm": 0.5286079049110413, "learning_rate": 8.866317169069464e-05, "loss": 2.3963, "step": 366 }, { "epoch": 0.47755367599219256, "grad_norm": 0.7923350930213928, "learning_rate": 8.863040629095675e-05, "loss": 1.7052, "step": 367 }, { "epoch": 0.47885491216655823, "grad_norm": 2.118337869644165, "learning_rate": 8.859764089121888e-05, "loss": 1.5932, "step": 368 }, { "epoch": 0.4801561483409239, "grad_norm": 0.9238174557685852, "learning_rate": 8.8564875491481e-05, "loss": 1.6651, "step": 369 }, { "epoch": 0.4814573845152895, "grad_norm": 1.0675678253173828, "learning_rate": 8.853211009174312e-05, "loss": 2.1186, "step": 370 }, { "epoch": 0.4827586206896552, "grad_norm": 0.6834080815315247, "learning_rate": 8.849934469200525e-05, "loss": 2.0991, "step": 371 }, { "epoch": 0.4840598568640208, "grad_norm": 0.7000769376754761, "learning_rate": 8.846657929226737e-05, "loss": 2.2579, "step": 372 }, { "epoch": 0.4853610930383865, "grad_norm": 0.6907331943511963, "learning_rate": 8.843381389252949e-05, "loss": 2.1976, "step": 373 }, { "epoch": 0.4866623292127521, "grad_norm": 2.6002399921417236, "learning_rate": 8.840104849279162e-05, "loss": 2.1534, "step": 374 }, { "epoch": 0.48796356538711777, "grad_norm": 0.7345831990242004, "learning_rate": 8.836828309305374e-05, "loss": 1.9498, "step": 375 }, { "epoch": 0.4892648015614834, "grad_norm": 0.9782259464263916, "learning_rate": 8.833551769331586e-05, "loss": 2.2896, "step": 376 }, { "epoch": 0.49056603773584906, "grad_norm": 0.8012757301330566, "learning_rate": 8.830275229357799e-05, "loss": 2.085, "step": 377 }, { "epoch": 0.49186727391021473, "grad_norm": 0.7021802067756653, "learning_rate": 8.82699868938401e-05, "loss": 2.0222, "step": 378 }, { "epoch": 0.49316851008458035, "grad_norm": 0.7505455613136292, "learning_rate": 8.823722149410222e-05, "loss": 2.2729, "step": 379 }, { "epoch": 0.494469746258946, "grad_norm": 0.7453184127807617, "learning_rate": 8.820445609436435e-05, "loss": 2.047, "step": 380 }, { "epoch": 0.49577098243331164, "grad_norm": 0.6470634937286377, "learning_rate": 8.817169069462647e-05, "loss": 2.2461, "step": 381 }, { "epoch": 0.4970722186076773, "grad_norm": 0.733517050743103, "learning_rate": 8.813892529488859e-05, "loss": 1.9082, "step": 382 }, { "epoch": 0.4983734547820429, "grad_norm": 1.4960238933563232, "learning_rate": 8.810615989515074e-05, "loss": 2.1898, "step": 383 }, { "epoch": 0.4996746909564086, "grad_norm": 3.3559377193450928, "learning_rate": 8.807339449541285e-05, "loss": 1.9085, "step": 384 }, { "epoch": 0.5009759271307742, "grad_norm": 0.662103533744812, "learning_rate": 8.804062909567497e-05, "loss": 1.961, "step": 385 }, { "epoch": 0.5022771633051398, "grad_norm": 0.708267331123352, "learning_rate": 8.80078636959371e-05, "loss": 2.0031, "step": 386 }, { "epoch": 0.5035783994795056, "grad_norm": 1.0726678371429443, "learning_rate": 8.797509829619922e-05, "loss": 2.5278, "step": 387 }, { "epoch": 0.5048796356538712, "grad_norm": 0.9425134658813477, "learning_rate": 8.794233289646134e-05, "loss": 2.2007, "step": 388 }, { "epoch": 0.5061808718282368, "grad_norm": 0.6030499935150146, "learning_rate": 8.790956749672347e-05, "loss": 1.5678, "step": 389 }, { "epoch": 0.5074821080026025, "grad_norm": 0.6018863916397095, "learning_rate": 8.787680209698559e-05, "loss": 1.8674, "step": 390 }, { "epoch": 0.5087833441769681, "grad_norm": 0.6354195475578308, "learning_rate": 8.78440366972477e-05, "loss": 2.0746, "step": 391 }, { "epoch": 0.5100845803513337, "grad_norm": 0.6799745559692383, "learning_rate": 8.781127129750984e-05, "loss": 2.1448, "step": 392 }, { "epoch": 0.5113858165256994, "grad_norm": 0.6732354760169983, "learning_rate": 8.777850589777196e-05, "loss": 2.0416, "step": 393 }, { "epoch": 0.5126870527000651, "grad_norm": 0.6804423928260803, "learning_rate": 8.774574049803407e-05, "loss": 2.1583, "step": 394 }, { "epoch": 0.5139882888744307, "grad_norm": 0.7107060551643372, "learning_rate": 8.77129750982962e-05, "loss": 2.3131, "step": 395 }, { "epoch": 0.5152895250487963, "grad_norm": 0.5837469696998596, "learning_rate": 8.768020969855832e-05, "loss": 2.274, "step": 396 }, { "epoch": 0.516590761223162, "grad_norm": 0.8852350115776062, "learning_rate": 8.764744429882044e-05, "loss": 1.7116, "step": 397 }, { "epoch": 0.5178919973975277, "grad_norm": 1.128977656364441, "learning_rate": 8.761467889908257e-05, "loss": 1.9106, "step": 398 }, { "epoch": 0.5191932335718933, "grad_norm": 0.6670336127281189, "learning_rate": 8.758191349934469e-05, "loss": 2.3, "step": 399 }, { "epoch": 0.5204944697462589, "grad_norm": 1.0632071495056152, "learning_rate": 8.754914809960681e-05, "loss": 1.7845, "step": 400 }, { "epoch": 0.5217957059206246, "grad_norm": 0.5537011623382568, "learning_rate": 8.751638269986894e-05, "loss": 1.9364, "step": 401 }, { "epoch": 0.5230969420949902, "grad_norm": 0.6794194579124451, "learning_rate": 8.748361730013107e-05, "loss": 2.0038, "step": 402 }, { "epoch": 0.5243981782693559, "grad_norm": 4.056316375732422, "learning_rate": 8.745085190039319e-05, "loss": 2.0182, "step": 403 }, { "epoch": 0.5256994144437215, "grad_norm": 0.6671653389930725, "learning_rate": 8.741808650065532e-05, "loss": 2.2098, "step": 404 }, { "epoch": 0.5270006506180872, "grad_norm": 0.794202983379364, "learning_rate": 8.738532110091744e-05, "loss": 2.0673, "step": 405 }, { "epoch": 0.5283018867924528, "grad_norm": 0.6568089127540588, "learning_rate": 8.735255570117956e-05, "loss": 2.1294, "step": 406 }, { "epoch": 0.5296031229668184, "grad_norm": 0.6698296666145325, "learning_rate": 8.731979030144169e-05, "loss": 1.8313, "step": 407 }, { "epoch": 0.5309043591411842, "grad_norm": 0.715895414352417, "learning_rate": 8.72870249017038e-05, "loss": 1.9679, "step": 408 }, { "epoch": 0.5322055953155498, "grad_norm": 0.6321094632148743, "learning_rate": 8.725425950196592e-05, "loss": 1.7009, "step": 409 }, { "epoch": 0.5335068314899154, "grad_norm": 0.8281283378601074, "learning_rate": 8.722149410222806e-05, "loss": 2.1704, "step": 410 }, { "epoch": 0.534808067664281, "grad_norm": 0.8965051770210266, "learning_rate": 8.718872870249017e-05, "loss": 1.9685, "step": 411 }, { "epoch": 0.5361093038386467, "grad_norm": 0.7773590087890625, "learning_rate": 8.715596330275229e-05, "loss": 2.3617, "step": 412 }, { "epoch": 0.5374105400130124, "grad_norm": 0.8172918558120728, "learning_rate": 8.712319790301442e-05, "loss": 2.0915, "step": 413 }, { "epoch": 0.538711776187378, "grad_norm": 0.6609477996826172, "learning_rate": 8.709043250327654e-05, "loss": 1.8857, "step": 414 }, { "epoch": 0.5400130123617437, "grad_norm": 0.6234973669052124, "learning_rate": 8.705766710353866e-05, "loss": 1.9387, "step": 415 }, { "epoch": 0.5413142485361093, "grad_norm": 1.364524245262146, "learning_rate": 8.702490170380079e-05, "loss": 1.6324, "step": 416 }, { "epoch": 0.5426154847104749, "grad_norm": 0.525234043598175, "learning_rate": 8.699213630406291e-05, "loss": 1.15, "step": 417 }, { "epoch": 0.5439167208848406, "grad_norm": 0.7136496305465698, "learning_rate": 8.695937090432503e-05, "loss": 1.7993, "step": 418 }, { "epoch": 0.5452179570592063, "grad_norm": 0.5480121374130249, "learning_rate": 8.692660550458716e-05, "loss": 2.0671, "step": 419 }, { "epoch": 0.5465191932335719, "grad_norm": 1.6092497110366821, "learning_rate": 8.689384010484928e-05, "loss": 2.3048, "step": 420 }, { "epoch": 0.5478204294079375, "grad_norm": 1.1331690549850464, "learning_rate": 8.686107470511141e-05, "loss": 2.051, "step": 421 }, { "epoch": 0.5491216655823032, "grad_norm": 0.581285834312439, "learning_rate": 8.682830930537354e-05, "loss": 1.8995, "step": 422 }, { "epoch": 0.5504229017566689, "grad_norm": 0.6273661851882935, "learning_rate": 8.679554390563566e-05, "loss": 2.0016, "step": 423 }, { "epoch": 0.5517241379310345, "grad_norm": 0.6980569958686829, "learning_rate": 8.676277850589778e-05, "loss": 2.0262, "step": 424 }, { "epoch": 0.5530253741054001, "grad_norm": 0.5972334742546082, "learning_rate": 8.67300131061599e-05, "loss": 2.012, "step": 425 }, { "epoch": 0.5543266102797658, "grad_norm": 0.6985610127449036, "learning_rate": 8.669724770642202e-05, "loss": 1.8135, "step": 426 }, { "epoch": 0.5556278464541314, "grad_norm": 1.2523705959320068, "learning_rate": 8.666448230668414e-05, "loss": 2.0182, "step": 427 }, { "epoch": 0.556929082628497, "grad_norm": 0.6394460797309875, "learning_rate": 8.663171690694627e-05, "loss": 1.9476, "step": 428 }, { "epoch": 0.5582303188028627, "grad_norm": 0.9993970394134521, "learning_rate": 8.659895150720839e-05, "loss": 1.7855, "step": 429 }, { "epoch": 0.5595315549772284, "grad_norm": 0.5940405130386353, "learning_rate": 8.656618610747051e-05, "loss": 2.026, "step": 430 }, { "epoch": 0.560832791151594, "grad_norm": 1.267106056213379, "learning_rate": 8.653342070773264e-05, "loss": 2.0443, "step": 431 }, { "epoch": 0.5621340273259596, "grad_norm": 1.6719133853912354, "learning_rate": 8.650065530799476e-05, "loss": 1.8779, "step": 432 }, { "epoch": 0.5634352635003254, "grad_norm": 0.8533623218536377, "learning_rate": 8.646788990825688e-05, "loss": 1.9577, "step": 433 }, { "epoch": 0.564736499674691, "grad_norm": 1.3418980836868286, "learning_rate": 8.643512450851901e-05, "loss": 1.8324, "step": 434 }, { "epoch": 0.5660377358490566, "grad_norm": 0.7948082089424133, "learning_rate": 8.640235910878113e-05, "loss": 1.8022, "step": 435 }, { "epoch": 0.5673389720234222, "grad_norm": 0.7929134368896484, "learning_rate": 8.636959370904324e-05, "loss": 1.6469, "step": 436 }, { "epoch": 0.5686402081977879, "grad_norm": 0.6629887819290161, "learning_rate": 8.633682830930538e-05, "loss": 1.7092, "step": 437 }, { "epoch": 0.5699414443721535, "grad_norm": 0.732011079788208, "learning_rate": 8.63040629095675e-05, "loss": 1.5328, "step": 438 }, { "epoch": 0.5712426805465192, "grad_norm": 1.5685492753982544, "learning_rate": 8.627129750982963e-05, "loss": 2.2893, "step": 439 }, { "epoch": 0.5725439167208849, "grad_norm": 0.6864717602729797, "learning_rate": 8.623853211009176e-05, "loss": 2.1985, "step": 440 }, { "epoch": 0.5738451528952505, "grad_norm": 1.009713053703308, "learning_rate": 8.620576671035388e-05, "loss": 1.8744, "step": 441 }, { "epoch": 0.5751463890696161, "grad_norm": 0.6402328014373779, "learning_rate": 8.6173001310616e-05, "loss": 2.1661, "step": 442 }, { "epoch": 0.5764476252439817, "grad_norm": 0.6982507109642029, "learning_rate": 8.614023591087812e-05, "loss": 2.3367, "step": 443 }, { "epoch": 0.5777488614183475, "grad_norm": 0.7935813665390015, "learning_rate": 8.610747051114024e-05, "loss": 2.4353, "step": 444 }, { "epoch": 0.5790500975927131, "grad_norm": 0.6260576248168945, "learning_rate": 8.607470511140236e-05, "loss": 1.7497, "step": 445 }, { "epoch": 0.5803513337670787, "grad_norm": 0.6957904100418091, "learning_rate": 8.604193971166449e-05, "loss": 2.2114, "step": 446 }, { "epoch": 0.5816525699414443, "grad_norm": 0.6901311874389648, "learning_rate": 8.600917431192661e-05, "loss": 2.1889, "step": 447 }, { "epoch": 0.58295380611581, "grad_norm": 0.9591718316078186, "learning_rate": 8.597640891218873e-05, "loss": 1.8797, "step": 448 }, { "epoch": 0.5842550422901757, "grad_norm": 0.9367043375968933, "learning_rate": 8.594364351245086e-05, "loss": 2.1629, "step": 449 }, { "epoch": 0.5855562784645413, "grad_norm": 0.9033549427986145, "learning_rate": 8.591087811271298e-05, "loss": 1.2868, "step": 450 }, { "epoch": 0.586857514638907, "grad_norm": 1.251720666885376, "learning_rate": 8.58781127129751e-05, "loss": 2.2314, "step": 451 }, { "epoch": 0.5881587508132726, "grad_norm": 0.6182752847671509, "learning_rate": 8.584534731323723e-05, "loss": 1.9024, "step": 452 }, { "epoch": 0.5894599869876382, "grad_norm": 2.1509897708892822, "learning_rate": 8.581258191349935e-05, "loss": 2.3362, "step": 453 }, { "epoch": 0.5907612231620039, "grad_norm": 0.7154418230056763, "learning_rate": 8.577981651376146e-05, "loss": 1.924, "step": 454 }, { "epoch": 0.5920624593363696, "grad_norm": 0.766367495059967, "learning_rate": 8.57470511140236e-05, "loss": 2.3431, "step": 455 }, { "epoch": 0.5933636955107352, "grad_norm": 0.6153212189674377, "learning_rate": 8.571428571428571e-05, "loss": 2.0476, "step": 456 }, { "epoch": 0.5946649316851008, "grad_norm": 0.9786802530288696, "learning_rate": 8.568152031454783e-05, "loss": 1.6108, "step": 457 }, { "epoch": 0.5959661678594665, "grad_norm": 0.7657652497291565, "learning_rate": 8.564875491480998e-05, "loss": 1.5773, "step": 458 }, { "epoch": 0.5972674040338322, "grad_norm": 0.656250536441803, "learning_rate": 8.56159895150721e-05, "loss": 2.3265, "step": 459 }, { "epoch": 0.5985686402081978, "grad_norm": 0.7800455689430237, "learning_rate": 8.558322411533421e-05, "loss": 1.9811, "step": 460 }, { "epoch": 0.5998698763825634, "grad_norm": 1.2002615928649902, "learning_rate": 8.555045871559634e-05, "loss": 1.2594, "step": 461 }, { "epoch": 0.6011711125569291, "grad_norm": 0.9347881078720093, "learning_rate": 8.551769331585846e-05, "loss": 1.8725, "step": 462 }, { "epoch": 0.6024723487312947, "grad_norm": 1.2587891817092896, "learning_rate": 8.548492791612058e-05, "loss": 2.1705, "step": 463 }, { "epoch": 0.6037735849056604, "grad_norm": 0.681530237197876, "learning_rate": 8.545216251638271e-05, "loss": 1.5772, "step": 464 }, { "epoch": 0.605074821080026, "grad_norm": 0.6633172035217285, "learning_rate": 8.541939711664483e-05, "loss": 2.3003, "step": 465 }, { "epoch": 0.6063760572543917, "grad_norm": 0.528300940990448, "learning_rate": 8.538663171690695e-05, "loss": 1.8082, "step": 466 }, { "epoch": 0.6076772934287573, "grad_norm": 0.5672366619110107, "learning_rate": 8.535386631716908e-05, "loss": 2.0697, "step": 467 }, { "epoch": 0.6089785296031229, "grad_norm": 0.6567136645317078, "learning_rate": 8.53211009174312e-05, "loss": 2.2792, "step": 468 }, { "epoch": 0.6102797657774887, "grad_norm": 0.6413310170173645, "learning_rate": 8.528833551769331e-05, "loss": 1.8863, "step": 469 }, { "epoch": 0.6115810019518543, "grad_norm": 0.6612983345985413, "learning_rate": 8.525557011795545e-05, "loss": 1.402, "step": 470 }, { "epoch": 0.6128822381262199, "grad_norm": 21.03385353088379, "learning_rate": 8.522280471821756e-05, "loss": 2.4167, "step": 471 }, { "epoch": 0.6141834743005855, "grad_norm": 3.7225208282470703, "learning_rate": 8.519003931847968e-05, "loss": 2.2199, "step": 472 }, { "epoch": 0.6154847104749512, "grad_norm": 0.6837932467460632, "learning_rate": 8.515727391874181e-05, "loss": 2.3552, "step": 473 }, { "epoch": 0.6167859466493169, "grad_norm": 0.6920175552368164, "learning_rate": 8.512450851900393e-05, "loss": 2.061, "step": 474 }, { "epoch": 0.6180871828236825, "grad_norm": 0.6425620317459106, "learning_rate": 8.509174311926605e-05, "loss": 2.1435, "step": 475 }, { "epoch": 0.6193884189980482, "grad_norm": 1.032080054283142, "learning_rate": 8.50589777195282e-05, "loss": 2.4223, "step": 476 }, { "epoch": 0.6206896551724138, "grad_norm": 0.7809975147247314, "learning_rate": 8.502621231979031e-05, "loss": 2.0291, "step": 477 }, { "epoch": 0.6219908913467794, "grad_norm": 0.929903507232666, "learning_rate": 8.499344692005243e-05, "loss": 1.7595, "step": 478 }, { "epoch": 0.623292127521145, "grad_norm": 0.7928426265716553, "learning_rate": 8.496068152031456e-05, "loss": 2.0062, "step": 479 }, { "epoch": 0.6245933636955108, "grad_norm": 0.7365553379058838, "learning_rate": 8.492791612057668e-05, "loss": 1.5081, "step": 480 }, { "epoch": 0.6258945998698764, "grad_norm": 1.1081006526947021, "learning_rate": 8.48951507208388e-05, "loss": 2.1533, "step": 481 }, { "epoch": 0.627195836044242, "grad_norm": 0.8265200257301331, "learning_rate": 8.486238532110093e-05, "loss": 1.8251, "step": 482 }, { "epoch": 0.6284970722186076, "grad_norm": 0.5814463496208191, "learning_rate": 8.482961992136305e-05, "loss": 1.8775, "step": 483 }, { "epoch": 0.6297983083929733, "grad_norm": 0.629000723361969, "learning_rate": 8.479685452162516e-05, "loss": 2.2109, "step": 484 }, { "epoch": 0.631099544567339, "grad_norm": 1.0443669557571411, "learning_rate": 8.47640891218873e-05, "loss": 2.1246, "step": 485 }, { "epoch": 0.6324007807417046, "grad_norm": 0.7475597858428955, "learning_rate": 8.473132372214941e-05, "loss": 2.0444, "step": 486 }, { "epoch": 0.6337020169160703, "grad_norm": 0.791822612285614, "learning_rate": 8.469855832241153e-05, "loss": 2.1059, "step": 487 }, { "epoch": 0.6350032530904359, "grad_norm": 0.6184449195861816, "learning_rate": 8.466579292267366e-05, "loss": 1.6941, "step": 488 }, { "epoch": 0.6363044892648015, "grad_norm": 0.7756689190864563, "learning_rate": 8.463302752293578e-05, "loss": 2.0997, "step": 489 }, { "epoch": 0.6376057254391672, "grad_norm": 1.160391092300415, "learning_rate": 8.46002621231979e-05, "loss": 2.2463, "step": 490 }, { "epoch": 0.6389069616135329, "grad_norm": 0.6284164786338806, "learning_rate": 8.456749672346003e-05, "loss": 2.0235, "step": 491 }, { "epoch": 0.6402081977878985, "grad_norm": 0.7206906080245972, "learning_rate": 8.453473132372215e-05, "loss": 2.0168, "step": 492 }, { "epoch": 0.6415094339622641, "grad_norm": 0.5965873003005981, "learning_rate": 8.450196592398427e-05, "loss": 2.0257, "step": 493 }, { "epoch": 0.6428106701366298, "grad_norm": 0.6971442699432373, "learning_rate": 8.44692005242464e-05, "loss": 2.4791, "step": 494 }, { "epoch": 0.6441119063109955, "grad_norm": 1.0114269256591797, "learning_rate": 8.443643512450853e-05, "loss": 2.0991, "step": 495 }, { "epoch": 0.6454131424853611, "grad_norm": 0.8301197290420532, "learning_rate": 8.440366972477065e-05, "loss": 1.6273, "step": 496 }, { "epoch": 0.6467143786597267, "grad_norm": 1.3567975759506226, "learning_rate": 8.437090432503278e-05, "loss": 1.371, "step": 497 }, { "epoch": 0.6480156148340924, "grad_norm": 0.7759819030761719, "learning_rate": 8.43381389252949e-05, "loss": 1.5086, "step": 498 }, { "epoch": 0.649316851008458, "grad_norm": 0.8976244926452637, "learning_rate": 8.430537352555702e-05, "loss": 2.1531, "step": 499 }, { "epoch": 0.6506180871828237, "grad_norm": 0.7300030589103699, "learning_rate": 8.427260812581915e-05, "loss": 1.7211, "step": 500 }, { "epoch": 0.6519193233571894, "grad_norm": 0.9635962247848511, "learning_rate": 8.423984272608126e-05, "loss": 2.3635, "step": 501 }, { "epoch": 0.653220559531555, "grad_norm": 1.0316168069839478, "learning_rate": 8.420707732634338e-05, "loss": 1.9955, "step": 502 }, { "epoch": 0.6545217957059206, "grad_norm": 0.7123159766197205, "learning_rate": 8.417431192660551e-05, "loss": 2.2681, "step": 503 }, { "epoch": 0.6558230318802862, "grad_norm": 0.8390327095985413, "learning_rate": 8.414154652686763e-05, "loss": 1.9471, "step": 504 }, { "epoch": 0.657124268054652, "grad_norm": 0.6674993634223938, "learning_rate": 8.410878112712975e-05, "loss": 2.3522, "step": 505 }, { "epoch": 0.6584255042290176, "grad_norm": 1.0912988185882568, "learning_rate": 8.407601572739188e-05, "loss": 2.342, "step": 506 }, { "epoch": 0.6597267404033832, "grad_norm": 0.5475256443023682, "learning_rate": 8.4043250327654e-05, "loss": 1.9138, "step": 507 }, { "epoch": 0.6610279765777488, "grad_norm": 1.3707528114318848, "learning_rate": 8.401048492791612e-05, "loss": 2.3726, "step": 508 }, { "epoch": 0.6623292127521145, "grad_norm": 0.6029149293899536, "learning_rate": 8.397771952817825e-05, "loss": 1.8632, "step": 509 }, { "epoch": 0.6636304489264802, "grad_norm": 2.3069968223571777, "learning_rate": 8.394495412844037e-05, "loss": 1.4412, "step": 510 }, { "epoch": 0.6649316851008458, "grad_norm": 1.1287693977355957, "learning_rate": 8.391218872870249e-05, "loss": 1.9024, "step": 511 }, { "epoch": 0.6662329212752115, "grad_norm": 0.6689860820770264, "learning_rate": 8.387942332896462e-05, "loss": 1.9518, "step": 512 }, { "epoch": 0.6675341574495771, "grad_norm": 1.125270962715149, "learning_rate": 8.384665792922673e-05, "loss": 2.0458, "step": 513 }, { "epoch": 0.6688353936239427, "grad_norm": 1.0174224376678467, "learning_rate": 8.381389252948887e-05, "loss": 2.2479, "step": 514 }, { "epoch": 0.6701366297983083, "grad_norm": 0.5917952656745911, "learning_rate": 8.3781127129751e-05, "loss": 2.1057, "step": 515 }, { "epoch": 0.6714378659726741, "grad_norm": 2.4607748985290527, "learning_rate": 8.374836173001312e-05, "loss": 1.9315, "step": 516 }, { "epoch": 0.6727391021470397, "grad_norm": 0.6085014343261719, "learning_rate": 8.371559633027523e-05, "loss": 1.8469, "step": 517 }, { "epoch": 0.6740403383214053, "grad_norm": 0.5860880613327026, "learning_rate": 8.368283093053736e-05, "loss": 2.3341, "step": 518 }, { "epoch": 0.675341574495771, "grad_norm": 0.6442152857780457, "learning_rate": 8.365006553079948e-05, "loss": 1.7819, "step": 519 }, { "epoch": 0.6766428106701367, "grad_norm": 0.5665480494499207, "learning_rate": 8.36173001310616e-05, "loss": 1.6682, "step": 520 }, { "epoch": 0.6779440468445023, "grad_norm": 0.7157456278800964, "learning_rate": 8.358453473132373e-05, "loss": 2.055, "step": 521 }, { "epoch": 0.6792452830188679, "grad_norm": 1.091575264930725, "learning_rate": 8.355176933158585e-05, "loss": 2.0094, "step": 522 }, { "epoch": 0.6805465191932336, "grad_norm": 1.0410292148590088, "learning_rate": 8.351900393184797e-05, "loss": 2.2704, "step": 523 }, { "epoch": 0.6818477553675992, "grad_norm": 1.1532857418060303, "learning_rate": 8.34862385321101e-05, "loss": 2.0091, "step": 524 }, { "epoch": 0.6831489915419648, "grad_norm": 0.702865481376648, "learning_rate": 8.345347313237222e-05, "loss": 1.8805, "step": 525 }, { "epoch": 0.6844502277163305, "grad_norm": 0.7025652527809143, "learning_rate": 8.342070773263434e-05, "loss": 1.8689, "step": 526 }, { "epoch": 0.6857514638906962, "grad_norm": 0.6347373127937317, "learning_rate": 8.338794233289647e-05, "loss": 2.0822, "step": 527 }, { "epoch": 0.6870527000650618, "grad_norm": 0.765911340713501, "learning_rate": 8.335517693315859e-05, "loss": 2.1028, "step": 528 }, { "epoch": 0.6883539362394274, "grad_norm": 0.631354033946991, "learning_rate": 8.33224115334207e-05, "loss": 2.1544, "step": 529 }, { "epoch": 0.6896551724137931, "grad_norm": 1.1134068965911865, "learning_rate": 8.328964613368283e-05, "loss": 2.1826, "step": 530 }, { "epoch": 0.6909564085881588, "grad_norm": 0.5473922491073608, "learning_rate": 8.325688073394495e-05, "loss": 1.7078, "step": 531 }, { "epoch": 0.6922576447625244, "grad_norm": 0.6061679124832153, "learning_rate": 8.322411533420708e-05, "loss": 2.1475, "step": 532 }, { "epoch": 0.69355888093689, "grad_norm": 1.0640592575073242, "learning_rate": 8.319134993446922e-05, "loss": 2.1972, "step": 533 }, { "epoch": 0.6948601171112557, "grad_norm": 0.6916198134422302, "learning_rate": 8.315858453473133e-05, "loss": 1.8341, "step": 534 }, { "epoch": 0.6961613532856213, "grad_norm": 0.6084563136100769, "learning_rate": 8.312581913499345e-05, "loss": 1.8063, "step": 535 }, { "epoch": 0.697462589459987, "grad_norm": 1.154311180114746, "learning_rate": 8.309305373525558e-05, "loss": 2.1344, "step": 536 }, { "epoch": 0.6987638256343527, "grad_norm": 0.7835094928741455, "learning_rate": 8.30602883355177e-05, "loss": 2.0937, "step": 537 }, { "epoch": 0.7000650618087183, "grad_norm": 2.2955360412597656, "learning_rate": 8.302752293577982e-05, "loss": 1.9407, "step": 538 }, { "epoch": 0.7013662979830839, "grad_norm": 1.0409023761749268, "learning_rate": 8.299475753604195e-05, "loss": 2.2683, "step": 539 }, { "epoch": 0.7026675341574495, "grad_norm": 1.4477421045303345, "learning_rate": 8.296199213630407e-05, "loss": 2.017, "step": 540 }, { "epoch": 0.7039687703318153, "grad_norm": 1.6405701637268066, "learning_rate": 8.292922673656619e-05, "loss": 1.8175, "step": 541 }, { "epoch": 0.7052700065061809, "grad_norm": 2.3871591091156006, "learning_rate": 8.289646133682832e-05, "loss": 1.6692, "step": 542 }, { "epoch": 0.7065712426805465, "grad_norm": 2.1627774238586426, "learning_rate": 8.286369593709044e-05, "loss": 2.0959, "step": 543 }, { "epoch": 0.7078724788549121, "grad_norm": 0.7536118626594543, "learning_rate": 8.283093053735255e-05, "loss": 1.981, "step": 544 }, { "epoch": 0.7091737150292778, "grad_norm": 0.7336663603782654, "learning_rate": 8.279816513761469e-05, "loss": 2.066, "step": 545 }, { "epoch": 0.7104749512036435, "grad_norm": 1.2380722761154175, "learning_rate": 8.27653997378768e-05, "loss": 1.9795, "step": 546 }, { "epoch": 0.7117761873780091, "grad_norm": 0.614871621131897, "learning_rate": 8.273263433813892e-05, "loss": 2.0767, "step": 547 }, { "epoch": 0.7130774235523748, "grad_norm": 0.7782257795333862, "learning_rate": 8.269986893840105e-05, "loss": 2.0681, "step": 548 }, { "epoch": 0.7143786597267404, "grad_norm": 0.78246009349823, "learning_rate": 8.266710353866317e-05, "loss": 1.836, "step": 549 }, { "epoch": 0.715679895901106, "grad_norm": 0.6442996263504028, "learning_rate": 8.263433813892529e-05, "loss": 1.9834, "step": 550 }, { "epoch": 0.7169811320754716, "grad_norm": 1.2663520574569702, "learning_rate": 8.260157273918743e-05, "loss": 2.3182, "step": 551 }, { "epoch": 0.7182823682498374, "grad_norm": 1.915406346321106, "learning_rate": 8.256880733944955e-05, "loss": 1.9818, "step": 552 }, { "epoch": 0.719583604424203, "grad_norm": 5.776204586029053, "learning_rate": 8.253604193971167e-05, "loss": 2.2314, "step": 553 }, { "epoch": 0.7208848405985686, "grad_norm": 0.5929477214813232, "learning_rate": 8.25032765399738e-05, "loss": 1.811, "step": 554 }, { "epoch": 0.7221860767729343, "grad_norm": 0.825469970703125, "learning_rate": 8.247051114023592e-05, "loss": 1.8466, "step": 555 }, { "epoch": 0.7234873129473, "grad_norm": 0.7743509411811829, "learning_rate": 8.243774574049804e-05, "loss": 1.8327, "step": 556 }, { "epoch": 0.7247885491216656, "grad_norm": 0.7877588868141174, "learning_rate": 8.240498034076017e-05, "loss": 2.1231, "step": 557 }, { "epoch": 0.7260897852960312, "grad_norm": 0.5968635082244873, "learning_rate": 8.237221494102229e-05, "loss": 2.1857, "step": 558 }, { "epoch": 0.7273910214703969, "grad_norm": 0.7503438591957092, "learning_rate": 8.23394495412844e-05, "loss": 2.3733, "step": 559 }, { "epoch": 0.7286922576447625, "grad_norm": 0.7403172850608826, "learning_rate": 8.230668414154654e-05, "loss": 1.58, "step": 560 }, { "epoch": 0.7299934938191281, "grad_norm": 0.6493086218833923, "learning_rate": 8.227391874180865e-05, "loss": 2.1179, "step": 561 }, { "epoch": 0.7312947299934938, "grad_norm": 0.6515160799026489, "learning_rate": 8.224115334207077e-05, "loss": 1.5892, "step": 562 }, { "epoch": 0.7325959661678595, "grad_norm": 0.6652805805206299, "learning_rate": 8.22083879423329e-05, "loss": 2.2296, "step": 563 }, { "epoch": 0.7338972023422251, "grad_norm": 0.7767995595932007, "learning_rate": 8.217562254259502e-05, "loss": 2.3633, "step": 564 }, { "epoch": 0.7351984385165907, "grad_norm": 0.7031087279319763, "learning_rate": 8.214285714285714e-05, "loss": 1.4568, "step": 565 }, { "epoch": 0.7364996746909565, "grad_norm": 0.7119211554527283, "learning_rate": 8.211009174311927e-05, "loss": 1.9931, "step": 566 }, { "epoch": 0.7378009108653221, "grad_norm": 0.7706268429756165, "learning_rate": 8.207732634338139e-05, "loss": 1.9735, "step": 567 }, { "epoch": 0.7391021470396877, "grad_norm": 0.6263856291770935, "learning_rate": 8.204456094364351e-05, "loss": 2.073, "step": 568 }, { "epoch": 0.7404033832140533, "grad_norm": 0.8670558929443359, "learning_rate": 8.201179554390564e-05, "loss": 1.775, "step": 569 }, { "epoch": 0.741704619388419, "grad_norm": 0.7861795425415039, "learning_rate": 8.197903014416777e-05, "loss": 1.7419, "step": 570 }, { "epoch": 0.7430058555627846, "grad_norm": 1.3358899354934692, "learning_rate": 8.194626474442989e-05, "loss": 1.9664, "step": 571 }, { "epoch": 0.7443070917371503, "grad_norm": 0.8691604733467102, "learning_rate": 8.191349934469202e-05, "loss": 2.0334, "step": 572 }, { "epoch": 0.745608327911516, "grad_norm": 1.152927279472351, "learning_rate": 8.188073394495414e-05, "loss": 1.8109, "step": 573 }, { "epoch": 0.7469095640858816, "grad_norm": 0.6424343585968018, "learning_rate": 8.184796854521626e-05, "loss": 2.0698, "step": 574 }, { "epoch": 0.7482108002602472, "grad_norm": 1.100721001625061, "learning_rate": 8.181520314547839e-05, "loss": 2.4046, "step": 575 }, { "epoch": 0.7495120364346128, "grad_norm": 0.7956762313842773, "learning_rate": 8.17824377457405e-05, "loss": 1.9949, "step": 576 }, { "epoch": 0.7508132726089786, "grad_norm": 0.7020511031150818, "learning_rate": 8.174967234600262e-05, "loss": 2.4662, "step": 577 }, { "epoch": 0.7521145087833442, "grad_norm": 0.8399629592895508, "learning_rate": 8.171690694626475e-05, "loss": 1.9542, "step": 578 }, { "epoch": 0.7534157449577098, "grad_norm": 0.5644176006317139, "learning_rate": 8.168414154652687e-05, "loss": 1.5838, "step": 579 }, { "epoch": 0.7547169811320755, "grad_norm": 0.6935431957244873, "learning_rate": 8.165137614678899e-05, "loss": 1.8846, "step": 580 }, { "epoch": 0.7560182173064411, "grad_norm": 0.7071988582611084, "learning_rate": 8.161861074705112e-05, "loss": 1.798, "step": 581 }, { "epoch": 0.7573194534808068, "grad_norm": 0.6755792498588562, "learning_rate": 8.158584534731324e-05, "loss": 2.0076, "step": 582 }, { "epoch": 0.7586206896551724, "grad_norm": 1.780258297920227, "learning_rate": 8.155307994757536e-05, "loss": 1.8675, "step": 583 }, { "epoch": 0.7599219258295381, "grad_norm": 0.8601011037826538, "learning_rate": 8.152031454783749e-05, "loss": 2.2614, "step": 584 }, { "epoch": 0.7612231620039037, "grad_norm": 0.7060094475746155, "learning_rate": 8.148754914809961e-05, "loss": 1.9638, "step": 585 }, { "epoch": 0.7625243981782693, "grad_norm": 0.6910088658332825, "learning_rate": 8.145478374836173e-05, "loss": 1.7306, "step": 586 }, { "epoch": 0.763825634352635, "grad_norm": 1.2240207195281982, "learning_rate": 8.142201834862386e-05, "loss": 2.2245, "step": 587 }, { "epoch": 0.7651268705270007, "grad_norm": 0.703241229057312, "learning_rate": 8.138925294888599e-05, "loss": 2.062, "step": 588 }, { "epoch": 0.7664281067013663, "grad_norm": 0.7265186905860901, "learning_rate": 8.13564875491481e-05, "loss": 2.0028, "step": 589 }, { "epoch": 0.7677293428757319, "grad_norm": 0.7734688520431519, "learning_rate": 8.132372214941022e-05, "loss": 2.135, "step": 590 }, { "epoch": 0.7690305790500976, "grad_norm": 4.211628437042236, "learning_rate": 8.129095674967236e-05, "loss": 2.1418, "step": 591 }, { "epoch": 0.7703318152244633, "grad_norm": 0.7884597778320312, "learning_rate": 8.125819134993447e-05, "loss": 2.356, "step": 592 }, { "epoch": 0.7716330513988289, "grad_norm": 0.6309873461723328, "learning_rate": 8.12254259501966e-05, "loss": 1.9859, "step": 593 }, { "epoch": 0.7729342875731945, "grad_norm": 0.6928600668907166, "learning_rate": 8.119266055045872e-05, "loss": 2.1101, "step": 594 }, { "epoch": 0.7742355237475602, "grad_norm": 1.0542515516281128, "learning_rate": 8.115989515072084e-05, "loss": 2.0072, "step": 595 }, { "epoch": 0.7755367599219258, "grad_norm": 0.7632449865341187, "learning_rate": 8.112712975098297e-05, "loss": 2.057, "step": 596 }, { "epoch": 0.7768379960962914, "grad_norm": 0.5281820297241211, "learning_rate": 8.109436435124509e-05, "loss": 1.7742, "step": 597 }, { "epoch": 0.7781392322706572, "grad_norm": 0.8348383903503418, "learning_rate": 8.106159895150721e-05, "loss": 1.7692, "step": 598 }, { "epoch": 0.7794404684450228, "grad_norm": 0.8070511221885681, "learning_rate": 8.102883355176934e-05, "loss": 1.3462, "step": 599 }, { "epoch": 0.7807417046193884, "grad_norm": 0.7010710835456848, "learning_rate": 8.099606815203146e-05, "loss": 2.2864, "step": 600 }, { "epoch": 0.782042940793754, "grad_norm": 0.677163302898407, "learning_rate": 8.096330275229358e-05, "loss": 1.9195, "step": 601 }, { "epoch": 0.7833441769681198, "grad_norm": 5.959586143493652, "learning_rate": 8.093053735255571e-05, "loss": 2.3337, "step": 602 }, { "epoch": 0.7846454131424854, "grad_norm": 1.8447562456130981, "learning_rate": 8.089777195281783e-05, "loss": 2.0443, "step": 603 }, { "epoch": 0.785946649316851, "grad_norm": 0.6984150409698486, "learning_rate": 8.086500655307994e-05, "loss": 2.205, "step": 604 }, { "epoch": 0.7872478854912166, "grad_norm": 1.7056511640548706, "learning_rate": 8.083224115334207e-05, "loss": 2.2735, "step": 605 }, { "epoch": 0.7885491216655823, "grad_norm": 0.7021740674972534, "learning_rate": 8.079947575360419e-05, "loss": 2.3609, "step": 606 }, { "epoch": 0.789850357839948, "grad_norm": 0.59489506483078, "learning_rate": 8.076671035386632e-05, "loss": 2.2962, "step": 607 }, { "epoch": 0.7911515940143136, "grad_norm": 0.8843306303024292, "learning_rate": 8.073394495412844e-05, "loss": 2.2983, "step": 608 }, { "epoch": 0.7924528301886793, "grad_norm": 0.5907214283943176, "learning_rate": 8.070117955439057e-05, "loss": 1.7583, "step": 609 }, { "epoch": 0.7937540663630449, "grad_norm": 0.6045522093772888, "learning_rate": 8.066841415465269e-05, "loss": 2.2757, "step": 610 }, { "epoch": 0.7950553025374105, "grad_norm": 0.9077786803245544, "learning_rate": 8.063564875491482e-05, "loss": 1.6024, "step": 611 }, { "epoch": 0.7963565387117761, "grad_norm": 1.039808988571167, "learning_rate": 8.060288335517694e-05, "loss": 2.2439, "step": 612 }, { "epoch": 0.7976577748861419, "grad_norm": 1.3366588354110718, "learning_rate": 8.057011795543906e-05, "loss": 2.1202, "step": 613 }, { "epoch": 0.7989590110605075, "grad_norm": 0.672241747379303, "learning_rate": 8.053735255570119e-05, "loss": 2.0362, "step": 614 }, { "epoch": 0.8002602472348731, "grad_norm": 0.6476652026176453, "learning_rate": 8.050458715596331e-05, "loss": 2.3619, "step": 615 }, { "epoch": 0.8015614834092388, "grad_norm": 1.2766649723052979, "learning_rate": 8.047182175622543e-05, "loss": 2.1384, "step": 616 }, { "epoch": 0.8028627195836044, "grad_norm": 1.1835253238677979, "learning_rate": 8.043905635648756e-05, "loss": 1.5025, "step": 617 }, { "epoch": 0.8041639557579701, "grad_norm": 0.5084010362625122, "learning_rate": 8.040629095674968e-05, "loss": 1.7892, "step": 618 }, { "epoch": 0.8054651919323357, "grad_norm": 0.6055337190628052, "learning_rate": 8.03735255570118e-05, "loss": 1.7955, "step": 619 }, { "epoch": 0.8067664281067014, "grad_norm": 0.7025715708732605, "learning_rate": 8.034076015727393e-05, "loss": 1.1933, "step": 620 }, { "epoch": 0.808067664281067, "grad_norm": 0.8497712016105652, "learning_rate": 8.030799475753604e-05, "loss": 2.0411, "step": 621 }, { "epoch": 0.8093689004554326, "grad_norm": 0.9987003803253174, "learning_rate": 8.027522935779816e-05, "loss": 2.1109, "step": 622 }, { "epoch": 0.8106701366297983, "grad_norm": 1.4276269674301147, "learning_rate": 8.024246395806029e-05, "loss": 2.0011, "step": 623 }, { "epoch": 0.811971372804164, "grad_norm": 0.631126344203949, "learning_rate": 8.020969855832241e-05, "loss": 2.0267, "step": 624 }, { "epoch": 0.8132726089785296, "grad_norm": 1.5128307342529297, "learning_rate": 8.017693315858454e-05, "loss": 1.9171, "step": 625 }, { "epoch": 0.8145738451528952, "grad_norm": 2.171869993209839, "learning_rate": 8.014416775884666e-05, "loss": 1.9663, "step": 626 }, { "epoch": 0.8158750813272609, "grad_norm": 0.6667686700820923, "learning_rate": 8.011140235910879e-05, "loss": 2.0859, "step": 627 }, { "epoch": 0.8171763175016266, "grad_norm": 0.6462613344192505, "learning_rate": 8.007863695937091e-05, "loss": 2.3756, "step": 628 }, { "epoch": 0.8184775536759922, "grad_norm": 0.739416241645813, "learning_rate": 8.004587155963303e-05, "loss": 2.0217, "step": 629 }, { "epoch": 0.8197787898503578, "grad_norm": 1.3353711366653442, "learning_rate": 8.001310615989516e-05, "loss": 2.1265, "step": 630 }, { "epoch": 0.8210800260247235, "grad_norm": 0.7359930276870728, "learning_rate": 7.998034076015728e-05, "loss": 1.9591, "step": 631 }, { "epoch": 0.8223812621990891, "grad_norm": 4.613391876220703, "learning_rate": 7.994757536041941e-05, "loss": 2.0568, "step": 632 }, { "epoch": 0.8236824983734548, "grad_norm": 0.7600522637367249, "learning_rate": 7.991480996068153e-05, "loss": 1.6821, "step": 633 }, { "epoch": 0.8249837345478205, "grad_norm": 0.5687493085861206, "learning_rate": 7.988204456094364e-05, "loss": 2.264, "step": 634 }, { "epoch": 0.8262849707221861, "grad_norm": 1.3861745595932007, "learning_rate": 7.984927916120578e-05, "loss": 1.3424, "step": 635 }, { "epoch": 0.8275862068965517, "grad_norm": 0.8707115054130554, "learning_rate": 7.98165137614679e-05, "loss": 2.2887, "step": 636 }, { "epoch": 0.8288874430709173, "grad_norm": 0.6625469923019409, "learning_rate": 7.978374836173001e-05, "loss": 1.6428, "step": 637 }, { "epoch": 0.8301886792452831, "grad_norm": 0.6107297539710999, "learning_rate": 7.975098296199214e-05, "loss": 1.8341, "step": 638 }, { "epoch": 0.8314899154196487, "grad_norm": 2.5626301765441895, "learning_rate": 7.971821756225426e-05, "loss": 2.1483, "step": 639 }, { "epoch": 0.8327911515940143, "grad_norm": 0.676650881767273, "learning_rate": 7.968545216251638e-05, "loss": 2.0969, "step": 640 }, { "epoch": 0.8340923877683799, "grad_norm": 0.7347112894058228, "learning_rate": 7.965268676277851e-05, "loss": 2.0655, "step": 641 }, { "epoch": 0.8353936239427456, "grad_norm": 1.561261534690857, "learning_rate": 7.961992136304063e-05, "loss": 1.341, "step": 642 }, { "epoch": 0.8366948601171112, "grad_norm": 0.6191935539245605, "learning_rate": 7.958715596330275e-05, "loss": 2.3476, "step": 643 }, { "epoch": 0.8379960962914769, "grad_norm": 0.6606898903846741, "learning_rate": 7.955439056356488e-05, "loss": 2.3396, "step": 644 }, { "epoch": 0.8392973324658426, "grad_norm": 0.8116512894630432, "learning_rate": 7.952162516382701e-05, "loss": 2.0669, "step": 645 }, { "epoch": 0.8405985686402082, "grad_norm": 0.639385998249054, "learning_rate": 7.948885976408913e-05, "loss": 2.1925, "step": 646 }, { "epoch": 0.8418998048145738, "grad_norm": 1.6013036966323853, "learning_rate": 7.945609436435125e-05, "loss": 1.5236, "step": 647 }, { "epoch": 0.8432010409889394, "grad_norm": 5.539115905761719, "learning_rate": 7.942332896461338e-05, "loss": 1.8435, "step": 648 }, { "epoch": 0.8445022771633052, "grad_norm": 0.6553117036819458, "learning_rate": 7.93905635648755e-05, "loss": 2.1855, "step": 649 }, { "epoch": 0.8458035133376708, "grad_norm": 6.037136554718018, "learning_rate": 7.935779816513761e-05, "loss": 1.908, "step": 650 }, { "epoch": 0.8471047495120364, "grad_norm": 0.8238575458526611, "learning_rate": 7.932503276539974e-05, "loss": 2.0715, "step": 651 }, { "epoch": 0.8484059856864021, "grad_norm": 0.647878110408783, "learning_rate": 7.929226736566186e-05, "loss": 1.4911, "step": 652 }, { "epoch": 0.8497072218607677, "grad_norm": 1.779747724533081, "learning_rate": 7.9259501965924e-05, "loss": 2.1675, "step": 653 }, { "epoch": 0.8510084580351334, "grad_norm": 0.6237292289733887, "learning_rate": 7.922673656618611e-05, "loss": 2.0786, "step": 654 }, { "epoch": 0.852309694209499, "grad_norm": 0.8392279148101807, "learning_rate": 7.919397116644823e-05, "loss": 1.6818, "step": 655 }, { "epoch": 0.8536109303838647, "grad_norm": 0.6477450728416443, "learning_rate": 7.916120576671036e-05, "loss": 2.0183, "step": 656 }, { "epoch": 0.8549121665582303, "grad_norm": 0.6381653547286987, "learning_rate": 7.912844036697248e-05, "loss": 2.0143, "step": 657 }, { "epoch": 0.8562134027325959, "grad_norm": 3.4131829738616943, "learning_rate": 7.90956749672346e-05, "loss": 1.8548, "step": 658 }, { "epoch": 0.8575146389069617, "grad_norm": 3.4522600173950195, "learning_rate": 7.906290956749673e-05, "loss": 2.1174, "step": 659 }, { "epoch": 0.8588158750813273, "grad_norm": 0.8925557732582092, "learning_rate": 7.903014416775885e-05, "loss": 2.2429, "step": 660 }, { "epoch": 0.8601171112556929, "grad_norm": 0.9122416377067566, "learning_rate": 7.899737876802097e-05, "loss": 1.8926, "step": 661 }, { "epoch": 0.8614183474300585, "grad_norm": 0.6765635013580322, "learning_rate": 7.89646133682831e-05, "loss": 1.7244, "step": 662 }, { "epoch": 0.8627195836044242, "grad_norm": 0.6188225150108337, "learning_rate": 7.893184796854523e-05, "loss": 2.1846, "step": 663 }, { "epoch": 0.8640208197787899, "grad_norm": 0.6332929730415344, "learning_rate": 7.889908256880735e-05, "loss": 1.5422, "step": 664 }, { "epoch": 0.8653220559531555, "grad_norm": 0.6681044101715088, "learning_rate": 7.886631716906946e-05, "loss": 2.1023, "step": 665 }, { "epoch": 0.8666232921275211, "grad_norm": 1.0010368824005127, "learning_rate": 7.88335517693316e-05, "loss": 2.2435, "step": 666 }, { "epoch": 0.8679245283018868, "grad_norm": 0.7121641039848328, "learning_rate": 7.880078636959371e-05, "loss": 1.9136, "step": 667 }, { "epoch": 0.8692257644762524, "grad_norm": 0.8264907598495483, "learning_rate": 7.876802096985583e-05, "loss": 2.0102, "step": 668 }, { "epoch": 0.870527000650618, "grad_norm": 0.7948901057243347, "learning_rate": 7.873525557011796e-05, "loss": 1.9752, "step": 669 }, { "epoch": 0.8718282368249838, "grad_norm": 0.6690786480903625, "learning_rate": 7.870249017038008e-05, "loss": 1.1575, "step": 670 }, { "epoch": 0.8731294729993494, "grad_norm": 0.7070730328559875, "learning_rate": 7.86697247706422e-05, "loss": 2.2688, "step": 671 }, { "epoch": 0.874430709173715, "grad_norm": 0.53987717628479, "learning_rate": 7.863695937090433e-05, "loss": 1.6972, "step": 672 }, { "epoch": 0.8757319453480806, "grad_norm": 0.7822424173355103, "learning_rate": 7.860419397116645e-05, "loss": 2.2058, "step": 673 }, { "epoch": 0.8770331815224464, "grad_norm": 0.633253276348114, "learning_rate": 7.857142857142858e-05, "loss": 1.8491, "step": 674 }, { "epoch": 0.878334417696812, "grad_norm": 0.6231401562690735, "learning_rate": 7.85386631716907e-05, "loss": 2.2362, "step": 675 }, { "epoch": 0.8796356538711776, "grad_norm": 0.6102195382118225, "learning_rate": 7.850589777195282e-05, "loss": 1.721, "step": 676 }, { "epoch": 0.8809368900455433, "grad_norm": 0.6224596500396729, "learning_rate": 7.847313237221495e-05, "loss": 1.6651, "step": 677 }, { "epoch": 0.8822381262199089, "grad_norm": 0.706207275390625, "learning_rate": 7.844036697247707e-05, "loss": 2.099, "step": 678 }, { "epoch": 0.8835393623942746, "grad_norm": 0.6496186852455139, "learning_rate": 7.840760157273918e-05, "loss": 2.2498, "step": 679 }, { "epoch": 0.8848405985686402, "grad_norm": 0.5604274272918701, "learning_rate": 7.837483617300131e-05, "loss": 1.5772, "step": 680 }, { "epoch": 0.8861418347430059, "grad_norm": 0.642518937587738, "learning_rate": 7.834207077326345e-05, "loss": 1.656, "step": 681 }, { "epoch": 0.8874430709173715, "grad_norm": 0.8863670825958252, "learning_rate": 7.830930537352556e-05, "loss": 2.142, "step": 682 }, { "epoch": 0.8887443070917371, "grad_norm": 0.6266924142837524, "learning_rate": 7.827653997378768e-05, "loss": 2.3261, "step": 683 }, { "epoch": 0.8900455432661027, "grad_norm": 0.6674096584320068, "learning_rate": 7.824377457404981e-05, "loss": 2.1702, "step": 684 }, { "epoch": 0.8913467794404685, "grad_norm": 1.4851038455963135, "learning_rate": 7.821100917431193e-05, "loss": 1.8092, "step": 685 }, { "epoch": 0.8926480156148341, "grad_norm": 1.7813481092453003, "learning_rate": 7.817824377457405e-05, "loss": 2.0302, "step": 686 }, { "epoch": 0.8939492517891997, "grad_norm": 0.7105625867843628, "learning_rate": 7.814547837483618e-05, "loss": 2.2159, "step": 687 }, { "epoch": 0.8952504879635654, "grad_norm": 0.6556017994880676, "learning_rate": 7.81127129750983e-05, "loss": 2.1836, "step": 688 }, { "epoch": 0.896551724137931, "grad_norm": 0.6890139579772949, "learning_rate": 7.807994757536042e-05, "loss": 1.9044, "step": 689 }, { "epoch": 0.8978529603122967, "grad_norm": 2.9150776863098145, "learning_rate": 7.804718217562255e-05, "loss": 1.781, "step": 690 }, { "epoch": 0.8991541964866623, "grad_norm": 0.5710819959640503, "learning_rate": 7.801441677588467e-05, "loss": 1.5897, "step": 691 }, { "epoch": 0.900455432661028, "grad_norm": 0.6684173345565796, "learning_rate": 7.79816513761468e-05, "loss": 2.1468, "step": 692 }, { "epoch": 0.9017566688353936, "grad_norm": 0.6897217035293579, "learning_rate": 7.794888597640892e-05, "loss": 1.9928, "step": 693 }, { "epoch": 0.9030579050097592, "grad_norm": 1.4636558294296265, "learning_rate": 7.791612057667103e-05, "loss": 1.6953, "step": 694 }, { "epoch": 0.904359141184125, "grad_norm": 0.885755181312561, "learning_rate": 7.788335517693317e-05, "loss": 2.1491, "step": 695 }, { "epoch": 0.9056603773584906, "grad_norm": 0.6940903067588806, "learning_rate": 7.785058977719528e-05, "loss": 2.0154, "step": 696 }, { "epoch": 0.9069616135328562, "grad_norm": 0.6563905477523804, "learning_rate": 7.78178243774574e-05, "loss": 1.914, "step": 697 }, { "epoch": 0.9082628497072218, "grad_norm": 0.7555827498435974, "learning_rate": 7.778505897771953e-05, "loss": 1.4059, "step": 698 }, { "epoch": 0.9095640858815875, "grad_norm": 0.6056782007217407, "learning_rate": 7.775229357798165e-05, "loss": 2.2077, "step": 699 }, { "epoch": 0.9108653220559532, "grad_norm": 1.9092819690704346, "learning_rate": 7.771952817824378e-05, "loss": 1.7023, "step": 700 }, { "epoch": 0.9121665582303188, "grad_norm": 0.8118913173675537, "learning_rate": 7.76867627785059e-05, "loss": 1.9982, "step": 701 }, { "epoch": 0.9134677944046844, "grad_norm": 0.7120208740234375, "learning_rate": 7.765399737876803e-05, "loss": 2.2168, "step": 702 }, { "epoch": 0.9147690305790501, "grad_norm": 0.797903835773468, "learning_rate": 7.762123197903015e-05, "loss": 1.7003, "step": 703 }, { "epoch": 0.9160702667534157, "grad_norm": 0.6420096158981323, "learning_rate": 7.758846657929227e-05, "loss": 1.9803, "step": 704 }, { "epoch": 0.9173715029277814, "grad_norm": 1.8542115688323975, "learning_rate": 7.75557011795544e-05, "loss": 2.2392, "step": 705 }, { "epoch": 0.9186727391021471, "grad_norm": 1.0761765241622925, "learning_rate": 7.752293577981652e-05, "loss": 2.0924, "step": 706 }, { "epoch": 0.9199739752765127, "grad_norm": 0.6920451521873474, "learning_rate": 7.749017038007864e-05, "loss": 1.4496, "step": 707 }, { "epoch": 0.9212752114508783, "grad_norm": 0.7719810605049133, "learning_rate": 7.745740498034077e-05, "loss": 2.1649, "step": 708 }, { "epoch": 0.9225764476252439, "grad_norm": 0.7573016285896301, "learning_rate": 7.742463958060288e-05, "loss": 2.1162, "step": 709 }, { "epoch": 0.9238776837996097, "grad_norm": 0.6613726615905762, "learning_rate": 7.7391874180865e-05, "loss": 2.0795, "step": 710 }, { "epoch": 0.9251789199739753, "grad_norm": 0.8162800669670105, "learning_rate": 7.735910878112713e-05, "loss": 2.3946, "step": 711 }, { "epoch": 0.9264801561483409, "grad_norm": 0.8106135725975037, "learning_rate": 7.732634338138925e-05, "loss": 2.2713, "step": 712 }, { "epoch": 0.9277813923227066, "grad_norm": 0.8148362040519714, "learning_rate": 7.729357798165138e-05, "loss": 1.5298, "step": 713 }, { "epoch": 0.9290826284970722, "grad_norm": 0.9177562594413757, "learning_rate": 7.72608125819135e-05, "loss": 1.7834, "step": 714 }, { "epoch": 0.9303838646714379, "grad_norm": 1.0574177503585815, "learning_rate": 7.722804718217562e-05, "loss": 1.9931, "step": 715 }, { "epoch": 0.9316851008458035, "grad_norm": 0.8264479041099548, "learning_rate": 7.719528178243775e-05, "loss": 1.5179, "step": 716 }, { "epoch": 0.9329863370201692, "grad_norm": 0.716833770275116, "learning_rate": 7.716251638269987e-05, "loss": 2.1369, "step": 717 }, { "epoch": 0.9342875731945348, "grad_norm": 1.135319471359253, "learning_rate": 7.7129750982962e-05, "loss": 2.1578, "step": 718 }, { "epoch": 0.9355888093689004, "grad_norm": 0.7853922843933105, "learning_rate": 7.709698558322412e-05, "loss": 2.0312, "step": 719 }, { "epoch": 0.936890045543266, "grad_norm": 2.1889078617095947, "learning_rate": 7.706422018348625e-05, "loss": 2.0558, "step": 720 }, { "epoch": 0.9381912817176318, "grad_norm": 0.5951920747756958, "learning_rate": 7.703145478374837e-05, "loss": 2.1726, "step": 721 }, { "epoch": 0.9394925178919974, "grad_norm": 0.7441930174827576, "learning_rate": 7.699868938401049e-05, "loss": 1.7051, "step": 722 }, { "epoch": 0.940793754066363, "grad_norm": 2.0994136333465576, "learning_rate": 7.696592398427262e-05, "loss": 1.8963, "step": 723 }, { "epoch": 0.9420949902407287, "grad_norm": 0.7093216180801392, "learning_rate": 7.693315858453474e-05, "loss": 1.5897, "step": 724 }, { "epoch": 0.9433962264150944, "grad_norm": 0.6063390970230103, "learning_rate": 7.690039318479685e-05, "loss": 2.2064, "step": 725 }, { "epoch": 0.94469746258946, "grad_norm": 0.765845775604248, "learning_rate": 7.686762778505899e-05, "loss": 2.0964, "step": 726 }, { "epoch": 0.9459986987638256, "grad_norm": 0.7569318413734436, "learning_rate": 7.68348623853211e-05, "loss": 1.4201, "step": 727 }, { "epoch": 0.9472999349381913, "grad_norm": 0.7499330639839172, "learning_rate": 7.680209698558322e-05, "loss": 2.2776, "step": 728 }, { "epoch": 0.9486011711125569, "grad_norm": 0.6492407917976379, "learning_rate": 7.676933158584535e-05, "loss": 2.2388, "step": 729 }, { "epoch": 0.9499024072869225, "grad_norm": 0.6955572366714478, "learning_rate": 7.673656618610747e-05, "loss": 2.1825, "step": 730 }, { "epoch": 0.9512036434612883, "grad_norm": 0.6345792412757874, "learning_rate": 7.670380078636959e-05, "loss": 2.1316, "step": 731 }, { "epoch": 0.9525048796356539, "grad_norm": 1.2171833515167236, "learning_rate": 7.667103538663172e-05, "loss": 2.0828, "step": 732 }, { "epoch": 0.9538061158100195, "grad_norm": 0.770027756690979, "learning_rate": 7.663826998689384e-05, "loss": 1.8927, "step": 733 }, { "epoch": 0.9551073519843851, "grad_norm": 0.6684175729751587, "learning_rate": 7.660550458715597e-05, "loss": 2.3555, "step": 734 }, { "epoch": 0.9564085881587508, "grad_norm": 0.7743199467658997, "learning_rate": 7.657273918741809e-05, "loss": 2.4471, "step": 735 }, { "epoch": 0.9577098243331165, "grad_norm": 1.261738657951355, "learning_rate": 7.65399737876802e-05, "loss": 1.8949, "step": 736 }, { "epoch": 0.9590110605074821, "grad_norm": 0.6188062429428101, "learning_rate": 7.650720838794234e-05, "loss": 2.3102, "step": 737 }, { "epoch": 0.9603122966818478, "grad_norm": 0.7198075652122498, "learning_rate": 7.647444298820447e-05, "loss": 1.9864, "step": 738 }, { "epoch": 0.9616135328562134, "grad_norm": 2.8937673568725586, "learning_rate": 7.644167758846659e-05, "loss": 2.0847, "step": 739 }, { "epoch": 0.962914769030579, "grad_norm": 0.5778738260269165, "learning_rate": 7.64089121887287e-05, "loss": 1.9013, "step": 740 }, { "epoch": 0.9642160052049447, "grad_norm": 0.8019165396690369, "learning_rate": 7.637614678899084e-05, "loss": 2.1367, "step": 741 }, { "epoch": 0.9655172413793104, "grad_norm": 0.6918816566467285, "learning_rate": 7.634338138925295e-05, "loss": 1.7862, "step": 742 }, { "epoch": 0.966818477553676, "grad_norm": 0.6876587867736816, "learning_rate": 7.631061598951507e-05, "loss": 1.981, "step": 743 }, { "epoch": 0.9681197137280416, "grad_norm": 0.6456512212753296, "learning_rate": 7.62778505897772e-05, "loss": 2.2527, "step": 744 }, { "epoch": 0.9694209499024072, "grad_norm": 1.267526626586914, "learning_rate": 7.624508519003932e-05, "loss": 2.352, "step": 745 }, { "epoch": 0.970722186076773, "grad_norm": 0.6587380766868591, "learning_rate": 7.621231979030144e-05, "loss": 1.6199, "step": 746 }, { "epoch": 0.9720234222511386, "grad_norm": 1.4538242816925049, "learning_rate": 7.617955439056357e-05, "loss": 1.7093, "step": 747 }, { "epoch": 0.9733246584255042, "grad_norm": 0.8302292823791504, "learning_rate": 7.614678899082569e-05, "loss": 1.9973, "step": 748 }, { "epoch": 0.9746258945998699, "grad_norm": 0.7210075855255127, "learning_rate": 7.61140235910878e-05, "loss": 2.1274, "step": 749 }, { "epoch": 0.9759271307742355, "grad_norm": 0.591905415058136, "learning_rate": 7.608125819134994e-05, "loss": 2.1526, "step": 750 }, { "epoch": 0.9772283669486012, "grad_norm": 0.5860458612442017, "learning_rate": 7.604849279161206e-05, "loss": 2.1363, "step": 751 }, { "epoch": 0.9785296031229668, "grad_norm": 0.7068420648574829, "learning_rate": 7.601572739187419e-05, "loss": 2.0309, "step": 752 }, { "epoch": 0.9798308392973325, "grad_norm": 0.728827178478241, "learning_rate": 7.59829619921363e-05, "loss": 2.1021, "step": 753 }, { "epoch": 0.9811320754716981, "grad_norm": 0.7892928719520569, "learning_rate": 7.595019659239842e-05, "loss": 1.6011, "step": 754 }, { "epoch": 0.9824333116460637, "grad_norm": 2.1846065521240234, "learning_rate": 7.591743119266055e-05, "loss": 1.7399, "step": 755 }, { "epoch": 0.9837345478204295, "grad_norm": 0.7677867412567139, "learning_rate": 7.588466579292269e-05, "loss": 1.9577, "step": 756 }, { "epoch": 0.9850357839947951, "grad_norm": 0.9964635968208313, "learning_rate": 7.58519003931848e-05, "loss": 2.0548, "step": 757 }, { "epoch": 0.9863370201691607, "grad_norm": 0.8075155019760132, "learning_rate": 7.581913499344692e-05, "loss": 2.1778, "step": 758 }, { "epoch": 0.9876382563435263, "grad_norm": 0.6600486040115356, "learning_rate": 7.578636959370905e-05, "loss": 1.9842, "step": 759 }, { "epoch": 0.988939492517892, "grad_norm": 0.6700686812400818, "learning_rate": 7.575360419397117e-05, "loss": 1.8483, "step": 760 }, { "epoch": 0.9902407286922577, "grad_norm": 0.6455413699150085, "learning_rate": 7.572083879423329e-05, "loss": 1.7979, "step": 761 }, { "epoch": 0.9915419648666233, "grad_norm": 0.7454836964607239, "learning_rate": 7.568807339449542e-05, "loss": 1.8957, "step": 762 }, { "epoch": 0.9928432010409889, "grad_norm": 1.7144545316696167, "learning_rate": 7.565530799475754e-05, "loss": 2.1974, "step": 763 }, { "epoch": 0.9941444372153546, "grad_norm": 0.6758603453636169, "learning_rate": 7.562254259501966e-05, "loss": 2.3481, "step": 764 }, { "epoch": 0.9954456733897202, "grad_norm": 0.594955325126648, "learning_rate": 7.558977719528179e-05, "loss": 2.0715, "step": 765 }, { "epoch": 0.9967469095640858, "grad_norm": 0.6053801774978638, "learning_rate": 7.555701179554391e-05, "loss": 2.0215, "step": 766 }, { "epoch": 0.9980481457384516, "grad_norm": 0.9625367522239685, "learning_rate": 7.552424639580602e-05, "loss": 1.7672, "step": 767 }, { "epoch": 0.9993493819128172, "grad_norm": 0.4736585021018982, "learning_rate": 7.549148099606816e-05, "loss": 2.1165, "step": 768 }, { "epoch": 1.0006506180871828, "grad_norm": 1.012812852859497, "learning_rate": 7.545871559633027e-05, "loss": 2.1922, "step": 769 }, { "epoch": 1.0019518542615484, "grad_norm": 0.7507654428482056, "learning_rate": 7.542595019659239e-05, "loss": 2.044, "step": 770 }, { "epoch": 1.003253090435914, "grad_norm": 0.6698552370071411, "learning_rate": 7.539318479685452e-05, "loss": 1.8061, "step": 771 }, { "epoch": 1.0045543266102797, "grad_norm": 0.8947458267211914, "learning_rate": 7.536041939711664e-05, "loss": 1.1584, "step": 772 }, { "epoch": 1.0058555627846455, "grad_norm": 0.6707432866096497, "learning_rate": 7.532765399737877e-05, "loss": 1.1051, "step": 773 }, { "epoch": 1.0071567989590111, "grad_norm": 1.726501226425171, "learning_rate": 7.52948885976409e-05, "loss": 1.7042, "step": 774 }, { "epoch": 1.0084580351333767, "grad_norm": 0.8528735637664795, "learning_rate": 7.526212319790302e-05, "loss": 1.6469, "step": 775 }, { "epoch": 1.0097592713077423, "grad_norm": 0.9077311754226685, "learning_rate": 7.522935779816514e-05, "loss": 1.9047, "step": 776 }, { "epoch": 1.011060507482108, "grad_norm": 1.1657183170318604, "learning_rate": 7.519659239842727e-05, "loss": 1.3191, "step": 777 }, { "epoch": 1.0123617436564736, "grad_norm": 1.2426878213882446, "learning_rate": 7.516382699868939e-05, "loss": 1.2303, "step": 778 }, { "epoch": 1.0136629798308392, "grad_norm": 0.89070725440979, "learning_rate": 7.513106159895151e-05, "loss": 1.6901, "step": 779 }, { "epoch": 1.014964216005205, "grad_norm": 1.6244542598724365, "learning_rate": 7.509829619921364e-05, "loss": 1.474, "step": 780 }, { "epoch": 1.0162654521795706, "grad_norm": 0.5806633830070496, "learning_rate": 7.506553079947576e-05, "loss": 1.3164, "step": 781 }, { "epoch": 1.0175666883539363, "grad_norm": 0.6697611212730408, "learning_rate": 7.503276539973788e-05, "loss": 1.5607, "step": 782 }, { "epoch": 1.0188679245283019, "grad_norm": 0.6641393303871155, "learning_rate": 7.500000000000001e-05, "loss": 1.3781, "step": 783 }, { "epoch": 1.0201691607026675, "grad_norm": 1.3475302457809448, "learning_rate": 7.496723460026212e-05, "loss": 1.4885, "step": 784 }, { "epoch": 1.0214703968770331, "grad_norm": 1.0945649147033691, "learning_rate": 7.493446920052424e-05, "loss": 1.5173, "step": 785 }, { "epoch": 1.0227716330513987, "grad_norm": 1.24889075756073, "learning_rate": 7.490170380078637e-05, "loss": 1.471, "step": 786 }, { "epoch": 1.0240728692257646, "grad_norm": 1.3837988376617432, "learning_rate": 7.486893840104849e-05, "loss": 1.657, "step": 787 }, { "epoch": 1.0253741054001302, "grad_norm": 0.9529482126235962, "learning_rate": 7.483617300131061e-05, "loss": 1.8515, "step": 788 }, { "epoch": 1.0266753415744958, "grad_norm": 0.9364893436431885, "learning_rate": 7.480340760157274e-05, "loss": 1.7632, "step": 789 }, { "epoch": 1.0279765777488614, "grad_norm": 0.7346935272216797, "learning_rate": 7.477064220183486e-05, "loss": 1.8827, "step": 790 }, { "epoch": 1.029277813923227, "grad_norm": 1.262599229812622, "learning_rate": 7.473787680209698e-05, "loss": 1.8554, "step": 791 }, { "epoch": 1.0305790500975927, "grad_norm": 0.9633623957633972, "learning_rate": 7.470511140235911e-05, "loss": 1.6254, "step": 792 }, { "epoch": 1.0318802862719583, "grad_norm": 0.8005169034004211, "learning_rate": 7.467234600262124e-05, "loss": 1.5009, "step": 793 }, { "epoch": 1.033181522446324, "grad_norm": 1.0864940881729126, "learning_rate": 7.463958060288336e-05, "loss": 1.4609, "step": 794 }, { "epoch": 1.0344827586206897, "grad_norm": 1.129428505897522, "learning_rate": 7.460681520314549e-05, "loss": 1.833, "step": 795 }, { "epoch": 1.0357839947950553, "grad_norm": 0.7501147985458374, "learning_rate": 7.457404980340761e-05, "loss": 2.0397, "step": 796 }, { "epoch": 1.037085230969421, "grad_norm": 0.7439454197883606, "learning_rate": 7.454128440366973e-05, "loss": 1.3381, "step": 797 }, { "epoch": 1.0383864671437866, "grad_norm": 1.4733548164367676, "learning_rate": 7.450851900393186e-05, "loss": 1.7147, "step": 798 }, { "epoch": 1.0396877033181522, "grad_norm": 0.8671987056732178, "learning_rate": 7.447575360419398e-05, "loss": 1.3758, "step": 799 }, { "epoch": 1.0409889394925178, "grad_norm": 1.0932918787002563, "learning_rate": 7.44429882044561e-05, "loss": 1.6429, "step": 800 }, { "epoch": 1.0422901756668836, "grad_norm": 0.790759265422821, "learning_rate": 7.441022280471823e-05, "loss": 1.4454, "step": 801 }, { "epoch": 1.0435914118412493, "grad_norm": 0.861865758895874, "learning_rate": 7.437745740498034e-05, "loss": 1.5109, "step": 802 }, { "epoch": 1.0448926480156149, "grad_norm": 0.718235969543457, "learning_rate": 7.434469200524246e-05, "loss": 1.7302, "step": 803 }, { "epoch": 1.0461938841899805, "grad_norm": 0.7785394787788391, "learning_rate": 7.431192660550459e-05, "loss": 1.6024, "step": 804 }, { "epoch": 1.047495120364346, "grad_norm": 0.8478546738624573, "learning_rate": 7.427916120576671e-05, "loss": 1.4325, "step": 805 }, { "epoch": 1.0487963565387117, "grad_norm": 0.878935694694519, "learning_rate": 7.424639580602883e-05, "loss": 1.5571, "step": 806 }, { "epoch": 1.0500975927130773, "grad_norm": 0.823961615562439, "learning_rate": 7.421363040629096e-05, "loss": 1.6106, "step": 807 }, { "epoch": 1.051398828887443, "grad_norm": 0.7457985877990723, "learning_rate": 7.418086500655308e-05, "loss": 1.914, "step": 808 }, { "epoch": 1.0527000650618088, "grad_norm": 0.7074294686317444, "learning_rate": 7.41480996068152e-05, "loss": 1.6562, "step": 809 }, { "epoch": 1.0540013012361744, "grad_norm": 0.998188316822052, "learning_rate": 7.411533420707733e-05, "loss": 1.7519, "step": 810 }, { "epoch": 1.05530253741054, "grad_norm": 0.6213899850845337, "learning_rate": 7.408256880733946e-05, "loss": 1.5179, "step": 811 }, { "epoch": 1.0566037735849056, "grad_norm": 0.7580050230026245, "learning_rate": 7.404980340760158e-05, "loss": 1.8291, "step": 812 }, { "epoch": 1.0579050097592713, "grad_norm": 0.8274865746498108, "learning_rate": 7.401703800786371e-05, "loss": 1.6377, "step": 813 }, { "epoch": 1.0592062459336369, "grad_norm": 1.025498867034912, "learning_rate": 7.398427260812583e-05, "loss": 1.3476, "step": 814 }, { "epoch": 1.0605074821080025, "grad_norm": 0.6941721439361572, "learning_rate": 7.395150720838794e-05, "loss": 0.9634, "step": 815 }, { "epoch": 1.0618087182823683, "grad_norm": 1.9061594009399414, "learning_rate": 7.391874180865008e-05, "loss": 1.315, "step": 816 }, { "epoch": 1.063109954456734, "grad_norm": 0.9292385578155518, "learning_rate": 7.38859764089122e-05, "loss": 1.8489, "step": 817 }, { "epoch": 1.0644111906310996, "grad_norm": 1.1003870964050293, "learning_rate": 7.385321100917431e-05, "loss": 1.2152, "step": 818 }, { "epoch": 1.0657124268054652, "grad_norm": 0.8102130889892578, "learning_rate": 7.382044560943644e-05, "loss": 1.8892, "step": 819 }, { "epoch": 1.0670136629798308, "grad_norm": 1.239367961883545, "learning_rate": 7.378768020969856e-05, "loss": 1.4286, "step": 820 }, { "epoch": 1.0683148991541964, "grad_norm": 1.4461933374404907, "learning_rate": 7.375491480996068e-05, "loss": 1.5756, "step": 821 }, { "epoch": 1.069616135328562, "grad_norm": 0.8706846833229065, "learning_rate": 7.372214941022281e-05, "loss": 1.5568, "step": 822 }, { "epoch": 1.0709173715029279, "grad_norm": 0.8357823491096497, "learning_rate": 7.368938401048493e-05, "loss": 1.3609, "step": 823 }, { "epoch": 1.0722186076772935, "grad_norm": 0.9755945205688477, "learning_rate": 7.365661861074705e-05, "loss": 1.6222, "step": 824 }, { "epoch": 1.073519843851659, "grad_norm": 0.7612958550453186, "learning_rate": 7.362385321100918e-05, "loss": 1.7909, "step": 825 }, { "epoch": 1.0748210800260247, "grad_norm": 1.1872872114181519, "learning_rate": 7.35910878112713e-05, "loss": 1.6512, "step": 826 }, { "epoch": 1.0761223162003903, "grad_norm": 0.852663516998291, "learning_rate": 7.355832241153341e-05, "loss": 1.6492, "step": 827 }, { "epoch": 1.077423552374756, "grad_norm": 0.9593597650527954, "learning_rate": 7.352555701179555e-05, "loss": 1.5555, "step": 828 }, { "epoch": 1.0787247885491216, "grad_norm": 0.9183743000030518, "learning_rate": 7.349279161205766e-05, "loss": 1.1173, "step": 829 }, { "epoch": 1.0800260247234874, "grad_norm": 0.7066413760185242, "learning_rate": 7.34600262123198e-05, "loss": 1.1601, "step": 830 }, { "epoch": 1.081327260897853, "grad_norm": 0.6245144009590149, "learning_rate": 7.342726081258193e-05, "loss": 1.405, "step": 831 }, { "epoch": 1.0826284970722186, "grad_norm": 0.7979201674461365, "learning_rate": 7.339449541284404e-05, "loss": 1.9568, "step": 832 }, { "epoch": 1.0839297332465843, "grad_norm": 1.1682201623916626, "learning_rate": 7.336173001310616e-05, "loss": 1.1665, "step": 833 }, { "epoch": 1.0852309694209499, "grad_norm": 0.6565560698509216, "learning_rate": 7.33289646133683e-05, "loss": 1.9448, "step": 834 }, { "epoch": 1.0865322055953155, "grad_norm": 0.7160550951957703, "learning_rate": 7.329619921363041e-05, "loss": 1.694, "step": 835 }, { "epoch": 1.087833441769681, "grad_norm": 0.7564948201179504, "learning_rate": 7.326343381389253e-05, "loss": 1.9116, "step": 836 }, { "epoch": 1.089134677944047, "grad_norm": 0.7771456837654114, "learning_rate": 7.323066841415466e-05, "loss": 1.6503, "step": 837 }, { "epoch": 1.0904359141184126, "grad_norm": 0.9096713662147522, "learning_rate": 7.319790301441678e-05, "loss": 1.69, "step": 838 }, { "epoch": 1.0917371502927782, "grad_norm": 0.9597713351249695, "learning_rate": 7.31651376146789e-05, "loss": 1.3628, "step": 839 }, { "epoch": 1.0930383864671438, "grad_norm": 0.8050462007522583, "learning_rate": 7.313237221494103e-05, "loss": 1.3571, "step": 840 }, { "epoch": 1.0943396226415094, "grad_norm": 0.8489325046539307, "learning_rate": 7.309960681520315e-05, "loss": 1.9349, "step": 841 }, { "epoch": 1.095640858815875, "grad_norm": 0.9899209141731262, "learning_rate": 7.306684141546526e-05, "loss": 1.6997, "step": 842 }, { "epoch": 1.0969420949902406, "grad_norm": 0.8374925851821899, "learning_rate": 7.30340760157274e-05, "loss": 1.7999, "step": 843 }, { "epoch": 1.0982433311646065, "grad_norm": 1.039319634437561, "learning_rate": 7.300131061598951e-05, "loss": 1.7518, "step": 844 }, { "epoch": 1.099544567338972, "grad_norm": 1.2312158346176147, "learning_rate": 7.296854521625163e-05, "loss": 0.8651, "step": 845 }, { "epoch": 1.1008458035133377, "grad_norm": 0.7502791881561279, "learning_rate": 7.293577981651376e-05, "loss": 1.9642, "step": 846 }, { "epoch": 1.1021470396877033, "grad_norm": 1.0275205373764038, "learning_rate": 7.290301441677588e-05, "loss": 1.5481, "step": 847 }, { "epoch": 1.103448275862069, "grad_norm": 0.7881213426589966, "learning_rate": 7.2870249017038e-05, "loss": 1.6561, "step": 848 }, { "epoch": 1.1047495120364346, "grad_norm": 0.807431161403656, "learning_rate": 7.283748361730014e-05, "loss": 1.6354, "step": 849 }, { "epoch": 1.1060507482108002, "grad_norm": 1.0335010290145874, "learning_rate": 7.280471821756226e-05, "loss": 1.5707, "step": 850 }, { "epoch": 1.1073519843851658, "grad_norm": 0.8796849846839905, "learning_rate": 7.277195281782438e-05, "loss": 1.739, "step": 851 }, { "epoch": 1.1086532205595316, "grad_norm": 0.9731112122535706, "learning_rate": 7.273918741808651e-05, "loss": 1.7129, "step": 852 }, { "epoch": 1.1099544567338973, "grad_norm": 2.0551133155822754, "learning_rate": 7.270642201834863e-05, "loss": 1.6471, "step": 853 }, { "epoch": 1.1112556929082629, "grad_norm": 1.3990224599838257, "learning_rate": 7.267365661861075e-05, "loss": 1.4277, "step": 854 }, { "epoch": 1.1125569290826285, "grad_norm": 0.8433036208152771, "learning_rate": 7.264089121887288e-05, "loss": 1.6816, "step": 855 }, { "epoch": 1.113858165256994, "grad_norm": 0.7869728207588196, "learning_rate": 7.2608125819135e-05, "loss": 1.7053, "step": 856 }, { "epoch": 1.1151594014313597, "grad_norm": 0.8131395578384399, "learning_rate": 7.257536041939712e-05, "loss": 1.9304, "step": 857 }, { "epoch": 1.1164606376057253, "grad_norm": 3.2505669593811035, "learning_rate": 7.254259501965925e-05, "loss": 1.5711, "step": 858 }, { "epoch": 1.1177618737800912, "grad_norm": 0.8182663321495056, "learning_rate": 7.250982961992137e-05, "loss": 1.6776, "step": 859 }, { "epoch": 1.1190631099544568, "grad_norm": 0.9222348928451538, "learning_rate": 7.247706422018348e-05, "loss": 1.5949, "step": 860 }, { "epoch": 1.1203643461288224, "grad_norm": 0.7952315807342529, "learning_rate": 7.244429882044561e-05, "loss": 1.5224, "step": 861 }, { "epoch": 1.121665582303188, "grad_norm": 0.820124626159668, "learning_rate": 7.241153342070773e-05, "loss": 1.6952, "step": 862 }, { "epoch": 1.1229668184775536, "grad_norm": 0.9836868047714233, "learning_rate": 7.237876802096985e-05, "loss": 1.9254, "step": 863 }, { "epoch": 1.1242680546519193, "grad_norm": 0.9996393322944641, "learning_rate": 7.234600262123198e-05, "loss": 1.4029, "step": 864 }, { "epoch": 1.1255692908262849, "grad_norm": 0.8443908095359802, "learning_rate": 7.23132372214941e-05, "loss": 1.4906, "step": 865 }, { "epoch": 1.1268705270006507, "grad_norm": 0.8313694000244141, "learning_rate": 7.228047182175622e-05, "loss": 1.5936, "step": 866 }, { "epoch": 1.1281717631750163, "grad_norm": 0.8157771825790405, "learning_rate": 7.224770642201836e-05, "loss": 1.4709, "step": 867 }, { "epoch": 1.129472999349382, "grad_norm": 0.6848510503768921, "learning_rate": 7.221494102228048e-05, "loss": 1.5201, "step": 868 }, { "epoch": 1.1307742355237476, "grad_norm": 0.7708297371864319, "learning_rate": 7.21821756225426e-05, "loss": 1.6741, "step": 869 }, { "epoch": 1.1320754716981132, "grad_norm": 1.009873628616333, "learning_rate": 7.214941022280473e-05, "loss": 1.3602, "step": 870 }, { "epoch": 1.1333767078724788, "grad_norm": 0.6941707134246826, "learning_rate": 7.211664482306685e-05, "loss": 1.7432, "step": 871 }, { "epoch": 1.1346779440468444, "grad_norm": 0.9382712841033936, "learning_rate": 7.208387942332897e-05, "loss": 1.2428, "step": 872 }, { "epoch": 1.1359791802212102, "grad_norm": 0.9410544633865356, "learning_rate": 7.20511140235911e-05, "loss": 2.0198, "step": 873 }, { "epoch": 1.1372804163955759, "grad_norm": 0.8933817148208618, "learning_rate": 7.201834862385322e-05, "loss": 1.3589, "step": 874 }, { "epoch": 1.1385816525699415, "grad_norm": 0.7369709014892578, "learning_rate": 7.198558322411533e-05, "loss": 1.5584, "step": 875 }, { "epoch": 1.139882888744307, "grad_norm": 0.6826525330543518, "learning_rate": 7.195281782437747e-05, "loss": 0.9168, "step": 876 }, { "epoch": 1.1411841249186727, "grad_norm": 0.7951777577400208, "learning_rate": 7.192005242463958e-05, "loss": 1.2314, "step": 877 }, { "epoch": 1.1424853610930383, "grad_norm": 0.8687210083007812, "learning_rate": 7.18872870249017e-05, "loss": 1.7736, "step": 878 }, { "epoch": 1.143786597267404, "grad_norm": 0.8880603909492493, "learning_rate": 7.185452162516383e-05, "loss": 1.7111, "step": 879 }, { "epoch": 1.1450878334417696, "grad_norm": 0.8484585285186768, "learning_rate": 7.182175622542595e-05, "loss": 1.9327, "step": 880 }, { "epoch": 1.1463890696161354, "grad_norm": 0.9476994276046753, "learning_rate": 7.178899082568807e-05, "loss": 1.2356, "step": 881 }, { "epoch": 1.147690305790501, "grad_norm": 0.7982966899871826, "learning_rate": 7.17562254259502e-05, "loss": 1.9195, "step": 882 }, { "epoch": 1.1489915419648666, "grad_norm": 0.999690055847168, "learning_rate": 7.172346002621232e-05, "loss": 1.3708, "step": 883 }, { "epoch": 1.1502927781392323, "grad_norm": 0.7487162351608276, "learning_rate": 7.169069462647444e-05, "loss": 1.8314, "step": 884 }, { "epoch": 1.1515940143135979, "grad_norm": 0.7988235950469971, "learning_rate": 7.165792922673657e-05, "loss": 1.7453, "step": 885 }, { "epoch": 1.1528952504879635, "grad_norm": 0.8307318687438965, "learning_rate": 7.16251638269987e-05, "loss": 1.4423, "step": 886 }, { "epoch": 1.1541964866623293, "grad_norm": 0.899943470954895, "learning_rate": 7.159239842726082e-05, "loss": 1.1952, "step": 887 }, { "epoch": 1.155497722836695, "grad_norm": 1.596401572227478, "learning_rate": 7.155963302752295e-05, "loss": 1.7688, "step": 888 }, { "epoch": 1.1567989590110606, "grad_norm": 0.8879856467247009, "learning_rate": 7.152686762778507e-05, "loss": 1.9922, "step": 889 }, { "epoch": 1.1581001951854262, "grad_norm": 1.1052894592285156, "learning_rate": 7.149410222804718e-05, "loss": 1.2166, "step": 890 }, { "epoch": 1.1594014313597918, "grad_norm": 1.3350752592086792, "learning_rate": 7.146133682830932e-05, "loss": 1.3195, "step": 891 }, { "epoch": 1.1607026675341574, "grad_norm": 0.8114393949508667, "learning_rate": 7.142857142857143e-05, "loss": 1.7033, "step": 892 }, { "epoch": 1.162003903708523, "grad_norm": 0.8815709352493286, "learning_rate": 7.139580602883355e-05, "loss": 1.7428, "step": 893 }, { "epoch": 1.1633051398828886, "grad_norm": 1.2895179986953735, "learning_rate": 7.136304062909568e-05, "loss": 1.4803, "step": 894 }, { "epoch": 1.1646063760572545, "grad_norm": 0.826259434223175, "learning_rate": 7.13302752293578e-05, "loss": 1.54, "step": 895 }, { "epoch": 1.16590761223162, "grad_norm": 1.7639037370681763, "learning_rate": 7.129750982961992e-05, "loss": 1.3712, "step": 896 }, { "epoch": 1.1672088484059857, "grad_norm": 0.7162357568740845, "learning_rate": 7.126474442988205e-05, "loss": 1.5744, "step": 897 }, { "epoch": 1.1685100845803513, "grad_norm": 0.7369087934494019, "learning_rate": 7.123197903014417e-05, "loss": 1.7011, "step": 898 }, { "epoch": 1.169811320754717, "grad_norm": 1.3359040021896362, "learning_rate": 7.119921363040629e-05, "loss": 1.4149, "step": 899 }, { "epoch": 1.1711125569290826, "grad_norm": 0.6675812602043152, "learning_rate": 7.116644823066842e-05, "loss": 1.9376, "step": 900 }, { "epoch": 1.1724137931034484, "grad_norm": 1.0194183588027954, "learning_rate": 7.113368283093054e-05, "loss": 1.5252, "step": 901 }, { "epoch": 1.173715029277814, "grad_norm": 0.8172527551651001, "learning_rate": 7.110091743119265e-05, "loss": 2.0328, "step": 902 }, { "epoch": 1.1750162654521796, "grad_norm": 0.9126254320144653, "learning_rate": 7.106815203145479e-05, "loss": 1.3174, "step": 903 }, { "epoch": 1.1763175016265452, "grad_norm": 0.9921519756317139, "learning_rate": 7.103538663171692e-05, "loss": 1.836, "step": 904 }, { "epoch": 1.1776187378009109, "grad_norm": 0.9495877027511597, "learning_rate": 7.100262123197904e-05, "loss": 1.5072, "step": 905 }, { "epoch": 1.1789199739752765, "grad_norm": 0.7284337878227234, "learning_rate": 7.096985583224117e-05, "loss": 1.54, "step": 906 }, { "epoch": 1.180221210149642, "grad_norm": 0.8445408344268799, "learning_rate": 7.093709043250328e-05, "loss": 1.6886, "step": 907 }, { "epoch": 1.1815224463240077, "grad_norm": 1.0153337717056274, "learning_rate": 7.09043250327654e-05, "loss": 2.0574, "step": 908 }, { "epoch": 1.1828236824983733, "grad_norm": 0.9691196084022522, "learning_rate": 7.087155963302753e-05, "loss": 1.9808, "step": 909 }, { "epoch": 1.1841249186727392, "grad_norm": 0.9183889031410217, "learning_rate": 7.083879423328965e-05, "loss": 1.8526, "step": 910 }, { "epoch": 1.1854261548471048, "grad_norm": 1.1176584959030151, "learning_rate": 7.080602883355177e-05, "loss": 1.5342, "step": 911 }, { "epoch": 1.1867273910214704, "grad_norm": 1.4232932329177856, "learning_rate": 7.07732634338139e-05, "loss": 1.3233, "step": 912 }, { "epoch": 1.188028627195836, "grad_norm": 0.8222576379776001, "learning_rate": 7.074049803407602e-05, "loss": 1.5247, "step": 913 }, { "epoch": 1.1893298633702016, "grad_norm": 0.7871713638305664, "learning_rate": 7.070773263433814e-05, "loss": 1.9079, "step": 914 }, { "epoch": 1.1906310995445673, "grad_norm": 1.3005493879318237, "learning_rate": 7.067496723460027e-05, "loss": 1.2838, "step": 915 }, { "epoch": 1.191932335718933, "grad_norm": 1.3268331289291382, "learning_rate": 7.064220183486239e-05, "loss": 1.6586, "step": 916 }, { "epoch": 1.1932335718932987, "grad_norm": 0.9947080612182617, "learning_rate": 7.06094364351245e-05, "loss": 1.5524, "step": 917 }, { "epoch": 1.1945348080676643, "grad_norm": 0.8267329931259155, "learning_rate": 7.057667103538664e-05, "loss": 2.1386, "step": 918 }, { "epoch": 1.19583604424203, "grad_norm": 0.7897701859474182, "learning_rate": 7.054390563564875e-05, "loss": 1.5923, "step": 919 }, { "epoch": 1.1971372804163956, "grad_norm": 0.9993418455123901, "learning_rate": 7.051114023591087e-05, "loss": 1.4217, "step": 920 }, { "epoch": 1.1984385165907612, "grad_norm": 1.1384029388427734, "learning_rate": 7.0478374836173e-05, "loss": 1.8282, "step": 921 }, { "epoch": 1.1997397527651268, "grad_norm": 0.6628587245941162, "learning_rate": 7.044560943643512e-05, "loss": 2.0185, "step": 922 }, { "epoch": 1.2010409889394924, "grad_norm": 1.2611730098724365, "learning_rate": 7.041284403669725e-05, "loss": 1.3037, "step": 923 }, { "epoch": 1.2023422251138582, "grad_norm": 1.1846648454666138, "learning_rate": 7.038007863695938e-05, "loss": 1.2162, "step": 924 }, { "epoch": 1.2036434612882239, "grad_norm": 0.7201260328292847, "learning_rate": 7.03473132372215e-05, "loss": 1.829, "step": 925 }, { "epoch": 1.2049446974625895, "grad_norm": 0.7791245579719543, "learning_rate": 7.031454783748362e-05, "loss": 1.6407, "step": 926 }, { "epoch": 1.206245933636955, "grad_norm": 1.0425821542739868, "learning_rate": 7.028178243774575e-05, "loss": 1.5613, "step": 927 }, { "epoch": 1.2075471698113207, "grad_norm": 1.8103893995285034, "learning_rate": 7.024901703800787e-05, "loss": 1.6977, "step": 928 }, { "epoch": 1.2088484059856863, "grad_norm": 0.8963273167610168, "learning_rate": 7.021625163826999e-05, "loss": 1.8668, "step": 929 }, { "epoch": 1.2101496421600522, "grad_norm": 0.9128831624984741, "learning_rate": 7.018348623853212e-05, "loss": 1.3119, "step": 930 }, { "epoch": 1.2114508783344178, "grad_norm": 0.7835903167724609, "learning_rate": 7.015072083879424e-05, "loss": 1.4542, "step": 931 }, { "epoch": 1.2127521145087834, "grad_norm": 0.6831282377243042, "learning_rate": 7.011795543905636e-05, "loss": 1.8037, "step": 932 }, { "epoch": 1.214053350683149, "grad_norm": 1.122541069984436, "learning_rate": 7.008519003931849e-05, "loss": 1.7345, "step": 933 }, { "epoch": 1.2153545868575146, "grad_norm": 0.7036309838294983, "learning_rate": 7.00524246395806e-05, "loss": 2.026, "step": 934 }, { "epoch": 1.2166558230318802, "grad_norm": 0.7880788445472717, "learning_rate": 7.001965923984272e-05, "loss": 1.705, "step": 935 }, { "epoch": 1.2179570592062459, "grad_norm": 0.793428897857666, "learning_rate": 6.998689384010485e-05, "loss": 1.7416, "step": 936 }, { "epoch": 1.2192582953806115, "grad_norm": 0.8155110478401184, "learning_rate": 6.995412844036697e-05, "loss": 1.6895, "step": 937 }, { "epoch": 1.2205595315549773, "grad_norm": 0.7480125427246094, "learning_rate": 6.992136304062909e-05, "loss": 1.5864, "step": 938 }, { "epoch": 1.221860767729343, "grad_norm": 0.8781589865684509, "learning_rate": 6.988859764089122e-05, "loss": 1.3822, "step": 939 }, { "epoch": 1.2231620039037086, "grad_norm": 1.0218061208724976, "learning_rate": 6.985583224115334e-05, "loss": 1.8464, "step": 940 }, { "epoch": 1.2244632400780742, "grad_norm": 0.7555110454559326, "learning_rate": 6.982306684141546e-05, "loss": 1.8565, "step": 941 }, { "epoch": 1.2257644762524398, "grad_norm": 0.7511152625083923, "learning_rate": 6.97903014416776e-05, "loss": 2.0294, "step": 942 }, { "epoch": 1.2270657124268054, "grad_norm": 1.0907107591629028, "learning_rate": 6.975753604193972e-05, "loss": 1.4102, "step": 943 }, { "epoch": 1.228366948601171, "grad_norm": 0.7848272323608398, "learning_rate": 6.972477064220184e-05, "loss": 1.3131, "step": 944 }, { "epoch": 1.2296681847755369, "grad_norm": 1.0624936819076538, "learning_rate": 6.969200524246397e-05, "loss": 1.4849, "step": 945 }, { "epoch": 1.2309694209499025, "grad_norm": 1.1583614349365234, "learning_rate": 6.965923984272609e-05, "loss": 1.4857, "step": 946 }, { "epoch": 1.232270657124268, "grad_norm": 0.6758008599281311, "learning_rate": 6.96264744429882e-05, "loss": 1.8642, "step": 947 }, { "epoch": 1.2335718932986337, "grad_norm": 0.5795673131942749, "learning_rate": 6.959370904325034e-05, "loss": 1.7914, "step": 948 }, { "epoch": 1.2348731294729993, "grad_norm": 1.3749276399612427, "learning_rate": 6.956094364351246e-05, "loss": 1.8115, "step": 949 }, { "epoch": 1.236174365647365, "grad_norm": 0.7618027329444885, "learning_rate": 6.952817824377457e-05, "loss": 1.2783, "step": 950 }, { "epoch": 1.2374756018217306, "grad_norm": 0.5912798643112183, "learning_rate": 6.94954128440367e-05, "loss": 1.8727, "step": 951 }, { "epoch": 1.2387768379960962, "grad_norm": 1.0082651376724243, "learning_rate": 6.946264744429882e-05, "loss": 1.4315, "step": 952 }, { "epoch": 1.240078074170462, "grad_norm": 1.0466032028198242, "learning_rate": 6.942988204456094e-05, "loss": 1.5635, "step": 953 }, { "epoch": 1.2413793103448276, "grad_norm": 0.7676052451133728, "learning_rate": 6.939711664482307e-05, "loss": 1.7831, "step": 954 }, { "epoch": 1.2426805465191932, "grad_norm": 1.0872946977615356, "learning_rate": 6.936435124508519e-05, "loss": 1.7658, "step": 955 }, { "epoch": 1.2439817826935589, "grad_norm": 0.7136397361755371, "learning_rate": 6.933158584534731e-05, "loss": 1.8095, "step": 956 }, { "epoch": 1.2452830188679245, "grad_norm": 0.7722171545028687, "learning_rate": 6.929882044560944e-05, "loss": 1.0245, "step": 957 }, { "epoch": 1.24658425504229, "grad_norm": 0.919909656047821, "learning_rate": 6.926605504587156e-05, "loss": 1.1152, "step": 958 }, { "epoch": 1.247885491216656, "grad_norm": 0.42374274134635925, "learning_rate": 6.923328964613368e-05, "loss": 0.7481, "step": 959 }, { "epoch": 1.2491867273910215, "grad_norm": 0.7017391324043274, "learning_rate": 6.920052424639582e-05, "loss": 1.4808, "step": 960 }, { "epoch": 1.2504879635653872, "grad_norm": 1.510352373123169, "learning_rate": 6.916775884665794e-05, "loss": 2.0383, "step": 961 }, { "epoch": 1.2517891997397528, "grad_norm": 0.6219519972801208, "learning_rate": 6.913499344692006e-05, "loss": 1.2555, "step": 962 }, { "epoch": 1.2530904359141184, "grad_norm": 1.082633137702942, "learning_rate": 6.910222804718219e-05, "loss": 2.0445, "step": 963 }, { "epoch": 1.254391672088484, "grad_norm": 1.224017858505249, "learning_rate": 6.90694626474443e-05, "loss": 1.6173, "step": 964 }, { "epoch": 1.2556929082628496, "grad_norm": 0.6671722531318665, "learning_rate": 6.903669724770642e-05, "loss": 1.7886, "step": 965 }, { "epoch": 1.2569941444372152, "grad_norm": 0.8693065047264099, "learning_rate": 6.900393184796856e-05, "loss": 1.8244, "step": 966 }, { "epoch": 1.258295380611581, "grad_norm": 0.7832571864128113, "learning_rate": 6.897116644823067e-05, "loss": 1.0748, "step": 967 }, { "epoch": 1.2595966167859467, "grad_norm": 1.1029421091079712, "learning_rate": 6.893840104849279e-05, "loss": 2.0041, "step": 968 }, { "epoch": 1.2608978529603123, "grad_norm": 0.7565571665763855, "learning_rate": 6.890563564875492e-05, "loss": 1.951, "step": 969 }, { "epoch": 1.262199089134678, "grad_norm": 0.8847816586494446, "learning_rate": 6.887287024901704e-05, "loss": 1.8209, "step": 970 }, { "epoch": 1.2635003253090435, "grad_norm": 0.7797966003417969, "learning_rate": 6.884010484927916e-05, "loss": 1.6448, "step": 971 }, { "epoch": 1.2648015614834092, "grad_norm": 0.8273074626922607, "learning_rate": 6.880733944954129e-05, "loss": 1.7516, "step": 972 }, { "epoch": 1.266102797657775, "grad_norm": 0.7152173519134521, "learning_rate": 6.877457404980341e-05, "loss": 1.1542, "step": 973 }, { "epoch": 1.2674040338321406, "grad_norm": 1.007451057434082, "learning_rate": 6.874180865006553e-05, "loss": 1.8292, "step": 974 }, { "epoch": 1.2687052700065062, "grad_norm": 0.7937166690826416, "learning_rate": 6.870904325032766e-05, "loss": 1.3528, "step": 975 }, { "epoch": 1.2700065061808719, "grad_norm": 1.276218295097351, "learning_rate": 6.867627785058978e-05, "loss": 1.4737, "step": 976 }, { "epoch": 1.2713077423552375, "grad_norm": 0.569302499294281, "learning_rate": 6.86435124508519e-05, "loss": 1.2827, "step": 977 }, { "epoch": 1.272608978529603, "grad_norm": 0.8590444922447205, "learning_rate": 6.861074705111403e-05, "loss": 1.4973, "step": 978 }, { "epoch": 1.2739102147039687, "grad_norm": 0.9621579051017761, "learning_rate": 6.857798165137616e-05, "loss": 1.578, "step": 979 }, { "epoch": 1.2752114508783343, "grad_norm": 1.1873749494552612, "learning_rate": 6.854521625163828e-05, "loss": 2.1051, "step": 980 }, { "epoch": 1.2765126870527, "grad_norm": 1.9063034057617188, "learning_rate": 6.851245085190041e-05, "loss": 1.8366, "step": 981 }, { "epoch": 1.2778139232270658, "grad_norm": 0.9726938009262085, "learning_rate": 6.847968545216252e-05, "loss": 1.4403, "step": 982 }, { "epoch": 1.2791151594014314, "grad_norm": 1.1927934885025024, "learning_rate": 6.844692005242464e-05, "loss": 1.4833, "step": 983 }, { "epoch": 1.280416395575797, "grad_norm": 0.8170132637023926, "learning_rate": 6.841415465268677e-05, "loss": 1.6857, "step": 984 }, { "epoch": 1.2817176317501626, "grad_norm": 0.7814056277275085, "learning_rate": 6.838138925294889e-05, "loss": 1.685, "step": 985 }, { "epoch": 1.2830188679245282, "grad_norm": 0.6441072821617126, "learning_rate": 6.834862385321101e-05, "loss": 2.1139, "step": 986 }, { "epoch": 1.284320104098894, "grad_norm": 0.7550761699676514, "learning_rate": 6.831585845347314e-05, "loss": 1.1923, "step": 987 }, { "epoch": 1.2856213402732597, "grad_norm": 1.2927114963531494, "learning_rate": 6.828309305373526e-05, "loss": 1.9201, "step": 988 }, { "epoch": 1.2869225764476253, "grad_norm": 1.1189708709716797, "learning_rate": 6.825032765399738e-05, "loss": 1.2497, "step": 989 }, { "epoch": 1.288223812621991, "grad_norm": 0.8026967644691467, "learning_rate": 6.821756225425951e-05, "loss": 1.8238, "step": 990 }, { "epoch": 1.2895250487963565, "grad_norm": 0.7840678691864014, "learning_rate": 6.818479685452163e-05, "loss": 1.8882, "step": 991 }, { "epoch": 1.2908262849707222, "grad_norm": 0.6625152826309204, "learning_rate": 6.815203145478375e-05, "loss": 2.0278, "step": 992 }, { "epoch": 1.2921275211450878, "grad_norm": 0.8204750418663025, "learning_rate": 6.811926605504588e-05, "loss": 1.7571, "step": 993 }, { "epoch": 1.2934287573194534, "grad_norm": 0.7582717537879944, "learning_rate": 6.8086500655308e-05, "loss": 1.8099, "step": 994 }, { "epoch": 1.294729993493819, "grad_norm": 0.7991265654563904, "learning_rate": 6.805373525557011e-05, "loss": 1.974, "step": 995 }, { "epoch": 1.2960312296681848, "grad_norm": 0.8572267293930054, "learning_rate": 6.802096985583224e-05, "loss": 1.6881, "step": 996 }, { "epoch": 1.2973324658425505, "grad_norm": 0.9528511762619019, "learning_rate": 6.798820445609436e-05, "loss": 2.0306, "step": 997 }, { "epoch": 1.298633702016916, "grad_norm": 1.756341576576233, "learning_rate": 6.79554390563565e-05, "loss": 1.953, "step": 998 }, { "epoch": 1.2999349381912817, "grad_norm": 1.701744556427002, "learning_rate": 6.792267365661862e-05, "loss": 1.8054, "step": 999 }, { "epoch": 1.3012361743656473, "grad_norm": 0.7313840985298157, "learning_rate": 6.788990825688074e-05, "loss": 1.8182, "step": 1000 }, { "epoch": 1.302537410540013, "grad_norm": 0.8177651762962341, "learning_rate": 6.785714285714286e-05, "loss": 1.7276, "step": 1001 }, { "epoch": 1.3038386467143788, "grad_norm": 1.0579301118850708, "learning_rate": 6.782437745740499e-05, "loss": 1.0632, "step": 1002 }, { "epoch": 1.3051398828887444, "grad_norm": 0.6622770428657532, "learning_rate": 6.779161205766711e-05, "loss": 1.0768, "step": 1003 }, { "epoch": 1.30644111906311, "grad_norm": 0.6889738440513611, "learning_rate": 6.775884665792923e-05, "loss": 2.1004, "step": 1004 }, { "epoch": 1.3077423552374756, "grad_norm": 0.7540472149848938, "learning_rate": 6.772608125819136e-05, "loss": 1.4457, "step": 1005 }, { "epoch": 1.3090435914118412, "grad_norm": 0.9096068739891052, "learning_rate": 6.769331585845348e-05, "loss": 1.5722, "step": 1006 }, { "epoch": 1.3103448275862069, "grad_norm": 0.7821688652038574, "learning_rate": 6.76605504587156e-05, "loss": 1.3282, "step": 1007 }, { "epoch": 1.3116460637605725, "grad_norm": 0.857506275177002, "learning_rate": 6.762778505897773e-05, "loss": 1.7513, "step": 1008 }, { "epoch": 1.312947299934938, "grad_norm": 0.7883008122444153, "learning_rate": 6.759501965923985e-05, "loss": 1.6103, "step": 1009 }, { "epoch": 1.3142485361093037, "grad_norm": 0.7950558662414551, "learning_rate": 6.756225425950196e-05, "loss": 1.2548, "step": 1010 }, { "epoch": 1.3155497722836695, "grad_norm": 0.7552403211593628, "learning_rate": 6.75294888597641e-05, "loss": 1.6214, "step": 1011 }, { "epoch": 1.3168510084580352, "grad_norm": 1.018877387046814, "learning_rate": 6.749672346002621e-05, "loss": 1.7069, "step": 1012 }, { "epoch": 1.3181522446324008, "grad_norm": 0.9168699383735657, "learning_rate": 6.746395806028833e-05, "loss": 1.5426, "step": 1013 }, { "epoch": 1.3194534808067664, "grad_norm": 0.8144714832305908, "learning_rate": 6.743119266055046e-05, "loss": 2.0193, "step": 1014 }, { "epoch": 1.320754716981132, "grad_norm": 0.7221648693084717, "learning_rate": 6.739842726081258e-05, "loss": 1.873, "step": 1015 }, { "epoch": 1.3220559531554978, "grad_norm": 0.8393880128860474, "learning_rate": 6.736566186107471e-05, "loss": 1.5477, "step": 1016 }, { "epoch": 1.3233571893298635, "grad_norm": 0.899076521396637, "learning_rate": 6.733289646133683e-05, "loss": 1.8426, "step": 1017 }, { "epoch": 1.324658425504229, "grad_norm": 0.9221969842910767, "learning_rate": 6.730013106159896e-05, "loss": 1.8101, "step": 1018 }, { "epoch": 1.3259596616785947, "grad_norm": 1.344862937927246, "learning_rate": 6.726736566186108e-05, "loss": 1.5446, "step": 1019 }, { "epoch": 1.3272608978529603, "grad_norm": 0.7244084477424622, "learning_rate": 6.723460026212321e-05, "loss": 1.9206, "step": 1020 }, { "epoch": 1.328562134027326, "grad_norm": 1.6741523742675781, "learning_rate": 6.720183486238533e-05, "loss": 1.8339, "step": 1021 }, { "epoch": 1.3298633702016915, "grad_norm": 0.7968908548355103, "learning_rate": 6.716906946264745e-05, "loss": 1.6694, "step": 1022 }, { "epoch": 1.3311646063760572, "grad_norm": 1.0471298694610596, "learning_rate": 6.713630406290958e-05, "loss": 1.861, "step": 1023 }, { "epoch": 1.3324658425504228, "grad_norm": 0.7247074246406555, "learning_rate": 6.71035386631717e-05, "loss": 1.6175, "step": 1024 }, { "epoch": 1.3337670787247886, "grad_norm": 0.870715320110321, "learning_rate": 6.707077326343381e-05, "loss": 1.3672, "step": 1025 }, { "epoch": 1.3350683148991542, "grad_norm": 0.9406316876411438, "learning_rate": 6.703800786369595e-05, "loss": 1.963, "step": 1026 }, { "epoch": 1.3363695510735198, "grad_norm": 1.0075844526290894, "learning_rate": 6.700524246395806e-05, "loss": 1.3704, "step": 1027 }, { "epoch": 1.3376707872478855, "grad_norm": 0.73492431640625, "learning_rate": 6.697247706422018e-05, "loss": 1.9451, "step": 1028 }, { "epoch": 1.338972023422251, "grad_norm": 0.6045480966567993, "learning_rate": 6.693971166448231e-05, "loss": 1.727, "step": 1029 }, { "epoch": 1.340273259596617, "grad_norm": 0.9422447085380554, "learning_rate": 6.690694626474443e-05, "loss": 2.1076, "step": 1030 }, { "epoch": 1.3415744957709825, "grad_norm": 0.8174301981925964, "learning_rate": 6.687418086500655e-05, "loss": 1.2132, "step": 1031 }, { "epoch": 1.3428757319453482, "grad_norm": 0.8765426278114319, "learning_rate": 6.684141546526868e-05, "loss": 1.61, "step": 1032 }, { "epoch": 1.3441769681197138, "grad_norm": 0.8334310054779053, "learning_rate": 6.68086500655308e-05, "loss": 1.7612, "step": 1033 }, { "epoch": 1.3454782042940794, "grad_norm": 0.8884925842285156, "learning_rate": 6.677588466579292e-05, "loss": 1.0347, "step": 1034 }, { "epoch": 1.346779440468445, "grad_norm": 0.8548498153686523, "learning_rate": 6.674311926605505e-05, "loss": 1.3418, "step": 1035 }, { "epoch": 1.3480806766428106, "grad_norm": 0.924168050289154, "learning_rate": 6.671035386631718e-05, "loss": 1.5072, "step": 1036 }, { "epoch": 1.3493819128171762, "grad_norm": 0.870715320110321, "learning_rate": 6.66775884665793e-05, "loss": 1.7957, "step": 1037 }, { "epoch": 1.3506831489915418, "grad_norm": 1.0462228059768677, "learning_rate": 6.664482306684143e-05, "loss": 1.6585, "step": 1038 }, { "epoch": 1.3519843851659077, "grad_norm": 0.9244155883789062, "learning_rate": 6.661205766710355e-05, "loss": 1.6313, "step": 1039 }, { "epoch": 1.3532856213402733, "grad_norm": 0.8604432940483093, "learning_rate": 6.657929226736566e-05, "loss": 1.8062, "step": 1040 }, { "epoch": 1.354586857514639, "grad_norm": 0.67204350233078, "learning_rate": 6.65465268676278e-05, "loss": 1.5743, "step": 1041 }, { "epoch": 1.3558880936890045, "grad_norm": 0.8795926570892334, "learning_rate": 6.651376146788991e-05, "loss": 1.7, "step": 1042 }, { "epoch": 1.3571893298633702, "grad_norm": 0.8056595325469971, "learning_rate": 6.648099606815203e-05, "loss": 1.7269, "step": 1043 }, { "epoch": 1.3584905660377358, "grad_norm": 0.7296256422996521, "learning_rate": 6.644823066841416e-05, "loss": 1.8088, "step": 1044 }, { "epoch": 1.3597918022121016, "grad_norm": 1.0687034130096436, "learning_rate": 6.641546526867628e-05, "loss": 1.9595, "step": 1045 }, { "epoch": 1.3610930383864672, "grad_norm": 0.9665893912315369, "learning_rate": 6.63826998689384e-05, "loss": 1.9174, "step": 1046 }, { "epoch": 1.3623942745608328, "grad_norm": 0.7311649918556213, "learning_rate": 6.634993446920053e-05, "loss": 1.6674, "step": 1047 }, { "epoch": 1.3636955107351985, "grad_norm": 0.8123180866241455, "learning_rate": 6.631716906946265e-05, "loss": 1.8399, "step": 1048 }, { "epoch": 1.364996746909564, "grad_norm": 0.9701938033103943, "learning_rate": 6.628440366972477e-05, "loss": 1.6836, "step": 1049 }, { "epoch": 1.3662979830839297, "grad_norm": 1.0750563144683838, "learning_rate": 6.62516382699869e-05, "loss": 1.9195, "step": 1050 }, { "epoch": 1.3675992192582953, "grad_norm": 0.7289652228355408, "learning_rate": 6.621887287024902e-05, "loss": 1.7797, "step": 1051 }, { "epoch": 1.368900455432661, "grad_norm": 0.8384411931037903, "learning_rate": 6.618610747051113e-05, "loss": 1.3019, "step": 1052 }, { "epoch": 1.3702016916070265, "grad_norm": 0.7335941791534424, "learning_rate": 6.615334207077327e-05, "loss": 1.8688, "step": 1053 }, { "epoch": 1.3715029277813924, "grad_norm": 0.8539342284202576, "learning_rate": 6.61205766710354e-05, "loss": 0.8873, "step": 1054 }, { "epoch": 1.372804163955758, "grad_norm": 0.8013415336608887, "learning_rate": 6.608781127129752e-05, "loss": 1.9041, "step": 1055 }, { "epoch": 1.3741054001301236, "grad_norm": 1.1399561166763306, "learning_rate": 6.605504587155963e-05, "loss": 1.9502, "step": 1056 }, { "epoch": 1.3754066363044892, "grad_norm": 0.7471815347671509, "learning_rate": 6.602228047182176e-05, "loss": 1.3018, "step": 1057 }, { "epoch": 1.3767078724788548, "grad_norm": 0.822234034538269, "learning_rate": 6.598951507208388e-05, "loss": 1.6571, "step": 1058 }, { "epoch": 1.3780091086532207, "grad_norm": 0.8119014501571655, "learning_rate": 6.595674967234601e-05, "loss": 1.7772, "step": 1059 }, { "epoch": 1.3793103448275863, "grad_norm": 0.6745425462722778, "learning_rate": 6.592398427260813e-05, "loss": 1.8354, "step": 1060 }, { "epoch": 1.380611581001952, "grad_norm": 0.9534369707107544, "learning_rate": 6.589121887287025e-05, "loss": 1.4247, "step": 1061 }, { "epoch": 1.3819128171763175, "grad_norm": 1.1248557567596436, "learning_rate": 6.585845347313238e-05, "loss": 1.2809, "step": 1062 }, { "epoch": 1.3832140533506831, "grad_norm": 1.3973145484924316, "learning_rate": 6.58256880733945e-05, "loss": 1.5382, "step": 1063 }, { "epoch": 1.3845152895250488, "grad_norm": 1.845997929573059, "learning_rate": 6.579292267365662e-05, "loss": 1.8051, "step": 1064 }, { "epoch": 1.3858165256994144, "grad_norm": 0.8176946043968201, "learning_rate": 6.576015727391875e-05, "loss": 1.4997, "step": 1065 }, { "epoch": 1.38711776187378, "grad_norm": 0.9224151968955994, "learning_rate": 6.572739187418087e-05, "loss": 1.3563, "step": 1066 }, { "epoch": 1.3884189980481456, "grad_norm": 0.7783791422843933, "learning_rate": 6.569462647444299e-05, "loss": 1.1043, "step": 1067 }, { "epoch": 1.3897202342225115, "grad_norm": 0.8008193373680115, "learning_rate": 6.566186107470512e-05, "loss": 1.9351, "step": 1068 }, { "epoch": 1.391021470396877, "grad_norm": 0.9052460789680481, "learning_rate": 6.562909567496723e-05, "loss": 1.7336, "step": 1069 }, { "epoch": 1.3923227065712427, "grad_norm": 0.9266341924667358, "learning_rate": 6.559633027522935e-05, "loss": 1.209, "step": 1070 }, { "epoch": 1.3936239427456083, "grad_norm": 0.8496604561805725, "learning_rate": 6.556356487549148e-05, "loss": 1.6029, "step": 1071 }, { "epoch": 1.394925178919974, "grad_norm": 1.528765082359314, "learning_rate": 6.553079947575362e-05, "loss": 1.1688, "step": 1072 }, { "epoch": 1.3962264150943398, "grad_norm": 0.7975212335586548, "learning_rate": 6.549803407601573e-05, "loss": 1.8567, "step": 1073 }, { "epoch": 1.3975276512687054, "grad_norm": 0.8588680624961853, "learning_rate": 6.546526867627785e-05, "loss": 1.8845, "step": 1074 }, { "epoch": 1.398828887443071, "grad_norm": 1.0010955333709717, "learning_rate": 6.543250327653998e-05, "loss": 1.6143, "step": 1075 }, { "epoch": 1.4001301236174366, "grad_norm": 0.7804883718490601, "learning_rate": 6.53997378768021e-05, "loss": 1.097, "step": 1076 }, { "epoch": 1.4014313597918022, "grad_norm": 1.0226562023162842, "learning_rate": 6.536697247706422e-05, "loss": 1.6786, "step": 1077 }, { "epoch": 1.4027325959661678, "grad_norm": 0.8562804460525513, "learning_rate": 6.533420707732635e-05, "loss": 1.8026, "step": 1078 }, { "epoch": 1.4040338321405335, "grad_norm": 0.6481124758720398, "learning_rate": 6.530144167758847e-05, "loss": 1.8555, "step": 1079 }, { "epoch": 1.405335068314899, "grad_norm": 0.8342714309692383, "learning_rate": 6.52686762778506e-05, "loss": 1.9993, "step": 1080 }, { "epoch": 1.4066363044892647, "grad_norm": 0.8712189197540283, "learning_rate": 6.523591087811272e-05, "loss": 1.3999, "step": 1081 }, { "epoch": 1.4079375406636305, "grad_norm": 0.9618698358535767, "learning_rate": 6.520314547837484e-05, "loss": 1.0846, "step": 1082 }, { "epoch": 1.4092387768379961, "grad_norm": 0.7544875144958496, "learning_rate": 6.517038007863697e-05, "loss": 1.6284, "step": 1083 }, { "epoch": 1.4105400130123618, "grad_norm": 0.8632603883743286, "learning_rate": 6.513761467889909e-05, "loss": 0.9584, "step": 1084 }, { "epoch": 1.4118412491867274, "grad_norm": 0.7552632093429565, "learning_rate": 6.51048492791612e-05, "loss": 2.0205, "step": 1085 }, { "epoch": 1.413142485361093, "grad_norm": 0.6472933888435364, "learning_rate": 6.507208387942333e-05, "loss": 1.3492, "step": 1086 }, { "epoch": 1.4144437215354586, "grad_norm": 0.9084169268608093, "learning_rate": 6.503931847968545e-05, "loss": 1.6551, "step": 1087 }, { "epoch": 1.4157449577098244, "grad_norm": 1.2157658338546753, "learning_rate": 6.500655307994757e-05, "loss": 1.3966, "step": 1088 }, { "epoch": 1.41704619388419, "grad_norm": 1.164246916770935, "learning_rate": 6.49737876802097e-05, "loss": 1.3238, "step": 1089 }, { "epoch": 1.4183474300585557, "grad_norm": 0.7860562801361084, "learning_rate": 6.494102228047182e-05, "loss": 1.8231, "step": 1090 }, { "epoch": 1.4196486662329213, "grad_norm": 0.6832435131072998, "learning_rate": 6.490825688073395e-05, "loss": 1.6712, "step": 1091 }, { "epoch": 1.420949902407287, "grad_norm": 0.9761883020401001, "learning_rate": 6.487549148099607e-05, "loss": 1.3216, "step": 1092 }, { "epoch": 1.4222511385816525, "grad_norm": 0.7399343848228455, "learning_rate": 6.48427260812582e-05, "loss": 2.0096, "step": 1093 }, { "epoch": 1.4235523747560181, "grad_norm": 1.4295475482940674, "learning_rate": 6.480996068152032e-05, "loss": 2.084, "step": 1094 }, { "epoch": 1.4248536109303838, "grad_norm": 0.7949061989784241, "learning_rate": 6.477719528178244e-05, "loss": 1.3487, "step": 1095 }, { "epoch": 1.4261548471047494, "grad_norm": 0.8276298642158508, "learning_rate": 6.474442988204457e-05, "loss": 1.2877, "step": 1096 }, { "epoch": 1.4274560832791152, "grad_norm": 0.9739117622375488, "learning_rate": 6.471166448230669e-05, "loss": 1.6807, "step": 1097 }, { "epoch": 1.4287573194534808, "grad_norm": 1.3009605407714844, "learning_rate": 6.46788990825688e-05, "loss": 1.5135, "step": 1098 }, { "epoch": 1.4300585556278465, "grad_norm": 0.9574446082115173, "learning_rate": 6.464613368283094e-05, "loss": 1.3971, "step": 1099 }, { "epoch": 1.431359791802212, "grad_norm": 0.8037934899330139, "learning_rate": 6.461336828309305e-05, "loss": 1.3967, "step": 1100 }, { "epoch": 1.4326610279765777, "grad_norm": 0.6004930138587952, "learning_rate": 6.458060288335519e-05, "loss": 1.8873, "step": 1101 }, { "epoch": 1.4339622641509435, "grad_norm": 0.9848613739013672, "learning_rate": 6.45478374836173e-05, "loss": 1.8884, "step": 1102 }, { "epoch": 1.4352635003253091, "grad_norm": 0.6587502956390381, "learning_rate": 6.451507208387942e-05, "loss": 1.484, "step": 1103 }, { "epoch": 1.4365647364996748, "grad_norm": 0.9272117018699646, "learning_rate": 6.448230668414155e-05, "loss": 1.0928, "step": 1104 }, { "epoch": 1.4378659726740404, "grad_norm": 1.442747712135315, "learning_rate": 6.444954128440367e-05, "loss": 2.1565, "step": 1105 }, { "epoch": 1.439167208848406, "grad_norm": 0.8426872491836548, "learning_rate": 6.441677588466579e-05, "loss": 1.3404, "step": 1106 }, { "epoch": 1.4404684450227716, "grad_norm": 1.6124274730682373, "learning_rate": 6.438401048492792e-05, "loss": 1.366, "step": 1107 }, { "epoch": 1.4417696811971372, "grad_norm": 2.1050853729248047, "learning_rate": 6.435124508519004e-05, "loss": 1.1148, "step": 1108 }, { "epoch": 1.4430709173715028, "grad_norm": 0.7825994491577148, "learning_rate": 6.431847968545217e-05, "loss": 1.7519, "step": 1109 }, { "epoch": 1.4443721535458685, "grad_norm": 1.343438982963562, "learning_rate": 6.428571428571429e-05, "loss": 1.2548, "step": 1110 }, { "epoch": 1.4456733897202343, "grad_norm": 0.9003665447235107, "learning_rate": 6.425294888597642e-05, "loss": 1.8878, "step": 1111 }, { "epoch": 1.4469746258946, "grad_norm": 0.9874491095542908, "learning_rate": 6.422018348623854e-05, "loss": 1.6371, "step": 1112 }, { "epoch": 1.4482758620689655, "grad_norm": 0.9111554026603699, "learning_rate": 6.418741808650066e-05, "loss": 1.4582, "step": 1113 }, { "epoch": 1.4495770982433311, "grad_norm": 0.7855933308601379, "learning_rate": 6.415465268676279e-05, "loss": 1.7017, "step": 1114 }, { "epoch": 1.4508783344176968, "grad_norm": 0.8819959759712219, "learning_rate": 6.41218872870249e-05, "loss": 2.1372, "step": 1115 }, { "epoch": 1.4521795705920626, "grad_norm": 0.8369624018669128, "learning_rate": 6.408912188728702e-05, "loss": 1.8581, "step": 1116 }, { "epoch": 1.4534808067664282, "grad_norm": 0.9745899438858032, "learning_rate": 6.405635648754915e-05, "loss": 1.8603, "step": 1117 }, { "epoch": 1.4547820429407938, "grad_norm": 0.912855863571167, "learning_rate": 6.402359108781127e-05, "loss": 1.573, "step": 1118 }, { "epoch": 1.4560832791151594, "grad_norm": 0.8576797246932983, "learning_rate": 6.39908256880734e-05, "loss": 1.8504, "step": 1119 }, { "epoch": 1.457384515289525, "grad_norm": 0.8096101880073547, "learning_rate": 6.395806028833552e-05, "loss": 1.7079, "step": 1120 }, { "epoch": 1.4586857514638907, "grad_norm": 0.7461526393890381, "learning_rate": 6.392529488859764e-05, "loss": 1.6695, "step": 1121 }, { "epoch": 1.4599869876382563, "grad_norm": 1.0096186399459839, "learning_rate": 6.389252948885977e-05, "loss": 1.864, "step": 1122 }, { "epoch": 1.461288223812622, "grad_norm": 0.8772860765457153, "learning_rate": 6.385976408912189e-05, "loss": 1.6772, "step": 1123 }, { "epoch": 1.4625894599869875, "grad_norm": 0.7923038005828857, "learning_rate": 6.382699868938401e-05, "loss": 1.595, "step": 1124 }, { "epoch": 1.4638906961613534, "grad_norm": 0.7237275242805481, "learning_rate": 6.379423328964614e-05, "loss": 1.3506, "step": 1125 }, { "epoch": 1.465191932335719, "grad_norm": 1.2001687288284302, "learning_rate": 6.376146788990826e-05, "loss": 1.2211, "step": 1126 }, { "epoch": 1.4664931685100846, "grad_norm": 0.8569205403327942, "learning_rate": 6.372870249017037e-05, "loss": 1.2983, "step": 1127 }, { "epoch": 1.4677944046844502, "grad_norm": 0.7708868384361267, "learning_rate": 6.36959370904325e-05, "loss": 1.6047, "step": 1128 }, { "epoch": 1.4690956408588158, "grad_norm": 1.5093634128570557, "learning_rate": 6.366317169069464e-05, "loss": 1.3179, "step": 1129 }, { "epoch": 1.4703968770331814, "grad_norm": 1.070641279220581, "learning_rate": 6.363040629095676e-05, "loss": 1.3494, "step": 1130 }, { "epoch": 1.4716981132075473, "grad_norm": 0.9528182148933411, "learning_rate": 6.359764089121887e-05, "loss": 1.5258, "step": 1131 }, { "epoch": 1.472999349381913, "grad_norm": 1.2565696239471436, "learning_rate": 6.3564875491481e-05, "loss": 1.57, "step": 1132 }, { "epoch": 1.4743005855562785, "grad_norm": 0.6364519000053406, "learning_rate": 6.353211009174312e-05, "loss": 1.2908, "step": 1133 }, { "epoch": 1.4756018217306441, "grad_norm": 1.163734793663025, "learning_rate": 6.349934469200524e-05, "loss": 1.6146, "step": 1134 }, { "epoch": 1.4769030579050098, "grad_norm": 1.1801888942718506, "learning_rate": 6.346657929226737e-05, "loss": 1.6644, "step": 1135 }, { "epoch": 1.4782042940793754, "grad_norm": 1.0038005113601685, "learning_rate": 6.343381389252949e-05, "loss": 1.4551, "step": 1136 }, { "epoch": 1.479505530253741, "grad_norm": 1.0582692623138428, "learning_rate": 6.340104849279161e-05, "loss": 1.2192, "step": 1137 }, { "epoch": 1.4808067664281066, "grad_norm": 1.0052828788757324, "learning_rate": 6.336828309305374e-05, "loss": 1.919, "step": 1138 }, { "epoch": 1.4821080026024722, "grad_norm": 0.8737993240356445, "learning_rate": 6.333551769331586e-05, "loss": 1.813, "step": 1139 }, { "epoch": 1.483409238776838, "grad_norm": 1.2508288621902466, "learning_rate": 6.330275229357799e-05, "loss": 1.7725, "step": 1140 }, { "epoch": 1.4847104749512037, "grad_norm": 1.3787819147109985, "learning_rate": 6.326998689384011e-05, "loss": 1.3241, "step": 1141 }, { "epoch": 1.4860117111255693, "grad_norm": 0.7962576746940613, "learning_rate": 6.323722149410223e-05, "loss": 1.6464, "step": 1142 }, { "epoch": 1.487312947299935, "grad_norm": 0.8308787941932678, "learning_rate": 6.320445609436436e-05, "loss": 1.4875, "step": 1143 }, { "epoch": 1.4886141834743005, "grad_norm": 1.185569167137146, "learning_rate": 6.317169069462647e-05, "loss": 1.7395, "step": 1144 }, { "epoch": 1.4899154196486664, "grad_norm": 0.8194323778152466, "learning_rate": 6.313892529488859e-05, "loss": 1.5849, "step": 1145 }, { "epoch": 1.491216655823032, "grad_norm": 0.5544121861457825, "learning_rate": 6.310615989515072e-05, "loss": 1.7962, "step": 1146 }, { "epoch": 1.4925178919973976, "grad_norm": 0.9360016584396362, "learning_rate": 6.307339449541286e-05, "loss": 1.712, "step": 1147 }, { "epoch": 1.4938191281717632, "grad_norm": 1.0033648014068604, "learning_rate": 6.304062909567497e-05, "loss": 1.6763, "step": 1148 }, { "epoch": 1.4951203643461288, "grad_norm": 1.007645606994629, "learning_rate": 6.300786369593709e-05, "loss": 1.457, "step": 1149 }, { "epoch": 1.4964216005204944, "grad_norm": 1.8160640001296997, "learning_rate": 6.297509829619922e-05, "loss": 1.5138, "step": 1150 }, { "epoch": 1.49772283669486, "grad_norm": 0.9474389553070068, "learning_rate": 6.294233289646134e-05, "loss": 1.6265, "step": 1151 }, { "epoch": 1.4990240728692257, "grad_norm": 0.8087038993835449, "learning_rate": 6.290956749672346e-05, "loss": 1.2093, "step": 1152 }, { "epoch": 1.5003253090435913, "grad_norm": 0.7290136218070984, "learning_rate": 6.287680209698559e-05, "loss": 0.9374, "step": 1153 }, { "epoch": 1.501626545217957, "grad_norm": 0.7479668259620667, "learning_rate": 6.284403669724771e-05, "loss": 1.5438, "step": 1154 }, { "epoch": 1.5029277813923227, "grad_norm": 1.4550060033798218, "learning_rate": 6.281127129750983e-05, "loss": 1.2301, "step": 1155 }, { "epoch": 1.5042290175666884, "grad_norm": 1.5139431953430176, "learning_rate": 6.277850589777196e-05, "loss": 1.1235, "step": 1156 }, { "epoch": 1.505530253741054, "grad_norm": 1.2523235082626343, "learning_rate": 6.274574049803408e-05, "loss": 1.5546, "step": 1157 }, { "epoch": 1.5068314899154196, "grad_norm": 2.3312652111053467, "learning_rate": 6.27129750982962e-05, "loss": 1.7413, "step": 1158 }, { "epoch": 1.5081327260897854, "grad_norm": 1.4471575021743774, "learning_rate": 6.268020969855833e-05, "loss": 1.7892, "step": 1159 }, { "epoch": 1.509433962264151, "grad_norm": 0.8336760401725769, "learning_rate": 6.264744429882044e-05, "loss": 1.8203, "step": 1160 }, { "epoch": 1.5107351984385167, "grad_norm": 2.6003682613372803, "learning_rate": 6.261467889908257e-05, "loss": 2.0735, "step": 1161 }, { "epoch": 1.5120364346128823, "grad_norm": 0.8530160784721375, "learning_rate": 6.258191349934469e-05, "loss": 1.7878, "step": 1162 }, { "epoch": 1.513337670787248, "grad_norm": 0.7864294648170471, "learning_rate": 6.254914809960681e-05, "loss": 1.7607, "step": 1163 }, { "epoch": 1.5146389069616135, "grad_norm": 1.2927407026290894, "learning_rate": 6.251638269986894e-05, "loss": 1.3908, "step": 1164 }, { "epoch": 1.5159401431359791, "grad_norm": 0.950560986995697, "learning_rate": 6.248361730013107e-05, "loss": 1.5974, "step": 1165 }, { "epoch": 1.5172413793103448, "grad_norm": 0.6577100157737732, "learning_rate": 6.245085190039319e-05, "loss": 1.128, "step": 1166 }, { "epoch": 1.5185426154847104, "grad_norm": 0.7616334557533264, "learning_rate": 6.241808650065531e-05, "loss": 1.4575, "step": 1167 }, { "epoch": 1.519843851659076, "grad_norm": 0.7936611771583557, "learning_rate": 6.238532110091744e-05, "loss": 1.2547, "step": 1168 }, { "epoch": 1.5211450878334418, "grad_norm": 1.831970453262329, "learning_rate": 6.235255570117956e-05, "loss": 1.7771, "step": 1169 }, { "epoch": 1.5224463240078074, "grad_norm": 1.1793321371078491, "learning_rate": 6.231979030144168e-05, "loss": 1.5821, "step": 1170 }, { "epoch": 1.523747560182173, "grad_norm": 0.9257839918136597, "learning_rate": 6.228702490170381e-05, "loss": 1.7149, "step": 1171 }, { "epoch": 1.5250487963565387, "grad_norm": 0.8139029741287231, "learning_rate": 6.225425950196593e-05, "loss": 1.6983, "step": 1172 }, { "epoch": 1.5263500325309045, "grad_norm": 0.6282765865325928, "learning_rate": 6.222149410222804e-05, "loss": 1.3643, "step": 1173 }, { "epoch": 1.5276512687052701, "grad_norm": 0.8392549753189087, "learning_rate": 6.218872870249018e-05, "loss": 1.4455, "step": 1174 }, { "epoch": 1.5289525048796357, "grad_norm": 0.871213972568512, "learning_rate": 6.21559633027523e-05, "loss": 1.6502, "step": 1175 }, { "epoch": 1.5302537410540014, "grad_norm": 1.046051263809204, "learning_rate": 6.212319790301441e-05, "loss": 1.7282, "step": 1176 }, { "epoch": 1.531554977228367, "grad_norm": 0.8911481499671936, "learning_rate": 6.209043250327654e-05, "loss": 1.889, "step": 1177 }, { "epoch": 1.5328562134027326, "grad_norm": 0.872871994972229, "learning_rate": 6.205766710353866e-05, "loss": 1.749, "step": 1178 }, { "epoch": 1.5341574495770982, "grad_norm": 1.353089451789856, "learning_rate": 6.202490170380078e-05, "loss": 1.9169, "step": 1179 }, { "epoch": 1.5354586857514638, "grad_norm": 0.6171141862869263, "learning_rate": 6.199213630406291e-05, "loss": 1.8315, "step": 1180 }, { "epoch": 1.5367599219258294, "grad_norm": 1.0929052829742432, "learning_rate": 6.195937090432503e-05, "loss": 1.3904, "step": 1181 }, { "epoch": 1.538061158100195, "grad_norm": 1.0012127161026, "learning_rate": 6.192660550458716e-05, "loss": 1.034, "step": 1182 }, { "epoch": 1.5393623942745607, "grad_norm": 0.6943049430847168, "learning_rate": 6.189384010484928e-05, "loss": 1.5894, "step": 1183 }, { "epoch": 1.5406636304489265, "grad_norm": 0.9747276902198792, "learning_rate": 6.186107470511141e-05, "loss": 1.7628, "step": 1184 }, { "epoch": 1.5419648666232921, "grad_norm": 0.8279460668563843, "learning_rate": 6.182830930537353e-05, "loss": 1.2019, "step": 1185 }, { "epoch": 1.5432661027976577, "grad_norm": 1.3655004501342773, "learning_rate": 6.179554390563566e-05, "loss": 1.35, "step": 1186 }, { "epoch": 1.5445673389720236, "grad_norm": 0.8635941743850708, "learning_rate": 6.176277850589778e-05, "loss": 1.2447, "step": 1187 }, { "epoch": 1.5458685751463892, "grad_norm": 0.7118312120437622, "learning_rate": 6.17300131061599e-05, "loss": 1.9323, "step": 1188 }, { "epoch": 1.5471698113207548, "grad_norm": 0.6626870632171631, "learning_rate": 6.169724770642203e-05, "loss": 1.4793, "step": 1189 }, { "epoch": 1.5484710474951204, "grad_norm": 1.6495893001556396, "learning_rate": 6.166448230668414e-05, "loss": 1.5694, "step": 1190 }, { "epoch": 1.549772283669486, "grad_norm": 1.768880844116211, "learning_rate": 6.163171690694626e-05, "loss": 1.7109, "step": 1191 }, { "epoch": 1.5510735198438517, "grad_norm": 0.8152806162834167, "learning_rate": 6.15989515072084e-05, "loss": 1.1047, "step": 1192 }, { "epoch": 1.5523747560182173, "grad_norm": 1.031057357788086, "learning_rate": 6.156618610747051e-05, "loss": 1.492, "step": 1193 }, { "epoch": 1.553675992192583, "grad_norm": 0.7994878888130188, "learning_rate": 6.153342070773263e-05, "loss": 1.3982, "step": 1194 }, { "epoch": 1.5549772283669485, "grad_norm": 1.0751738548278809, "learning_rate": 6.150065530799476e-05, "loss": 1.6716, "step": 1195 }, { "epoch": 1.5562784645413141, "grad_norm": 0.8919134140014648, "learning_rate": 6.146788990825688e-05, "loss": 1.959, "step": 1196 }, { "epoch": 1.5575797007156797, "grad_norm": 0.8019439578056335, "learning_rate": 6.1435124508519e-05, "loss": 1.7191, "step": 1197 }, { "epoch": 1.5588809368900456, "grad_norm": 0.9781663417816162, "learning_rate": 6.140235910878113e-05, "loss": 1.2862, "step": 1198 }, { "epoch": 1.5601821730644112, "grad_norm": 0.8487160205841064, "learning_rate": 6.136959370904325e-05, "loss": 1.6679, "step": 1199 }, { "epoch": 1.5614834092387768, "grad_norm": 1.0659089088439941, "learning_rate": 6.133682830930538e-05, "loss": 1.7828, "step": 1200 }, { "epoch": 1.5627846454131424, "grad_norm": 0.5615077018737793, "learning_rate": 6.13040629095675e-05, "loss": 1.6567, "step": 1201 }, { "epoch": 1.5640858815875083, "grad_norm": 1.119652509689331, "learning_rate": 6.127129750982963e-05, "loss": 1.3214, "step": 1202 }, { "epoch": 1.565387117761874, "grad_norm": 0.856385350227356, "learning_rate": 6.123853211009175e-05, "loss": 1.9286, "step": 1203 }, { "epoch": 1.5666883539362395, "grad_norm": 1.092742681503296, "learning_rate": 6.120576671035388e-05, "loss": 1.4958, "step": 1204 }, { "epoch": 1.5679895901106051, "grad_norm": 1.1529244184494019, "learning_rate": 6.1173001310616e-05, "loss": 0.936, "step": 1205 }, { "epoch": 1.5692908262849707, "grad_norm": 1.6845803260803223, "learning_rate": 6.114023591087811e-05, "loss": 1.3533, "step": 1206 }, { "epoch": 1.5705920624593364, "grad_norm": 1.0080698728561401, "learning_rate": 6.110747051114025e-05, "loss": 1.3822, "step": 1207 }, { "epoch": 1.571893298633702, "grad_norm": 0.6692062616348267, "learning_rate": 6.107470511140236e-05, "loss": 1.483, "step": 1208 }, { "epoch": 1.5731945348080676, "grad_norm": 0.8528201580047607, "learning_rate": 6.104193971166448e-05, "loss": 1.6514, "step": 1209 }, { "epoch": 1.5744957709824332, "grad_norm": 0.8522144556045532, "learning_rate": 6.1009174311926606e-05, "loss": 1.5596, "step": 1210 }, { "epoch": 1.5757970071567988, "grad_norm": 1.7467012405395508, "learning_rate": 6.097640891218873e-05, "loss": 1.4239, "step": 1211 }, { "epoch": 1.5770982433311644, "grad_norm": 1.2707619667053223, "learning_rate": 6.0943643512450855e-05, "loss": 1.3344, "step": 1212 }, { "epoch": 1.5783994795055303, "grad_norm": 1.0228770971298218, "learning_rate": 6.091087811271298e-05, "loss": 1.516, "step": 1213 }, { "epoch": 1.579700715679896, "grad_norm": 1.007389783859253, "learning_rate": 6.08781127129751e-05, "loss": 1.6075, "step": 1214 }, { "epoch": 1.5810019518542615, "grad_norm": 2.219559669494629, "learning_rate": 6.084534731323722e-05, "loss": 1.4189, "step": 1215 }, { "epoch": 1.5823031880286273, "grad_norm": 0.8490458130836487, "learning_rate": 6.081258191349935e-05, "loss": 2.0421, "step": 1216 }, { "epoch": 1.583604424202993, "grad_norm": 0.8757844567298889, "learning_rate": 6.0779816513761465e-05, "loss": 1.1222, "step": 1217 }, { "epoch": 1.5849056603773586, "grad_norm": 0.5634575486183167, "learning_rate": 6.074705111402359e-05, "loss": 2.1876, "step": 1218 }, { "epoch": 1.5862068965517242, "grad_norm": 0.9601777195930481, "learning_rate": 6.0714285714285715e-05, "loss": 1.3114, "step": 1219 }, { "epoch": 1.5875081327260898, "grad_norm": 1.8756211996078491, "learning_rate": 6.068152031454783e-05, "loss": 1.5854, "step": 1220 }, { "epoch": 1.5888093689004554, "grad_norm": 0.9435736536979675, "learning_rate": 6.064875491480997e-05, "loss": 1.7522, "step": 1221 }, { "epoch": 1.590110605074821, "grad_norm": 1.1617294549942017, "learning_rate": 6.061598951507209e-05, "loss": 1.6143, "step": 1222 }, { "epoch": 1.5914118412491867, "grad_norm": 0.8611693382263184, "learning_rate": 6.0583224115334214e-05, "loss": 1.3279, "step": 1223 }, { "epoch": 1.5927130774235523, "grad_norm": 1.3238295316696167, "learning_rate": 6.055045871559634e-05, "loss": 1.1199, "step": 1224 }, { "epoch": 1.594014313597918, "grad_norm": 0.9723123908042908, "learning_rate": 6.0517693315858456e-05, "loss": 1.531, "step": 1225 }, { "epoch": 1.5953155497722835, "grad_norm": 0.9327991604804993, "learning_rate": 6.048492791612058e-05, "loss": 2.2747, "step": 1226 }, { "epoch": 1.5966167859466494, "grad_norm": 1.1837098598480225, "learning_rate": 6.0452162516382706e-05, "loss": 1.7511, "step": 1227 }, { "epoch": 1.597918022121015, "grad_norm": 1.7035322189331055, "learning_rate": 6.0419397116644824e-05, "loss": 1.058, "step": 1228 }, { "epoch": 1.5992192582953806, "grad_norm": 1.612220048904419, "learning_rate": 6.038663171690695e-05, "loss": 1.5569, "step": 1229 }, { "epoch": 1.6005204944697464, "grad_norm": 1.9672068357467651, "learning_rate": 6.035386631716907e-05, "loss": 1.612, "step": 1230 }, { "epoch": 1.601821730644112, "grad_norm": 2.0985801219940186, "learning_rate": 6.03211009174312e-05, "loss": 1.3146, "step": 1231 }, { "epoch": 1.6031229668184777, "grad_norm": 1.588752031326294, "learning_rate": 6.0288335517693316e-05, "loss": 1.8115, "step": 1232 }, { "epoch": 1.6044242029928433, "grad_norm": 0.8232167959213257, "learning_rate": 6.025557011795544e-05, "loss": 1.7929, "step": 1233 }, { "epoch": 1.605725439167209, "grad_norm": 0.7996879816055298, "learning_rate": 6.0222804718217566e-05, "loss": 1.751, "step": 1234 }, { "epoch": 1.6070266753415745, "grad_norm": 0.9442285895347595, "learning_rate": 6.0190039318479683e-05, "loss": 1.6543, "step": 1235 }, { "epoch": 1.6083279115159401, "grad_norm": 0.8134985566139221, "learning_rate": 6.015727391874181e-05, "loss": 1.421, "step": 1236 }, { "epoch": 1.6096291476903057, "grad_norm": 0.8941827416419983, "learning_rate": 6.012450851900393e-05, "loss": 1.7549, "step": 1237 }, { "epoch": 1.6109303838646714, "grad_norm": 0.9611232876777649, "learning_rate": 6.009174311926605e-05, "loss": 1.6263, "step": 1238 }, { "epoch": 1.612231620039037, "grad_norm": 0.9388418197631836, "learning_rate": 6.005897771952819e-05, "loss": 1.831, "step": 1239 }, { "epoch": 1.6135328562134026, "grad_norm": 1.1481419801712036, "learning_rate": 6.002621231979031e-05, "loss": 1.2717, "step": 1240 }, { "epoch": 1.6148340923877684, "grad_norm": 0.8942838311195374, "learning_rate": 5.999344692005243e-05, "loss": 1.6085, "step": 1241 }, { "epoch": 1.616135328562134, "grad_norm": 0.9600407481193542, "learning_rate": 5.996068152031456e-05, "loss": 1.8099, "step": 1242 }, { "epoch": 1.6174365647364997, "grad_norm": 1.3834632635116577, "learning_rate": 5.9927916120576675e-05, "loss": 1.9349, "step": 1243 }, { "epoch": 1.6187378009108653, "grad_norm": 0.8740050196647644, "learning_rate": 5.98951507208388e-05, "loss": 1.6077, "step": 1244 }, { "epoch": 1.6200390370852311, "grad_norm": 0.8861376643180847, "learning_rate": 5.9862385321100924e-05, "loss": 1.4232, "step": 1245 }, { "epoch": 1.6213402732595967, "grad_norm": 2.1153194904327393, "learning_rate": 5.982961992136304e-05, "loss": 1.4794, "step": 1246 }, { "epoch": 1.6226415094339623, "grad_norm": 1.0009846687316895, "learning_rate": 5.979685452162517e-05, "loss": 1.6769, "step": 1247 }, { "epoch": 1.623942745608328, "grad_norm": 0.7223239541053772, "learning_rate": 5.976408912188729e-05, "loss": 1.6734, "step": 1248 }, { "epoch": 1.6252439817826936, "grad_norm": 1.2335052490234375, "learning_rate": 5.973132372214941e-05, "loss": 1.4342, "step": 1249 }, { "epoch": 1.6265452179570592, "grad_norm": 0.6680229306221008, "learning_rate": 5.9698558322411534e-05, "loss": 1.3768, "step": 1250 }, { "epoch": 1.6278464541314248, "grad_norm": 0.9401698708534241, "learning_rate": 5.966579292267366e-05, "loss": 1.5336, "step": 1251 }, { "epoch": 1.6291476903057904, "grad_norm": 1.2229256629943848, "learning_rate": 5.9633027522935784e-05, "loss": 1.6277, "step": 1252 }, { "epoch": 1.630448926480156, "grad_norm": 0.7878890633583069, "learning_rate": 5.96002621231979e-05, "loss": 1.9739, "step": 1253 }, { "epoch": 1.6317501626545217, "grad_norm": 0.9758756160736084, "learning_rate": 5.9567496723460026e-05, "loss": 1.5312, "step": 1254 }, { "epoch": 1.6330513988288873, "grad_norm": 0.9070915579795837, "learning_rate": 5.953473132372215e-05, "loss": 1.6084, "step": 1255 }, { "epoch": 1.6343526350032531, "grad_norm": 0.7687110304832458, "learning_rate": 5.950196592398427e-05, "loss": 1.8252, "step": 1256 }, { "epoch": 1.6356538711776187, "grad_norm": 0.8874659538269043, "learning_rate": 5.9469200524246394e-05, "loss": 1.6395, "step": 1257 }, { "epoch": 1.6369551073519844, "grad_norm": 0.6303523778915405, "learning_rate": 5.9436435124508525e-05, "loss": 1.506, "step": 1258 }, { "epoch": 1.6382563435263502, "grad_norm": 0.9561724066734314, "learning_rate": 5.940366972477065e-05, "loss": 1.4347, "step": 1259 }, { "epoch": 1.6395575797007158, "grad_norm": 0.7272549867630005, "learning_rate": 5.9370904325032775e-05, "loss": 1.9043, "step": 1260 }, { "epoch": 1.6408588158750814, "grad_norm": 0.7584100961685181, "learning_rate": 5.933813892529489e-05, "loss": 1.5032, "step": 1261 }, { "epoch": 1.642160052049447, "grad_norm": 0.9081423878669739, "learning_rate": 5.930537352555702e-05, "loss": 1.8106, "step": 1262 }, { "epoch": 1.6434612882238127, "grad_norm": 0.7488301396369934, "learning_rate": 5.927260812581914e-05, "loss": 1.8565, "step": 1263 }, { "epoch": 1.6447625243981783, "grad_norm": 0.816564679145813, "learning_rate": 5.923984272608126e-05, "loss": 1.8307, "step": 1264 }, { "epoch": 1.6460637605725439, "grad_norm": 0.9332618117332458, "learning_rate": 5.9207077326343385e-05, "loss": 1.6485, "step": 1265 }, { "epoch": 1.6473649967469095, "grad_norm": 0.920454204082489, "learning_rate": 5.917431192660551e-05, "loss": 1.9373, "step": 1266 }, { "epoch": 1.6486662329212751, "grad_norm": 1.034496545791626, "learning_rate": 5.914154652686763e-05, "loss": 1.6889, "step": 1267 }, { "epoch": 1.6499674690956407, "grad_norm": 0.9994304180145264, "learning_rate": 5.910878112712975e-05, "loss": 1.3893, "step": 1268 }, { "epoch": 1.6512687052700064, "grad_norm": 0.9362581372261047, "learning_rate": 5.907601572739188e-05, "loss": 1.5984, "step": 1269 }, { "epoch": 1.6525699414443722, "grad_norm": 1.0308774709701538, "learning_rate": 5.9043250327653995e-05, "loss": 1.7971, "step": 1270 }, { "epoch": 1.6538711776187378, "grad_norm": 0.8662400841712952, "learning_rate": 5.901048492791612e-05, "loss": 1.958, "step": 1271 }, { "epoch": 1.6551724137931034, "grad_norm": 0.6282533407211304, "learning_rate": 5.8977719528178245e-05, "loss": 1.1665, "step": 1272 }, { "epoch": 1.656473649967469, "grad_norm": 0.7988092303276062, "learning_rate": 5.894495412844037e-05, "loss": 0.9046, "step": 1273 }, { "epoch": 1.6577748861418349, "grad_norm": 0.9004459977149963, "learning_rate": 5.891218872870249e-05, "loss": 1.385, "step": 1274 }, { "epoch": 1.6590761223162005, "grad_norm": 1.0633572340011597, "learning_rate": 5.887942332896461e-05, "loss": 1.9659, "step": 1275 }, { "epoch": 1.6603773584905661, "grad_norm": 0.8603759407997131, "learning_rate": 5.884665792922674e-05, "loss": 1.8093, "step": 1276 }, { "epoch": 1.6616785946649317, "grad_norm": 0.7988016605377197, "learning_rate": 5.881389252948887e-05, "loss": 1.2606, "step": 1277 }, { "epoch": 1.6629798308392973, "grad_norm": 0.7725502848625183, "learning_rate": 5.878112712975099e-05, "loss": 2.0869, "step": 1278 }, { "epoch": 1.664281067013663, "grad_norm": 1.182335376739502, "learning_rate": 5.874836173001311e-05, "loss": 1.9087, "step": 1279 }, { "epoch": 1.6655823031880286, "grad_norm": 0.7543607354164124, "learning_rate": 5.8715596330275236e-05, "loss": 1.704, "step": 1280 }, { "epoch": 1.6668835393623942, "grad_norm": 0.7613018751144409, "learning_rate": 5.868283093053736e-05, "loss": 1.4808, "step": 1281 }, { "epoch": 1.6681847755367598, "grad_norm": 0.9603800773620605, "learning_rate": 5.865006553079948e-05, "loss": 1.748, "step": 1282 }, { "epoch": 1.6694860117111254, "grad_norm": 0.709215521812439, "learning_rate": 5.86173001310616e-05, "loss": 1.8837, "step": 1283 }, { "epoch": 1.6707872478854913, "grad_norm": 1.9300236701965332, "learning_rate": 5.858453473132373e-05, "loss": 1.1281, "step": 1284 }, { "epoch": 1.6720884840598569, "grad_norm": 0.9000341296195984, "learning_rate": 5.8551769331585846e-05, "loss": 1.5188, "step": 1285 }, { "epoch": 1.6733897202342225, "grad_norm": 1.2377146482467651, "learning_rate": 5.851900393184797e-05, "loss": 1.6957, "step": 1286 }, { "epoch": 1.6746909564085881, "grad_norm": 1.0615878105163574, "learning_rate": 5.8486238532110095e-05, "loss": 1.7395, "step": 1287 }, { "epoch": 1.675992192582954, "grad_norm": 1.0639220476150513, "learning_rate": 5.845347313237221e-05, "loss": 1.7231, "step": 1288 }, { "epoch": 1.6772934287573196, "grad_norm": 1.5860646963119507, "learning_rate": 5.842070773263434e-05, "loss": 1.3009, "step": 1289 }, { "epoch": 1.6785946649316852, "grad_norm": 1.1232824325561523, "learning_rate": 5.838794233289646e-05, "loss": 1.7223, "step": 1290 }, { "epoch": 1.6798959011060508, "grad_norm": 1.4127318859100342, "learning_rate": 5.835517693315858e-05, "loss": 1.5521, "step": 1291 }, { "epoch": 1.6811971372804164, "grad_norm": 0.7141563296318054, "learning_rate": 5.8322411533420705e-05, "loss": 1.6007, "step": 1292 }, { "epoch": 1.682498373454782, "grad_norm": 1.6581467390060425, "learning_rate": 5.828964613368283e-05, "loss": 1.6661, "step": 1293 }, { "epoch": 1.6837996096291477, "grad_norm": 1.11501944065094, "learning_rate": 5.8256880733944955e-05, "loss": 1.6775, "step": 1294 }, { "epoch": 1.6851008458035133, "grad_norm": 1.3256689310073853, "learning_rate": 5.8224115334207086e-05, "loss": 1.8439, "step": 1295 }, { "epoch": 1.6864020819778789, "grad_norm": 0.9576137065887451, "learning_rate": 5.819134993446921e-05, "loss": 1.6261, "step": 1296 }, { "epoch": 1.6877033181522445, "grad_norm": 1.0834840536117554, "learning_rate": 5.815858453473133e-05, "loss": 1.4344, "step": 1297 }, { "epoch": 1.6890045543266101, "grad_norm": 1.2587910890579224, "learning_rate": 5.8125819134993454e-05, "loss": 1.5148, "step": 1298 }, { "epoch": 1.690305790500976, "grad_norm": 0.7421240210533142, "learning_rate": 5.809305373525558e-05, "loss": 1.4023, "step": 1299 }, { "epoch": 1.6916070266753416, "grad_norm": 1.422673225402832, "learning_rate": 5.8060288335517697e-05, "loss": 1.2466, "step": 1300 }, { "epoch": 1.6929082628497072, "grad_norm": 1.1174432039260864, "learning_rate": 5.802752293577982e-05, "loss": 1.6694, "step": 1301 }, { "epoch": 1.694209499024073, "grad_norm": 0.6614477634429932, "learning_rate": 5.7994757536041946e-05, "loss": 1.5658, "step": 1302 }, { "epoch": 1.6955107351984386, "grad_norm": 0.7686635851860046, "learning_rate": 5.7961992136304064e-05, "loss": 1.8286, "step": 1303 }, { "epoch": 1.6968119713728043, "grad_norm": 0.7135832905769348, "learning_rate": 5.792922673656619e-05, "loss": 1.948, "step": 1304 }, { "epoch": 1.6981132075471699, "grad_norm": 1.2493455410003662, "learning_rate": 5.7896461336828313e-05, "loss": 1.7434, "step": 1305 }, { "epoch": 1.6994144437215355, "grad_norm": 0.7642462849617004, "learning_rate": 5.786369593709043e-05, "loss": 1.8674, "step": 1306 }, { "epoch": 1.7007156798959011, "grad_norm": 1.4130662679672241, "learning_rate": 5.7830930537352556e-05, "loss": 1.289, "step": 1307 }, { "epoch": 1.7020169160702667, "grad_norm": 1.069481372833252, "learning_rate": 5.779816513761468e-05, "loss": 1.3902, "step": 1308 }, { "epoch": 1.7033181522446323, "grad_norm": 0.8745287656784058, "learning_rate": 5.77653997378768e-05, "loss": 1.5818, "step": 1309 }, { "epoch": 1.704619388418998, "grad_norm": 1.4404197931289673, "learning_rate": 5.7732634338138924e-05, "loss": 1.7842, "step": 1310 }, { "epoch": 1.7059206245933636, "grad_norm": 1.1287171840667725, "learning_rate": 5.769986893840105e-05, "loss": 2.0195, "step": 1311 }, { "epoch": 1.7072218607677292, "grad_norm": 0.8815336227416992, "learning_rate": 5.766710353866317e-05, "loss": 1.652, "step": 1312 }, { "epoch": 1.708523096942095, "grad_norm": 0.7912148833274841, "learning_rate": 5.763433813892529e-05, "loss": 1.5972, "step": 1313 }, { "epoch": 1.7098243331164606, "grad_norm": 0.8649274706840515, "learning_rate": 5.760157273918743e-05, "loss": 1.6156, "step": 1314 }, { "epoch": 1.7111255692908263, "grad_norm": 0.8079018592834473, "learning_rate": 5.756880733944955e-05, "loss": 1.813, "step": 1315 }, { "epoch": 1.7124268054651919, "grad_norm": 0.8285290598869324, "learning_rate": 5.753604193971167e-05, "loss": 1.5997, "step": 1316 }, { "epoch": 1.7137280416395577, "grad_norm": 1.4027760028839111, "learning_rate": 5.75032765399738e-05, "loss": 1.3977, "step": 1317 }, { "epoch": 1.7150292778139233, "grad_norm": 1.06057608127594, "learning_rate": 5.7470511140235915e-05, "loss": 1.5532, "step": 1318 }, { "epoch": 1.716330513988289, "grad_norm": 2.0062673091888428, "learning_rate": 5.743774574049804e-05, "loss": 0.7484, "step": 1319 }, { "epoch": 1.7176317501626546, "grad_norm": 1.4249314069747925, "learning_rate": 5.7404980340760164e-05, "loss": 1.4156, "step": 1320 }, { "epoch": 1.7189329863370202, "grad_norm": 1.2333252429962158, "learning_rate": 5.737221494102228e-05, "loss": 0.6668, "step": 1321 }, { "epoch": 1.7202342225113858, "grad_norm": 0.9967629313468933, "learning_rate": 5.733944954128441e-05, "loss": 1.778, "step": 1322 }, { "epoch": 1.7215354586857514, "grad_norm": 0.8972412943840027, "learning_rate": 5.730668414154653e-05, "loss": 1.7958, "step": 1323 }, { "epoch": 1.722836694860117, "grad_norm": 1.312430739402771, "learning_rate": 5.727391874180865e-05, "loss": 1.1825, "step": 1324 }, { "epoch": 1.7241379310344827, "grad_norm": 0.7032502889633179, "learning_rate": 5.7241153342070774e-05, "loss": 1.1172, "step": 1325 }, { "epoch": 1.7254391672088483, "grad_norm": 0.7538443207740784, "learning_rate": 5.72083879423329e-05, "loss": 1.7655, "step": 1326 }, { "epoch": 1.726740403383214, "grad_norm": 0.8027899861335754, "learning_rate": 5.717562254259502e-05, "loss": 1.7376, "step": 1327 }, { "epoch": 1.7280416395575797, "grad_norm": 0.8091989755630493, "learning_rate": 5.714285714285714e-05, "loss": 1.5911, "step": 1328 }, { "epoch": 1.7293428757319453, "grad_norm": 0.789446234703064, "learning_rate": 5.7110091743119266e-05, "loss": 2.0121, "step": 1329 }, { "epoch": 1.730644111906311, "grad_norm": 0.7457113265991211, "learning_rate": 5.7077326343381384e-05, "loss": 1.4448, "step": 1330 }, { "epoch": 1.7319453480806768, "grad_norm": 0.7524446845054626, "learning_rate": 5.704456094364351e-05, "loss": 1.3837, "step": 1331 }, { "epoch": 1.7332465842550424, "grad_norm": 0.6823481321334839, "learning_rate": 5.701179554390564e-05, "loss": 1.7144, "step": 1332 }, { "epoch": 1.734547820429408, "grad_norm": 0.9352553486824036, "learning_rate": 5.6979030144167765e-05, "loss": 1.9177, "step": 1333 }, { "epoch": 1.7358490566037736, "grad_norm": 0.8687691688537598, "learning_rate": 5.694626474442989e-05, "loss": 1.7277, "step": 1334 }, { "epoch": 1.7371502927781393, "grad_norm": 0.907910168170929, "learning_rate": 5.6913499344692015e-05, "loss": 1.614, "step": 1335 }, { "epoch": 1.7384515289525049, "grad_norm": 0.7125810384750366, "learning_rate": 5.688073394495413e-05, "loss": 1.2761, "step": 1336 }, { "epoch": 1.7397527651268705, "grad_norm": 1.5011221170425415, "learning_rate": 5.684796854521626e-05, "loss": 1.386, "step": 1337 }, { "epoch": 1.741054001301236, "grad_norm": 0.8201001882553101, "learning_rate": 5.681520314547838e-05, "loss": 1.526, "step": 1338 }, { "epoch": 1.7423552374756017, "grad_norm": 0.9148080945014954, "learning_rate": 5.67824377457405e-05, "loss": 1.4143, "step": 1339 }, { "epoch": 1.7436564736499673, "grad_norm": 0.8619223833084106, "learning_rate": 5.6749672346002625e-05, "loss": 1.4509, "step": 1340 }, { "epoch": 1.744957709824333, "grad_norm": 1.244593620300293, "learning_rate": 5.671690694626475e-05, "loss": 1.8314, "step": 1341 }, { "epoch": 1.7462589459986988, "grad_norm": 0.8283818960189819, "learning_rate": 5.668414154652687e-05, "loss": 1.7718, "step": 1342 }, { "epoch": 1.7475601821730644, "grad_norm": 0.87214595079422, "learning_rate": 5.665137614678899e-05, "loss": 1.7313, "step": 1343 }, { "epoch": 1.74886141834743, "grad_norm": 0.8282403945922852, "learning_rate": 5.661861074705112e-05, "loss": 1.0677, "step": 1344 }, { "epoch": 1.7501626545217959, "grad_norm": 0.6050186157226562, "learning_rate": 5.6585845347313235e-05, "loss": 1.8923, "step": 1345 }, { "epoch": 1.7514638906961615, "grad_norm": 2.5374372005462646, "learning_rate": 5.655307994757536e-05, "loss": 1.4676, "step": 1346 }, { "epoch": 1.752765126870527, "grad_norm": 0.9410498738288879, "learning_rate": 5.6520314547837485e-05, "loss": 1.5792, "step": 1347 }, { "epoch": 1.7540663630448927, "grad_norm": 0.7076353430747986, "learning_rate": 5.64875491480996e-05, "loss": 1.6558, "step": 1348 }, { "epoch": 1.7553675992192583, "grad_norm": 1.1652634143829346, "learning_rate": 5.645478374836173e-05, "loss": 1.7442, "step": 1349 }, { "epoch": 1.756668835393624, "grad_norm": 1.0979132652282715, "learning_rate": 5.642201834862385e-05, "loss": 1.4177, "step": 1350 }, { "epoch": 1.7579700715679896, "grad_norm": 0.6790662407875061, "learning_rate": 5.6389252948885984e-05, "loss": 1.8792, "step": 1351 }, { "epoch": 1.7592713077423552, "grad_norm": 0.7219562530517578, "learning_rate": 5.635648754914811e-05, "loss": 1.7211, "step": 1352 }, { "epoch": 1.7605725439167208, "grad_norm": 1.6354771852493286, "learning_rate": 5.6323722149410226e-05, "loss": 1.0611, "step": 1353 }, { "epoch": 1.7618737800910864, "grad_norm": 1.0574191808700562, "learning_rate": 5.629095674967235e-05, "loss": 1.3256, "step": 1354 }, { "epoch": 1.763175016265452, "grad_norm": 0.9465879201889038, "learning_rate": 5.6258191349934476e-05, "loss": 1.739, "step": 1355 }, { "epoch": 1.7644762524398179, "grad_norm": 1.0286582708358765, "learning_rate": 5.62254259501966e-05, "loss": 1.6666, "step": 1356 }, { "epoch": 1.7657774886141835, "grad_norm": 0.8684209585189819, "learning_rate": 5.619266055045872e-05, "loss": 1.3912, "step": 1357 }, { "epoch": 1.767078724788549, "grad_norm": 1.2097471952438354, "learning_rate": 5.615989515072084e-05, "loss": 1.5971, "step": 1358 }, { "epoch": 1.7683799609629147, "grad_norm": 0.8580326437950134, "learning_rate": 5.612712975098297e-05, "loss": 1.5281, "step": 1359 }, { "epoch": 1.7696811971372806, "grad_norm": 3.1507327556610107, "learning_rate": 5.6094364351245086e-05, "loss": 1.9234, "step": 1360 }, { "epoch": 1.7709824333116462, "grad_norm": 0.8449283242225647, "learning_rate": 5.606159895150721e-05, "loss": 1.6971, "step": 1361 }, { "epoch": 1.7722836694860118, "grad_norm": 0.7687345743179321, "learning_rate": 5.6028833551769335e-05, "loss": 1.7973, "step": 1362 }, { "epoch": 1.7735849056603774, "grad_norm": 1.121598482131958, "learning_rate": 5.599606815203145e-05, "loss": 1.4383, "step": 1363 }, { "epoch": 1.774886141834743, "grad_norm": 0.7450833320617676, "learning_rate": 5.596330275229358e-05, "loss": 1.7668, "step": 1364 }, { "epoch": 1.7761873780091086, "grad_norm": 1.326856255531311, "learning_rate": 5.59305373525557e-05, "loss": 1.5553, "step": 1365 }, { "epoch": 1.7774886141834743, "grad_norm": 1.0393309593200684, "learning_rate": 5.589777195281782e-05, "loss": 0.9825, "step": 1366 }, { "epoch": 1.7787898503578399, "grad_norm": 1.1575599908828735, "learning_rate": 5.5865006553079945e-05, "loss": 1.7818, "step": 1367 }, { "epoch": 1.7800910865322055, "grad_norm": 1.105177640914917, "learning_rate": 5.583224115334207e-05, "loss": 1.6526, "step": 1368 }, { "epoch": 1.781392322706571, "grad_norm": 0.9981535077095032, "learning_rate": 5.579947575360419e-05, "loss": 1.3065, "step": 1369 }, { "epoch": 1.7826935588809367, "grad_norm": 1.126455545425415, "learning_rate": 5.5766710353866326e-05, "loss": 1.8539, "step": 1370 }, { "epoch": 1.7839947950553026, "grad_norm": 0.6846060752868652, "learning_rate": 5.5733944954128444e-05, "loss": 1.302, "step": 1371 }, { "epoch": 1.7852960312296682, "grad_norm": 0.7944363355636597, "learning_rate": 5.570117955439057e-05, "loss": 1.2462, "step": 1372 }, { "epoch": 1.7865972674040338, "grad_norm": 0.9040448069572449, "learning_rate": 5.5668414154652694e-05, "loss": 1.4224, "step": 1373 }, { "epoch": 1.7878985035783996, "grad_norm": 0.7150500416755676, "learning_rate": 5.563564875491482e-05, "loss": 1.8481, "step": 1374 }, { "epoch": 1.7891997397527653, "grad_norm": 0.7792714238166809, "learning_rate": 5.5602883355176937e-05, "loss": 2.0927, "step": 1375 }, { "epoch": 1.7905009759271309, "grad_norm": 1.6151827573776245, "learning_rate": 5.557011795543906e-05, "loss": 1.4005, "step": 1376 }, { "epoch": 1.7918022121014965, "grad_norm": 0.7339690923690796, "learning_rate": 5.5537352555701186e-05, "loss": 2.1056, "step": 1377 }, { "epoch": 1.793103448275862, "grad_norm": 0.9656336307525635, "learning_rate": 5.5504587155963304e-05, "loss": 1.4301, "step": 1378 }, { "epoch": 1.7944046844502277, "grad_norm": 1.057574987411499, "learning_rate": 5.547182175622543e-05, "loss": 1.6986, "step": 1379 }, { "epoch": 1.7957059206245933, "grad_norm": 0.7473500967025757, "learning_rate": 5.5439056356487554e-05, "loss": 1.2338, "step": 1380 }, { "epoch": 1.797007156798959, "grad_norm": 0.7734534740447998, "learning_rate": 5.540629095674967e-05, "loss": 1.7592, "step": 1381 }, { "epoch": 1.7983083929733246, "grad_norm": 0.7359315752983093, "learning_rate": 5.5373525557011796e-05, "loss": 1.614, "step": 1382 }, { "epoch": 1.7996096291476902, "grad_norm": 0.9160071611404419, "learning_rate": 5.534076015727392e-05, "loss": 1.3302, "step": 1383 }, { "epoch": 1.8009108653220558, "grad_norm": 1.1746207475662231, "learning_rate": 5.530799475753604e-05, "loss": 1.8886, "step": 1384 }, { "epoch": 1.8022121014964216, "grad_norm": 0.8206449151039124, "learning_rate": 5.5275229357798164e-05, "loss": 1.1001, "step": 1385 }, { "epoch": 1.8035133376707873, "grad_norm": 0.7496719360351562, "learning_rate": 5.524246395806029e-05, "loss": 1.4534, "step": 1386 }, { "epoch": 1.8048145738451529, "grad_norm": 0.8007135391235352, "learning_rate": 5.5209698558322406e-05, "loss": 1.8034, "step": 1387 }, { "epoch": 1.8061158100195187, "grad_norm": 1.0449694395065308, "learning_rate": 5.5176933158584545e-05, "loss": 1.2597, "step": 1388 }, { "epoch": 1.8074170461938843, "grad_norm": 0.8926443457603455, "learning_rate": 5.514416775884666e-05, "loss": 1.6355, "step": 1389 }, { "epoch": 1.80871828236825, "grad_norm": 1.0529966354370117, "learning_rate": 5.511140235910879e-05, "loss": 1.862, "step": 1390 }, { "epoch": 1.8100195185426156, "grad_norm": 0.9239261746406555, "learning_rate": 5.507863695937091e-05, "loss": 1.8862, "step": 1391 }, { "epoch": 1.8113207547169812, "grad_norm": 0.8153102993965149, "learning_rate": 5.504587155963303e-05, "loss": 1.5273, "step": 1392 }, { "epoch": 1.8126219908913468, "grad_norm": 0.8597419857978821, "learning_rate": 5.5013106159895155e-05, "loss": 2.1963, "step": 1393 }, { "epoch": 1.8139232270657124, "grad_norm": 1.1515299081802368, "learning_rate": 5.498034076015728e-05, "loss": 1.6666, "step": 1394 }, { "epoch": 1.815224463240078, "grad_norm": 0.7869179248809814, "learning_rate": 5.4947575360419404e-05, "loss": 1.7049, "step": 1395 }, { "epoch": 1.8165256994144436, "grad_norm": 1.0353889465332031, "learning_rate": 5.491480996068152e-05, "loss": 1.6176, "step": 1396 }, { "epoch": 1.8178269355888093, "grad_norm": 0.8661072850227356, "learning_rate": 5.488204456094365e-05, "loss": 1.8704, "step": 1397 }, { "epoch": 1.8191281717631749, "grad_norm": 0.8428536057472229, "learning_rate": 5.484927916120577e-05, "loss": 1.9338, "step": 1398 }, { "epoch": 1.8204294079375407, "grad_norm": 0.9810117483139038, "learning_rate": 5.481651376146789e-05, "loss": 2.0392, "step": 1399 }, { "epoch": 1.8217306441119063, "grad_norm": 0.8401241302490234, "learning_rate": 5.4783748361730014e-05, "loss": 1.79, "step": 1400 }, { "epoch": 1.823031880286272, "grad_norm": 0.8241257667541504, "learning_rate": 5.475098296199214e-05, "loss": 1.7933, "step": 1401 }, { "epoch": 1.8243331164606376, "grad_norm": 1.6130050420761108, "learning_rate": 5.471821756225426e-05, "loss": 1.3922, "step": 1402 }, { "epoch": 1.8256343526350034, "grad_norm": 0.8556849956512451, "learning_rate": 5.468545216251638e-05, "loss": 1.5977, "step": 1403 }, { "epoch": 1.826935588809369, "grad_norm": 0.8359464406967163, "learning_rate": 5.4652686762778507e-05, "loss": 1.7573, "step": 1404 }, { "epoch": 1.8282368249837346, "grad_norm": 0.8708614706993103, "learning_rate": 5.4619921363040625e-05, "loss": 1.8867, "step": 1405 }, { "epoch": 1.8295380611581002, "grad_norm": 0.8282567858695984, "learning_rate": 5.458715596330275e-05, "loss": 1.4282, "step": 1406 }, { "epoch": 1.8308392973324659, "grad_norm": 1.3901506662368774, "learning_rate": 5.455439056356488e-05, "loss": 0.9984, "step": 1407 }, { "epoch": 1.8321405335068315, "grad_norm": 0.8276293873786926, "learning_rate": 5.4521625163827006e-05, "loss": 1.9774, "step": 1408 }, { "epoch": 1.833441769681197, "grad_norm": 0.7125580310821533, "learning_rate": 5.448885976408913e-05, "loss": 1.749, "step": 1409 }, { "epoch": 1.8347430058555627, "grad_norm": 1.2422622442245483, "learning_rate": 5.445609436435125e-05, "loss": 1.4241, "step": 1410 }, { "epoch": 1.8360442420299283, "grad_norm": 1.5017950534820557, "learning_rate": 5.442332896461337e-05, "loss": 1.0211, "step": 1411 }, { "epoch": 1.837345478204294, "grad_norm": 0.9342747926712036, "learning_rate": 5.43905635648755e-05, "loss": 1.5704, "step": 1412 }, { "epoch": 1.8386467143786596, "grad_norm": 1.1894510984420776, "learning_rate": 5.4357798165137616e-05, "loss": 1.0449, "step": 1413 }, { "epoch": 1.8399479505530254, "grad_norm": 0.9351675510406494, "learning_rate": 5.432503276539974e-05, "loss": 1.5916, "step": 1414 }, { "epoch": 1.841249186727391, "grad_norm": 1.3590888977050781, "learning_rate": 5.4292267365661865e-05, "loss": 1.2592, "step": 1415 }, { "epoch": 1.8425504229017566, "grad_norm": 0.744843065738678, "learning_rate": 5.425950196592399e-05, "loss": 1.775, "step": 1416 }, { "epoch": 1.8438516590761225, "grad_norm": 0.6935780048370361, "learning_rate": 5.422673656618611e-05, "loss": 1.8251, "step": 1417 }, { "epoch": 1.845152895250488, "grad_norm": 0.7879618406295776, "learning_rate": 5.419397116644823e-05, "loss": 1.3497, "step": 1418 }, { "epoch": 1.8464541314248537, "grad_norm": 1.0574324131011963, "learning_rate": 5.416120576671036e-05, "loss": 1.4241, "step": 1419 }, { "epoch": 1.8477553675992193, "grad_norm": 1.0004535913467407, "learning_rate": 5.4128440366972475e-05, "loss": 1.3217, "step": 1420 }, { "epoch": 1.849056603773585, "grad_norm": 0.9466626048088074, "learning_rate": 5.40956749672346e-05, "loss": 1.6196, "step": 1421 }, { "epoch": 1.8503578399479506, "grad_norm": 1.2953803539276123, "learning_rate": 5.4062909567496725e-05, "loss": 2.0772, "step": 1422 }, { "epoch": 1.8516590761223162, "grad_norm": 0.9695630073547363, "learning_rate": 5.403014416775884e-05, "loss": 1.224, "step": 1423 }, { "epoch": 1.8529603122966818, "grad_norm": 0.7857170701026917, "learning_rate": 5.399737876802097e-05, "loss": 1.5073, "step": 1424 }, { "epoch": 1.8542615484710474, "grad_norm": 0.903986394405365, "learning_rate": 5.396461336828309e-05, "loss": 1.1997, "step": 1425 }, { "epoch": 1.855562784645413, "grad_norm": 0.7546539306640625, "learning_rate": 5.3931847968545224e-05, "loss": 1.8865, "step": 1426 }, { "epoch": 1.8568640208197786, "grad_norm": 0.9077516794204712, "learning_rate": 5.389908256880735e-05, "loss": 1.3596, "step": 1427 }, { "epoch": 1.8581652569941445, "grad_norm": 0.9269987940788269, "learning_rate": 5.3866317169069466e-05, "loss": 1.8118, "step": 1428 }, { "epoch": 1.85946649316851, "grad_norm": 0.8261504769325256, "learning_rate": 5.383355176933159e-05, "loss": 2.0948, "step": 1429 }, { "epoch": 1.8607677293428757, "grad_norm": 0.8071203231811523, "learning_rate": 5.3800786369593716e-05, "loss": 1.9964, "step": 1430 }, { "epoch": 1.8620689655172413, "grad_norm": 0.8667087554931641, "learning_rate": 5.3768020969855834e-05, "loss": 2.0657, "step": 1431 }, { "epoch": 1.8633702016916072, "grad_norm": 3.173273801803589, "learning_rate": 5.373525557011796e-05, "loss": 1.8571, "step": 1432 }, { "epoch": 1.8646714378659728, "grad_norm": 1.0166431665420532, "learning_rate": 5.370249017038008e-05, "loss": 1.604, "step": 1433 }, { "epoch": 1.8659726740403384, "grad_norm": 0.8658209443092346, "learning_rate": 5.36697247706422e-05, "loss": 0.9805, "step": 1434 }, { "epoch": 1.867273910214704, "grad_norm": 1.2935165166854858, "learning_rate": 5.3636959370904326e-05, "loss": 1.4211, "step": 1435 }, { "epoch": 1.8685751463890696, "grad_norm": 1.0045682191848755, "learning_rate": 5.360419397116645e-05, "loss": 1.6179, "step": 1436 }, { "epoch": 1.8698763825634352, "grad_norm": 1.090926170349121, "learning_rate": 5.3571428571428575e-05, "loss": 1.8983, "step": 1437 }, { "epoch": 1.8711776187378009, "grad_norm": 1.1800062656402588, "learning_rate": 5.3538663171690693e-05, "loss": 1.5327, "step": 1438 }, { "epoch": 1.8724788549121665, "grad_norm": 1.088571548461914, "learning_rate": 5.350589777195282e-05, "loss": 1.3919, "step": 1439 }, { "epoch": 1.873780091086532, "grad_norm": 0.8415693640708923, "learning_rate": 5.347313237221494e-05, "loss": 1.7088, "step": 1440 }, { "epoch": 1.8750813272608977, "grad_norm": 0.9621151685714722, "learning_rate": 5.344036697247706e-05, "loss": 1.6948, "step": 1441 }, { "epoch": 1.8763825634352636, "grad_norm": 0.847240686416626, "learning_rate": 5.3407601572739186e-05, "loss": 1.3693, "step": 1442 }, { "epoch": 1.8776837996096292, "grad_norm": 1.2606687545776367, "learning_rate": 5.337483617300131e-05, "loss": 1.4843, "step": 1443 }, { "epoch": 1.8789850357839948, "grad_norm": 1.54637610912323, "learning_rate": 5.334207077326344e-05, "loss": 1.1136, "step": 1444 }, { "epoch": 1.8802862719583604, "grad_norm": 0.7429446578025818, "learning_rate": 5.3309305373525567e-05, "loss": 1.6539, "step": 1445 }, { "epoch": 1.8815875081327262, "grad_norm": 0.6358936429023743, "learning_rate": 5.3276539973787685e-05, "loss": 2.2369, "step": 1446 }, { "epoch": 1.8828887443070919, "grad_norm": 0.8868154287338257, "learning_rate": 5.324377457404981e-05, "loss": 1.8179, "step": 1447 }, { "epoch": 1.8841899804814575, "grad_norm": 0.7209793925285339, "learning_rate": 5.3211009174311934e-05, "loss": 1.8777, "step": 1448 }, { "epoch": 1.885491216655823, "grad_norm": 1.3081629276275635, "learning_rate": 5.317824377457405e-05, "loss": 1.681, "step": 1449 }, { "epoch": 1.8867924528301887, "grad_norm": 0.7984788417816162, "learning_rate": 5.314547837483618e-05, "loss": 1.4036, "step": 1450 }, { "epoch": 1.8880936890045543, "grad_norm": 0.7381587624549866, "learning_rate": 5.31127129750983e-05, "loss": 1.2209, "step": 1451 }, { "epoch": 1.88939492517892, "grad_norm": 0.8740935921669006, "learning_rate": 5.307994757536042e-05, "loss": 1.5195, "step": 1452 }, { "epoch": 1.8906961613532856, "grad_norm": 0.7735045552253723, "learning_rate": 5.3047182175622544e-05, "loss": 1.9139, "step": 1453 }, { "epoch": 1.8919973975276512, "grad_norm": 0.9340476989746094, "learning_rate": 5.301441677588467e-05, "loss": 1.6611, "step": 1454 }, { "epoch": 1.8932986337020168, "grad_norm": 1.066355586051941, "learning_rate": 5.2981651376146794e-05, "loss": 1.7168, "step": 1455 }, { "epoch": 1.8945998698763824, "grad_norm": 0.9642564058303833, "learning_rate": 5.294888597640891e-05, "loss": 1.3523, "step": 1456 }, { "epoch": 1.8959011060507482, "grad_norm": 1.074836254119873, "learning_rate": 5.2916120576671036e-05, "loss": 1.4109, "step": 1457 }, { "epoch": 1.8972023422251139, "grad_norm": 1.0979022979736328, "learning_rate": 5.288335517693316e-05, "loss": 1.8037, "step": 1458 }, { "epoch": 1.8985035783994795, "grad_norm": 1.0604757070541382, "learning_rate": 5.285058977719528e-05, "loss": 1.605, "step": 1459 }, { "epoch": 1.8998048145738453, "grad_norm": 1.2102012634277344, "learning_rate": 5.2817824377457404e-05, "loss": 1.8226, "step": 1460 }, { "epoch": 1.901106050748211, "grad_norm": 0.857421875, "learning_rate": 5.278505897771953e-05, "loss": 1.9219, "step": 1461 }, { "epoch": 1.9024072869225765, "grad_norm": 0.5498221516609192, "learning_rate": 5.2752293577981646e-05, "loss": 0.9959, "step": 1462 }, { "epoch": 1.9037085230969422, "grad_norm": 0.762774646282196, "learning_rate": 5.2719528178243785e-05, "loss": 1.7322, "step": 1463 }, { "epoch": 1.9050097592713078, "grad_norm": 1.2149394750595093, "learning_rate": 5.26867627785059e-05, "loss": 1.6321, "step": 1464 }, { "epoch": 1.9063109954456734, "grad_norm": 0.7591372728347778, "learning_rate": 5.265399737876803e-05, "loss": 1.8645, "step": 1465 }, { "epoch": 1.907612231620039, "grad_norm": 0.912999153137207, "learning_rate": 5.262123197903015e-05, "loss": 1.6038, "step": 1466 }, { "epoch": 1.9089134677944046, "grad_norm": 1.2057464122772217, "learning_rate": 5.258846657929227e-05, "loss": 1.4657, "step": 1467 }, { "epoch": 1.9102147039687702, "grad_norm": 1.3563973903656006, "learning_rate": 5.2555701179554395e-05, "loss": 1.8751, "step": 1468 }, { "epoch": 1.9115159401431359, "grad_norm": 0.8025534152984619, "learning_rate": 5.252293577981652e-05, "loss": 1.9264, "step": 1469 }, { "epoch": 1.9128171763175015, "grad_norm": 1.8615543842315674, "learning_rate": 5.249017038007864e-05, "loss": 1.2595, "step": 1470 }, { "epoch": 1.9141184124918673, "grad_norm": 0.7627991437911987, "learning_rate": 5.245740498034076e-05, "loss": 1.8178, "step": 1471 }, { "epoch": 1.915419648666233, "grad_norm": 0.7815648913383484, "learning_rate": 5.242463958060289e-05, "loss": 1.7485, "step": 1472 }, { "epoch": 1.9167208848405985, "grad_norm": 0.7962321639060974, "learning_rate": 5.2391874180865005e-05, "loss": 1.6345, "step": 1473 }, { "epoch": 1.9180221210149642, "grad_norm": 1.4795897006988525, "learning_rate": 5.235910878112713e-05, "loss": 1.5985, "step": 1474 }, { "epoch": 1.91932335718933, "grad_norm": 1.5672683715820312, "learning_rate": 5.2326343381389254e-05, "loss": 1.9985, "step": 1475 }, { "epoch": 1.9206245933636956, "grad_norm": 0.900195837020874, "learning_rate": 5.229357798165138e-05, "loss": 1.4742, "step": 1476 }, { "epoch": 1.9219258295380612, "grad_norm": 0.6943009495735168, "learning_rate": 5.22608125819135e-05, "loss": 1.5063, "step": 1477 }, { "epoch": 1.9232270657124269, "grad_norm": 1.0624326467514038, "learning_rate": 5.222804718217562e-05, "loss": 1.4184, "step": 1478 }, { "epoch": 1.9245283018867925, "grad_norm": 1.066646933555603, "learning_rate": 5.219528178243775e-05, "loss": 1.785, "step": 1479 }, { "epoch": 1.925829538061158, "grad_norm": 1.1457258462905884, "learning_rate": 5.2162516382699865e-05, "loss": 1.5598, "step": 1480 }, { "epoch": 1.9271307742355237, "grad_norm": 0.8055323362350464, "learning_rate": 5.2129750982962e-05, "loss": 1.4574, "step": 1481 }, { "epoch": 1.9284320104098893, "grad_norm": 1.1001447439193726, "learning_rate": 5.209698558322412e-05, "loss": 1.5951, "step": 1482 }, { "epoch": 1.929733246584255, "grad_norm": 1.2033569812774658, "learning_rate": 5.2064220183486246e-05, "loss": 1.2816, "step": 1483 }, { "epoch": 1.9310344827586206, "grad_norm": 0.8143841624259949, "learning_rate": 5.203145478374837e-05, "loss": 1.874, "step": 1484 }, { "epoch": 1.9323357189329864, "grad_norm": 1.0956439971923828, "learning_rate": 5.199868938401049e-05, "loss": 1.9067, "step": 1485 }, { "epoch": 1.933636955107352, "grad_norm": 0.6541485786437988, "learning_rate": 5.196592398427261e-05, "loss": 1.3633, "step": 1486 }, { "epoch": 1.9349381912817176, "grad_norm": 1.3213372230529785, "learning_rate": 5.193315858453474e-05, "loss": 1.1323, "step": 1487 }, { "epoch": 1.9362394274560832, "grad_norm": 0.9041945338249207, "learning_rate": 5.1900393184796856e-05, "loss": 1.5861, "step": 1488 }, { "epoch": 1.937540663630449, "grad_norm": 1.278778076171875, "learning_rate": 5.186762778505898e-05, "loss": 1.8102, "step": 1489 }, { "epoch": 1.9388418998048147, "grad_norm": 0.6795538663864136, "learning_rate": 5.1834862385321105e-05, "loss": 1.3586, "step": 1490 }, { "epoch": 1.9401431359791803, "grad_norm": 0.7272860407829285, "learning_rate": 5.180209698558322e-05, "loss": 1.5789, "step": 1491 }, { "epoch": 1.941444372153546, "grad_norm": 0.863021194934845, "learning_rate": 5.176933158584535e-05, "loss": 0.9847, "step": 1492 }, { "epoch": 1.9427456083279115, "grad_norm": 0.9420701861381531, "learning_rate": 5.173656618610747e-05, "loss": 1.7531, "step": 1493 }, { "epoch": 1.9440468445022772, "grad_norm": 0.7522069215774536, "learning_rate": 5.170380078636959e-05, "loss": 1.8269, "step": 1494 }, { "epoch": 1.9453480806766428, "grad_norm": 1.0577008724212646, "learning_rate": 5.1671035386631715e-05, "loss": 1.7825, "step": 1495 }, { "epoch": 1.9466493168510084, "grad_norm": 1.0069128274917603, "learning_rate": 5.163826998689384e-05, "loss": 1.7126, "step": 1496 }, { "epoch": 1.947950553025374, "grad_norm": 1.022254228591919, "learning_rate": 5.1605504587155965e-05, "loss": 1.5209, "step": 1497 }, { "epoch": 1.9492517891997396, "grad_norm": 0.5808350443840027, "learning_rate": 5.157273918741808e-05, "loss": 1.531, "step": 1498 }, { "epoch": 1.9505530253741052, "grad_norm": 0.9027696251869202, "learning_rate": 5.153997378768021e-05, "loss": 1.815, "step": 1499 }, { "epoch": 1.951854261548471, "grad_norm": 0.8607378005981445, "learning_rate": 5.150720838794234e-05, "loss": 1.7533, "step": 1500 }, { "epoch": 1.9531554977228367, "grad_norm": 0.661959171295166, "learning_rate": 5.1474442988204464e-05, "loss": 1.4988, "step": 1501 }, { "epoch": 1.9544567338972023, "grad_norm": 1.1740003824234009, "learning_rate": 5.144167758846659e-05, "loss": 1.9157, "step": 1502 }, { "epoch": 1.9557579700715682, "grad_norm": 0.9466021656990051, "learning_rate": 5.1408912188728706e-05, "loss": 1.7212, "step": 1503 }, { "epoch": 1.9570592062459338, "grad_norm": 0.8103978633880615, "learning_rate": 5.137614678899083e-05, "loss": 1.8557, "step": 1504 }, { "epoch": 1.9583604424202994, "grad_norm": 0.8399762511253357, "learning_rate": 5.1343381389252956e-05, "loss": 1.6301, "step": 1505 }, { "epoch": 1.959661678594665, "grad_norm": 0.8513437509536743, "learning_rate": 5.1310615989515074e-05, "loss": 1.4996, "step": 1506 }, { "epoch": 1.9609629147690306, "grad_norm": 2.153118848800659, "learning_rate": 5.12778505897772e-05, "loss": 1.7136, "step": 1507 }, { "epoch": 1.9622641509433962, "grad_norm": 1.0941529273986816, "learning_rate": 5.124508519003932e-05, "loss": 1.8111, "step": 1508 }, { "epoch": 1.9635653871177619, "grad_norm": 0.9551283121109009, "learning_rate": 5.121231979030144e-05, "loss": 1.2272, "step": 1509 }, { "epoch": 1.9648666232921275, "grad_norm": 0.8522684574127197, "learning_rate": 5.1179554390563566e-05, "loss": 1.9693, "step": 1510 }, { "epoch": 1.966167859466493, "grad_norm": 1.3089511394500732, "learning_rate": 5.114678899082569e-05, "loss": 1.5247, "step": 1511 }, { "epoch": 1.9674690956408587, "grad_norm": 1.205409049987793, "learning_rate": 5.111402359108781e-05, "loss": 1.0243, "step": 1512 }, { "epoch": 1.9687703318152243, "grad_norm": 0.9327765107154846, "learning_rate": 5.1081258191349934e-05, "loss": 1.7439, "step": 1513 }, { "epoch": 1.9700715679895902, "grad_norm": 0.8080999255180359, "learning_rate": 5.104849279161206e-05, "loss": 0.8292, "step": 1514 }, { "epoch": 1.9713728041639558, "grad_norm": 0.6233455538749695, "learning_rate": 5.101572739187418e-05, "loss": 1.1482, "step": 1515 }, { "epoch": 1.9726740403383214, "grad_norm": 0.9758902192115784, "learning_rate": 5.09829619921363e-05, "loss": 1.6185, "step": 1516 }, { "epoch": 1.973975276512687, "grad_norm": 1.7408273220062256, "learning_rate": 5.0950196592398426e-05, "loss": 1.6147, "step": 1517 }, { "epoch": 1.9752765126870528, "grad_norm": 1.3517011404037476, "learning_rate": 5.091743119266055e-05, "loss": 1.7693, "step": 1518 }, { "epoch": 1.9765777488614185, "grad_norm": 0.8202415108680725, "learning_rate": 5.088466579292268e-05, "loss": 1.8401, "step": 1519 }, { "epoch": 1.977878985035784, "grad_norm": 1.1633862257003784, "learning_rate": 5.085190039318481e-05, "loss": 1.6278, "step": 1520 }, { "epoch": 1.9791802212101497, "grad_norm": 1.1108871698379517, "learning_rate": 5.0819134993446925e-05, "loss": 1.3117, "step": 1521 }, { "epoch": 1.9804814573845153, "grad_norm": 0.7587534189224243, "learning_rate": 5.078636959370905e-05, "loss": 1.889, "step": 1522 }, { "epoch": 1.981782693558881, "grad_norm": 0.9130533933639526, "learning_rate": 5.0753604193971174e-05, "loss": 0.7309, "step": 1523 }, { "epoch": 1.9830839297332465, "grad_norm": 1.3101400136947632, "learning_rate": 5.072083879423329e-05, "loss": 1.2313, "step": 1524 }, { "epoch": 1.9843851659076122, "grad_norm": 0.8804966807365417, "learning_rate": 5.068807339449542e-05, "loss": 1.4378, "step": 1525 }, { "epoch": 1.9856864020819778, "grad_norm": 0.9668269753456116, "learning_rate": 5.065530799475754e-05, "loss": 1.3434, "step": 1526 }, { "epoch": 1.9869876382563434, "grad_norm": 0.8098539710044861, "learning_rate": 5.062254259501966e-05, "loss": 1.9046, "step": 1527 }, { "epoch": 1.988288874430709, "grad_norm": 1.191476583480835, "learning_rate": 5.0589777195281784e-05, "loss": 1.6676, "step": 1528 }, { "epoch": 1.9895901106050748, "grad_norm": 0.8276931643486023, "learning_rate": 5.055701179554391e-05, "loss": 1.6836, "step": 1529 }, { "epoch": 1.9908913467794405, "grad_norm": 1.040406346321106, "learning_rate": 5.052424639580603e-05, "loss": 1.5061, "step": 1530 }, { "epoch": 1.992192582953806, "grad_norm": 0.8748639822006226, "learning_rate": 5.049148099606815e-05, "loss": 1.3962, "step": 1531 }, { "epoch": 1.993493819128172, "grad_norm": 0.845103919506073, "learning_rate": 5.0458715596330276e-05, "loss": 1.736, "step": 1532 }, { "epoch": 1.9947950553025375, "grad_norm": 0.8671653270721436, "learning_rate": 5.0425950196592394e-05, "loss": 1.8832, "step": 1533 }, { "epoch": 1.9960962914769032, "grad_norm": 0.7623544931411743, "learning_rate": 5.039318479685452e-05, "loss": 1.9467, "step": 1534 }, { "epoch": 1.9973975276512688, "grad_norm": 0.8267562389373779, "learning_rate": 5.0360419397116644e-05, "loss": 1.5016, "step": 1535 }, { "epoch": 1.9986987638256344, "grad_norm": 0.9423741102218628, "learning_rate": 5.032765399737877e-05, "loss": 1.9156, "step": 1536 }, { "epoch": 2.0, "grad_norm": 0.5853365063667297, "learning_rate": 5.02948885976409e-05, "loss": 1.8695, "step": 1537 }, { "epoch": 2.0013012361743656, "grad_norm": 0.8747848272323608, "learning_rate": 5.0262123197903025e-05, "loss": 1.604, "step": 1538 }, { "epoch": 2.0026024723487312, "grad_norm": 1.1049861907958984, "learning_rate": 5.022935779816514e-05, "loss": 1.2607, "step": 1539 }, { "epoch": 2.003903708523097, "grad_norm": 1.0614516735076904, "learning_rate": 5.019659239842727e-05, "loss": 1.1407, "step": 1540 }, { "epoch": 2.0052049446974625, "grad_norm": 1.4323190450668335, "learning_rate": 5.016382699868939e-05, "loss": 0.6036, "step": 1541 }, { "epoch": 2.006506180871828, "grad_norm": 1.739090919494629, "learning_rate": 5.013106159895151e-05, "loss": 1.1248, "step": 1542 }, { "epoch": 2.0078074170461937, "grad_norm": 1.420047402381897, "learning_rate": 5.0098296199213635e-05, "loss": 0.9523, "step": 1543 }, { "epoch": 2.0091086532205593, "grad_norm": 1.069200038909912, "learning_rate": 5.006553079947576e-05, "loss": 0.5338, "step": 1544 }, { "epoch": 2.0104098893949254, "grad_norm": 1.660895824432373, "learning_rate": 5.003276539973788e-05, "loss": 0.9191, "step": 1545 }, { "epoch": 2.011711125569291, "grad_norm": 1.3858795166015625, "learning_rate": 5e-05, "loss": 1.5051, "step": 1546 }, { "epoch": 2.0130123617436566, "grad_norm": 1.1761192083358765, "learning_rate": 4.996723460026213e-05, "loss": 0.8699, "step": 1547 }, { "epoch": 2.0143135979180222, "grad_norm": 1.3125555515289307, "learning_rate": 4.9934469200524245e-05, "loss": 1.299, "step": 1548 }, { "epoch": 2.015614834092388, "grad_norm": 0.9753734469413757, "learning_rate": 4.990170380078637e-05, "loss": 0.9928, "step": 1549 }, { "epoch": 2.0169160702667535, "grad_norm": 0.9737604856491089, "learning_rate": 4.9868938401048495e-05, "loss": 0.9391, "step": 1550 }, { "epoch": 2.018217306441119, "grad_norm": 1.09866464138031, "learning_rate": 4.983617300131062e-05, "loss": 1.3485, "step": 1551 }, { "epoch": 2.0195185426154847, "grad_norm": 1.2723747491836548, "learning_rate": 4.9803407601572744e-05, "loss": 0.9651, "step": 1552 }, { "epoch": 2.0208197787898503, "grad_norm": 1.0172873735427856, "learning_rate": 4.977064220183487e-05, "loss": 1.1386, "step": 1553 }, { "epoch": 2.022121014964216, "grad_norm": 1.0208176374435425, "learning_rate": 4.973787680209699e-05, "loss": 1.1893, "step": 1554 }, { "epoch": 2.0234222511385815, "grad_norm": 1.3184807300567627, "learning_rate": 4.970511140235911e-05, "loss": 1.0886, "step": 1555 }, { "epoch": 2.024723487312947, "grad_norm": 1.1757022142410278, "learning_rate": 4.9672346002621236e-05, "loss": 1.4373, "step": 1556 }, { "epoch": 2.0260247234873128, "grad_norm": 1.1617107391357422, "learning_rate": 4.9639580602883354e-05, "loss": 1.5008, "step": 1557 }, { "epoch": 2.0273259596616784, "grad_norm": 1.2220615148544312, "learning_rate": 4.960681520314548e-05, "loss": 1.042, "step": 1558 }, { "epoch": 2.0286271958360445, "grad_norm": 1.081470251083374, "learning_rate": 4.9574049803407604e-05, "loss": 1.1267, "step": 1559 }, { "epoch": 2.02992843201041, "grad_norm": 1.0622777938842773, "learning_rate": 4.954128440366973e-05, "loss": 0.7194, "step": 1560 }, { "epoch": 2.0312296681847757, "grad_norm": 1.1995891332626343, "learning_rate": 4.950851900393185e-05, "loss": 0.869, "step": 1561 }, { "epoch": 2.0325309043591413, "grad_norm": 1.1450697183609009, "learning_rate": 4.947575360419398e-05, "loss": 0.9334, "step": 1562 }, { "epoch": 2.033832140533507, "grad_norm": 0.6443825960159302, "learning_rate": 4.9442988204456096e-05, "loss": 0.7141, "step": 1563 }, { "epoch": 2.0351333767078725, "grad_norm": 0.8342128396034241, "learning_rate": 4.941022280471822e-05, "loss": 1.112, "step": 1564 }, { "epoch": 2.036434612882238, "grad_norm": 1.3539769649505615, "learning_rate": 4.9377457404980345e-05, "loss": 0.8884, "step": 1565 }, { "epoch": 2.0377358490566038, "grad_norm": 1.065962553024292, "learning_rate": 4.934469200524246e-05, "loss": 1.2504, "step": 1566 }, { "epoch": 2.0390370852309694, "grad_norm": 1.6590265035629272, "learning_rate": 4.931192660550459e-05, "loss": 0.9536, "step": 1567 }, { "epoch": 2.040338321405335, "grad_norm": 1.3187304735183716, "learning_rate": 4.927916120576671e-05, "loss": 1.1402, "step": 1568 }, { "epoch": 2.0416395575797006, "grad_norm": 1.1892246007919312, "learning_rate": 4.924639580602883e-05, "loss": 1.4802, "step": 1569 }, { "epoch": 2.0429407937540662, "grad_norm": 0.9943690896034241, "learning_rate": 4.921363040629096e-05, "loss": 1.2214, "step": 1570 }, { "epoch": 2.044242029928432, "grad_norm": 1.2764736413955688, "learning_rate": 4.918086500655309e-05, "loss": 1.3829, "step": 1571 }, { "epoch": 2.0455432661027975, "grad_norm": 1.2649831771850586, "learning_rate": 4.9148099606815205e-05, "loss": 1.0856, "step": 1572 }, { "epoch": 2.046844502277163, "grad_norm": 1.0045454502105713, "learning_rate": 4.911533420707733e-05, "loss": 1.2924, "step": 1573 }, { "epoch": 2.048145738451529, "grad_norm": 0.8199225664138794, "learning_rate": 4.9082568807339454e-05, "loss": 1.5562, "step": 1574 }, { "epoch": 2.0494469746258948, "grad_norm": 1.1477433443069458, "learning_rate": 4.904980340760157e-05, "loss": 1.0808, "step": 1575 }, { "epoch": 2.0507482108002604, "grad_norm": 0.9502068161964417, "learning_rate": 4.90170380078637e-05, "loss": 0.7945, "step": 1576 }, { "epoch": 2.052049446974626, "grad_norm": 0.9427710771560669, "learning_rate": 4.898427260812582e-05, "loss": 1.2128, "step": 1577 }, { "epoch": 2.0533506831489916, "grad_norm": 1.0262330770492554, "learning_rate": 4.895150720838794e-05, "loss": 1.3688, "step": 1578 }, { "epoch": 2.0546519193233572, "grad_norm": 1.626828670501709, "learning_rate": 4.891874180865007e-05, "loss": 0.5906, "step": 1579 }, { "epoch": 2.055953155497723, "grad_norm": 0.9333571195602417, "learning_rate": 4.8885976408912196e-05, "loss": 1.1146, "step": 1580 }, { "epoch": 2.0572543916720885, "grad_norm": 1.0774143934249878, "learning_rate": 4.8853211009174314e-05, "loss": 1.0885, "step": 1581 }, { "epoch": 2.058555627846454, "grad_norm": 0.8160280585289001, "learning_rate": 4.882044560943644e-05, "loss": 1.3463, "step": 1582 }, { "epoch": 2.0598568640208197, "grad_norm": 0.8435689210891724, "learning_rate": 4.8787680209698563e-05, "loss": 0.7485, "step": 1583 }, { "epoch": 2.0611581001951853, "grad_norm": 2.3531250953674316, "learning_rate": 4.875491480996068e-05, "loss": 1.0048, "step": 1584 }, { "epoch": 2.062459336369551, "grad_norm": 1.3617610931396484, "learning_rate": 4.8722149410222806e-05, "loss": 0.6205, "step": 1585 }, { "epoch": 2.0637605725439165, "grad_norm": 1.1325969696044922, "learning_rate": 4.868938401048493e-05, "loss": 0.8982, "step": 1586 }, { "epoch": 2.065061808718282, "grad_norm": 1.539128303527832, "learning_rate": 4.865661861074705e-05, "loss": 0.867, "step": 1587 }, { "epoch": 2.066363044892648, "grad_norm": 1.2391165494918823, "learning_rate": 4.862385321100918e-05, "loss": 1.1673, "step": 1588 }, { "epoch": 2.067664281067014, "grad_norm": 1.0094211101531982, "learning_rate": 4.8591087811271305e-05, "loss": 0.8076, "step": 1589 }, { "epoch": 2.0689655172413794, "grad_norm": 1.1777911186218262, "learning_rate": 4.855832241153342e-05, "loss": 0.6835, "step": 1590 }, { "epoch": 2.070266753415745, "grad_norm": 0.8212752938270569, "learning_rate": 4.852555701179555e-05, "loss": 0.9898, "step": 1591 }, { "epoch": 2.0715679895901107, "grad_norm": 1.0944513082504272, "learning_rate": 4.849279161205767e-05, "loss": 1.0286, "step": 1592 }, { "epoch": 2.0728692257644763, "grad_norm": 1.1023021936416626, "learning_rate": 4.846002621231979e-05, "loss": 1.372, "step": 1593 }, { "epoch": 2.074170461938842, "grad_norm": 1.0297638177871704, "learning_rate": 4.8427260812581915e-05, "loss": 0.5553, "step": 1594 }, { "epoch": 2.0754716981132075, "grad_norm": 1.3770753145217896, "learning_rate": 4.839449541284404e-05, "loss": 0.5435, "step": 1595 }, { "epoch": 2.076772934287573, "grad_norm": 1.4080607891082764, "learning_rate": 4.836173001310616e-05, "loss": 0.9436, "step": 1596 }, { "epoch": 2.0780741704619388, "grad_norm": 1.716141700744629, "learning_rate": 4.832896461336828e-05, "loss": 0.6949, "step": 1597 }, { "epoch": 2.0793754066363044, "grad_norm": 1.2890692949295044, "learning_rate": 4.8296199213630414e-05, "loss": 0.5834, "step": 1598 }, { "epoch": 2.08067664281067, "grad_norm": 1.701550006866455, "learning_rate": 4.826343381389253e-05, "loss": 1.1361, "step": 1599 }, { "epoch": 2.0819778789850356, "grad_norm": 0.9375856518745422, "learning_rate": 4.823066841415466e-05, "loss": 0.91, "step": 1600 }, { "epoch": 2.0832791151594012, "grad_norm": 0.9030932784080505, "learning_rate": 4.819790301441678e-05, "loss": 1.1656, "step": 1601 }, { "epoch": 2.0845803513337673, "grad_norm": 0.8589449524879456, "learning_rate": 4.81651376146789e-05, "loss": 1.0045, "step": 1602 }, { "epoch": 2.085881587508133, "grad_norm": 1.6834168434143066, "learning_rate": 4.8132372214941024e-05, "loss": 1.1192, "step": 1603 }, { "epoch": 2.0871828236824985, "grad_norm": 1.1657421588897705, "learning_rate": 4.809960681520315e-05, "loss": 1.4225, "step": 1604 }, { "epoch": 2.088484059856864, "grad_norm": 1.0284310579299927, "learning_rate": 4.806684141546527e-05, "loss": 1.3403, "step": 1605 }, { "epoch": 2.0897852960312298, "grad_norm": 1.3400870561599731, "learning_rate": 4.803407601572739e-05, "loss": 0.8927, "step": 1606 }, { "epoch": 2.0910865322055954, "grad_norm": 0.914268970489502, "learning_rate": 4.8001310615989516e-05, "loss": 0.5144, "step": 1607 }, { "epoch": 2.092387768379961, "grad_norm": 1.402650237083435, "learning_rate": 4.796854521625164e-05, "loss": 1.2361, "step": 1608 }, { "epoch": 2.0936890045543266, "grad_norm": 1.3522999286651611, "learning_rate": 4.7935779816513766e-05, "loss": 1.1351, "step": 1609 }, { "epoch": 2.094990240728692, "grad_norm": 1.1513317823410034, "learning_rate": 4.790301441677589e-05, "loss": 0.8585, "step": 1610 }, { "epoch": 2.096291476903058, "grad_norm": 0.8446826934814453, "learning_rate": 4.787024901703801e-05, "loss": 0.9679, "step": 1611 }, { "epoch": 2.0975927130774235, "grad_norm": 0.9937654137611389, "learning_rate": 4.7837483617300133e-05, "loss": 1.132, "step": 1612 }, { "epoch": 2.098893949251789, "grad_norm": 1.0347068309783936, "learning_rate": 4.780471821756226e-05, "loss": 1.248, "step": 1613 }, { "epoch": 2.1001951854261547, "grad_norm": 1.235241174697876, "learning_rate": 4.7771952817824376e-05, "loss": 1.0922, "step": 1614 }, { "epoch": 2.1014964216005203, "grad_norm": 1.3549911975860596, "learning_rate": 4.77391874180865e-05, "loss": 0.9129, "step": 1615 }, { "epoch": 2.102797657774886, "grad_norm": 2.1170382499694824, "learning_rate": 4.7706422018348626e-05, "loss": 1.1368, "step": 1616 }, { "epoch": 2.104098893949252, "grad_norm": 2.1323060989379883, "learning_rate": 4.767365661861075e-05, "loss": 1.0272, "step": 1617 }, { "epoch": 2.1054001301236176, "grad_norm": 0.8987525701522827, "learning_rate": 4.7640891218872875e-05, "loss": 1.1814, "step": 1618 }, { "epoch": 2.106701366297983, "grad_norm": 1.2682104110717773, "learning_rate": 4.7608125819135e-05, "loss": 1.3335, "step": 1619 }, { "epoch": 2.108002602472349, "grad_norm": 1.365801215171814, "learning_rate": 4.757536041939712e-05, "loss": 0.5186, "step": 1620 }, { "epoch": 2.1093038386467144, "grad_norm": 1.1000806093215942, "learning_rate": 4.754259501965924e-05, "loss": 1.3373, "step": 1621 }, { "epoch": 2.11060507482108, "grad_norm": 1.1319280862808228, "learning_rate": 4.750982961992137e-05, "loss": 1.2111, "step": 1622 }, { "epoch": 2.1119063109954457, "grad_norm": 1.371524453163147, "learning_rate": 4.7477064220183485e-05, "loss": 1.1495, "step": 1623 }, { "epoch": 2.1132075471698113, "grad_norm": 1.5801575183868408, "learning_rate": 4.744429882044561e-05, "loss": 1.0985, "step": 1624 }, { "epoch": 2.114508783344177, "grad_norm": 1.0830000638961792, "learning_rate": 4.7411533420707735e-05, "loss": 1.2572, "step": 1625 }, { "epoch": 2.1158100195185425, "grad_norm": 0.8902335166931152, "learning_rate": 4.737876802096986e-05, "loss": 0.4571, "step": 1626 }, { "epoch": 2.117111255692908, "grad_norm": 1.0787659883499146, "learning_rate": 4.7346002621231984e-05, "loss": 1.0726, "step": 1627 }, { "epoch": 2.1184124918672738, "grad_norm": 1.0098564624786377, "learning_rate": 4.731323722149411e-05, "loss": 1.5829, "step": 1628 }, { "epoch": 2.1197137280416394, "grad_norm": 1.003749132156372, "learning_rate": 4.728047182175623e-05, "loss": 0.7456, "step": 1629 }, { "epoch": 2.121014964216005, "grad_norm": 1.2271437644958496, "learning_rate": 4.724770642201835e-05, "loss": 0.8562, "step": 1630 }, { "epoch": 2.122316200390371, "grad_norm": 0.9275023937225342, "learning_rate": 4.7214941022280476e-05, "loss": 1.4424, "step": 1631 }, { "epoch": 2.1236174365647367, "grad_norm": 1.1140565872192383, "learning_rate": 4.7182175622542594e-05, "loss": 1.0996, "step": 1632 }, { "epoch": 2.1249186727391023, "grad_norm": 1.2821550369262695, "learning_rate": 4.714941022280472e-05, "loss": 1.1353, "step": 1633 }, { "epoch": 2.126219908913468, "grad_norm": 1.5123871564865112, "learning_rate": 4.7116644823066844e-05, "loss": 0.9416, "step": 1634 }, { "epoch": 2.1275211450878335, "grad_norm": 1.4036561250686646, "learning_rate": 4.708387942332897e-05, "loss": 1.179, "step": 1635 }, { "epoch": 2.128822381262199, "grad_norm": 1.2531670331954956, "learning_rate": 4.705111402359109e-05, "loss": 1.0575, "step": 1636 }, { "epoch": 2.1301236174365648, "grad_norm": 1.1193475723266602, "learning_rate": 4.701834862385321e-05, "loss": 1.1047, "step": 1637 }, { "epoch": 2.1314248536109304, "grad_norm": 1.306211233139038, "learning_rate": 4.6985583224115336e-05, "loss": 0.7175, "step": 1638 }, { "epoch": 2.132726089785296, "grad_norm": 1.006590485572815, "learning_rate": 4.695281782437746e-05, "loss": 1.1499, "step": 1639 }, { "epoch": 2.1340273259596616, "grad_norm": 1.1653376817703247, "learning_rate": 4.6920052424639585e-05, "loss": 0.977, "step": 1640 }, { "epoch": 2.135328562134027, "grad_norm": 0.9865640997886658, "learning_rate": 4.68872870249017e-05, "loss": 1.3172, "step": 1641 }, { "epoch": 2.136629798308393, "grad_norm": 0.9688118100166321, "learning_rate": 4.685452162516383e-05, "loss": 1.2622, "step": 1642 }, { "epoch": 2.1379310344827585, "grad_norm": 1.077317476272583, "learning_rate": 4.682175622542595e-05, "loss": 1.231, "step": 1643 }, { "epoch": 2.139232270657124, "grad_norm": 1.3578321933746338, "learning_rate": 4.678899082568808e-05, "loss": 1.1093, "step": 1644 }, { "epoch": 2.14053350683149, "grad_norm": 1.0971338748931885, "learning_rate": 4.67562254259502e-05, "loss": 0.6432, "step": 1645 }, { "epoch": 2.1418347430058557, "grad_norm": 1.4313931465148926, "learning_rate": 4.672346002621232e-05, "loss": 1.1392, "step": 1646 }, { "epoch": 2.1431359791802214, "grad_norm": 1.017157793045044, "learning_rate": 4.6690694626474445e-05, "loss": 1.3366, "step": 1647 }, { "epoch": 2.144437215354587, "grad_norm": 1.4007941484451294, "learning_rate": 4.665792922673657e-05, "loss": 0.9171, "step": 1648 }, { "epoch": 2.1457384515289526, "grad_norm": 1.0594149827957153, "learning_rate": 4.6625163826998694e-05, "loss": 0.851, "step": 1649 }, { "epoch": 2.147039687703318, "grad_norm": 1.0466200113296509, "learning_rate": 4.659239842726081e-05, "loss": 0.7778, "step": 1650 }, { "epoch": 2.148340923877684, "grad_norm": 1.3441708087921143, "learning_rate": 4.655963302752294e-05, "loss": 1.294, "step": 1651 }, { "epoch": 2.1496421600520494, "grad_norm": 1.1013283729553223, "learning_rate": 4.652686762778506e-05, "loss": 1.5974, "step": 1652 }, { "epoch": 2.150943396226415, "grad_norm": 1.1627789735794067, "learning_rate": 4.649410222804718e-05, "loss": 1.1011, "step": 1653 }, { "epoch": 2.1522446324007807, "grad_norm": 1.2281039953231812, "learning_rate": 4.646133682830931e-05, "loss": 0.8508, "step": 1654 }, { "epoch": 2.1535458685751463, "grad_norm": 1.0901905298233032, "learning_rate": 4.642857142857143e-05, "loss": 1.1271, "step": 1655 }, { "epoch": 2.154847104749512, "grad_norm": 1.100314736366272, "learning_rate": 4.6395806028833554e-05, "loss": 1.381, "step": 1656 }, { "epoch": 2.1561483409238775, "grad_norm": 1.0517851114273071, "learning_rate": 4.636304062909568e-05, "loss": 0.6231, "step": 1657 }, { "epoch": 2.157449577098243, "grad_norm": 1.3014403581619263, "learning_rate": 4.6330275229357804e-05, "loss": 1.0283, "step": 1658 }, { "epoch": 2.1587508132726088, "grad_norm": 1.0441150665283203, "learning_rate": 4.629750982961992e-05, "loss": 1.0697, "step": 1659 }, { "epoch": 2.160052049446975, "grad_norm": 1.4786328077316284, "learning_rate": 4.6264744429882046e-05, "loss": 0.885, "step": 1660 }, { "epoch": 2.1613532856213404, "grad_norm": 1.1570929288864136, "learning_rate": 4.623197903014417e-05, "loss": 0.7112, "step": 1661 }, { "epoch": 2.162654521795706, "grad_norm": 1.2718278169631958, "learning_rate": 4.619921363040629e-05, "loss": 0.4245, "step": 1662 }, { "epoch": 2.1639557579700717, "grad_norm": 1.211918592453003, "learning_rate": 4.616644823066842e-05, "loss": 0.561, "step": 1663 }, { "epoch": 2.1652569941444373, "grad_norm": 1.1968903541564941, "learning_rate": 4.613368283093054e-05, "loss": 0.4963, "step": 1664 }, { "epoch": 2.166558230318803, "grad_norm": 1.1355044841766357, "learning_rate": 4.610091743119266e-05, "loss": 0.9623, "step": 1665 }, { "epoch": 2.1678594664931685, "grad_norm": 1.165217399597168, "learning_rate": 4.606815203145479e-05, "loss": 1.1308, "step": 1666 }, { "epoch": 2.169160702667534, "grad_norm": 1.1886589527130127, "learning_rate": 4.6035386631716906e-05, "loss": 0.8139, "step": 1667 }, { "epoch": 2.1704619388418998, "grad_norm": 0.9523638486862183, "learning_rate": 4.600262123197903e-05, "loss": 1.1714, "step": 1668 }, { "epoch": 2.1717631750162654, "grad_norm": 1.1140670776367188, "learning_rate": 4.5969855832241155e-05, "loss": 1.1357, "step": 1669 }, { "epoch": 2.173064411190631, "grad_norm": 1.256430983543396, "learning_rate": 4.593709043250328e-05, "loss": 1.1237, "step": 1670 }, { "epoch": 2.1743656473649966, "grad_norm": 1.15438711643219, "learning_rate": 4.59043250327654e-05, "loss": 1.5401, "step": 1671 }, { "epoch": 2.175666883539362, "grad_norm": 1.1931722164154053, "learning_rate": 4.587155963302753e-05, "loss": 0.7907, "step": 1672 }, { "epoch": 2.176968119713728, "grad_norm": 1.5109527111053467, "learning_rate": 4.583879423328965e-05, "loss": 0.9427, "step": 1673 }, { "epoch": 2.178269355888094, "grad_norm": 1.5962048768997192, "learning_rate": 4.580602883355177e-05, "loss": 1.1794, "step": 1674 }, { "epoch": 2.1795705920624595, "grad_norm": 1.1751439571380615, "learning_rate": 4.57732634338139e-05, "loss": 1.5399, "step": 1675 }, { "epoch": 2.180871828236825, "grad_norm": 1.3315186500549316, "learning_rate": 4.5740498034076015e-05, "loss": 0.8075, "step": 1676 }, { "epoch": 2.1821730644111907, "grad_norm": 1.4134680032730103, "learning_rate": 4.570773263433814e-05, "loss": 0.7102, "step": 1677 }, { "epoch": 2.1834743005855564, "grad_norm": 1.128707766532898, "learning_rate": 4.5674967234600264e-05, "loss": 0.7843, "step": 1678 }, { "epoch": 2.184775536759922, "grad_norm": 1.2361034154891968, "learning_rate": 4.564220183486239e-05, "loss": 0.8429, "step": 1679 }, { "epoch": 2.1860767729342876, "grad_norm": 0.9949290156364441, "learning_rate": 4.560943643512451e-05, "loss": 1.5436, "step": 1680 }, { "epoch": 2.187378009108653, "grad_norm": 0.9801502227783203, "learning_rate": 4.557667103538664e-05, "loss": 0.8815, "step": 1681 }, { "epoch": 2.188679245283019, "grad_norm": 0.9306057691574097, "learning_rate": 4.5543905635648757e-05, "loss": 0.5702, "step": 1682 }, { "epoch": 2.1899804814573844, "grad_norm": 1.1577486991882324, "learning_rate": 4.551114023591088e-05, "loss": 1.0968, "step": 1683 }, { "epoch": 2.19128171763175, "grad_norm": 1.1442582607269287, "learning_rate": 4.5478374836173006e-05, "loss": 0.7556, "step": 1684 }, { "epoch": 2.1925829538061157, "grad_norm": 1.0106816291809082, "learning_rate": 4.5445609436435124e-05, "loss": 0.6281, "step": 1685 }, { "epoch": 2.1938841899804813, "grad_norm": 1.0324820280075073, "learning_rate": 4.541284403669725e-05, "loss": 1.2752, "step": 1686 }, { "epoch": 2.195185426154847, "grad_norm": 1.153624176979065, "learning_rate": 4.5380078636959374e-05, "loss": 1.1186, "step": 1687 }, { "epoch": 2.196486662329213, "grad_norm": 1.268710732460022, "learning_rate": 4.534731323722149e-05, "loss": 1.2103, "step": 1688 }, { "epoch": 2.1977878985035786, "grad_norm": 1.1776955127716064, "learning_rate": 4.5314547837483616e-05, "loss": 0.7624, "step": 1689 }, { "epoch": 2.199089134677944, "grad_norm": 1.0704044103622437, "learning_rate": 4.528178243774574e-05, "loss": 0.3542, "step": 1690 }, { "epoch": 2.20039037085231, "grad_norm": 0.7507322430610657, "learning_rate": 4.5249017038007866e-05, "loss": 1.647, "step": 1691 }, { "epoch": 2.2016916070266754, "grad_norm": 1.5887032747268677, "learning_rate": 4.521625163826999e-05, "loss": 1.461, "step": 1692 }, { "epoch": 2.202992843201041, "grad_norm": 1.0627305507659912, "learning_rate": 4.5183486238532115e-05, "loss": 1.0028, "step": 1693 }, { "epoch": 2.2042940793754067, "grad_norm": 1.5157203674316406, "learning_rate": 4.515072083879423e-05, "loss": 0.7626, "step": 1694 }, { "epoch": 2.2055953155497723, "grad_norm": 1.226893424987793, "learning_rate": 4.511795543905636e-05, "loss": 1.0362, "step": 1695 }, { "epoch": 2.206896551724138, "grad_norm": 1.214036464691162, "learning_rate": 4.508519003931848e-05, "loss": 1.4573, "step": 1696 }, { "epoch": 2.2081977878985035, "grad_norm": 1.0246754884719849, "learning_rate": 4.50524246395806e-05, "loss": 0.7875, "step": 1697 }, { "epoch": 2.209499024072869, "grad_norm": 1.1532909870147705, "learning_rate": 4.5019659239842725e-05, "loss": 0.7727, "step": 1698 }, { "epoch": 2.2108002602472347, "grad_norm": 0.9722341299057007, "learning_rate": 4.498689384010485e-05, "loss": 0.3739, "step": 1699 }, { "epoch": 2.2121014964216004, "grad_norm": 0.8917628526687622, "learning_rate": 4.4954128440366975e-05, "loss": 1.0667, "step": 1700 }, { "epoch": 2.213402732595966, "grad_norm": 1.0711785554885864, "learning_rate": 4.49213630406291e-05, "loss": 0.7571, "step": 1701 }, { "epoch": 2.2147039687703316, "grad_norm": 1.2819799184799194, "learning_rate": 4.4888597640891224e-05, "loss": 1.1021, "step": 1702 }, { "epoch": 2.2160052049446977, "grad_norm": 1.1146317720413208, "learning_rate": 4.485583224115334e-05, "loss": 0.7639, "step": 1703 }, { "epoch": 2.2173064411190633, "grad_norm": 1.4317576885223389, "learning_rate": 4.482306684141547e-05, "loss": 1.222, "step": 1704 }, { "epoch": 2.218607677293429, "grad_norm": 0.9012517929077148, "learning_rate": 4.479030144167759e-05, "loss": 0.7199, "step": 1705 }, { "epoch": 2.2199089134677945, "grad_norm": 1.0288690328598022, "learning_rate": 4.475753604193971e-05, "loss": 1.2392, "step": 1706 }, { "epoch": 2.22121014964216, "grad_norm": 1.078005075454712, "learning_rate": 4.4724770642201834e-05, "loss": 1.474, "step": 1707 }, { "epoch": 2.2225113858165257, "grad_norm": 0.9777456521987915, "learning_rate": 4.469200524246396e-05, "loss": 1.4645, "step": 1708 }, { "epoch": 2.2238126219908914, "grad_norm": 1.3188382387161255, "learning_rate": 4.4659239842726084e-05, "loss": 1.2543, "step": 1709 }, { "epoch": 2.225113858165257, "grad_norm": 1.2461968660354614, "learning_rate": 4.462647444298821e-05, "loss": 1.1357, "step": 1710 }, { "epoch": 2.2264150943396226, "grad_norm": 1.0244475603103638, "learning_rate": 4.459370904325033e-05, "loss": 1.2632, "step": 1711 }, { "epoch": 2.227716330513988, "grad_norm": 1.197827696800232, "learning_rate": 4.456094364351245e-05, "loss": 0.8907, "step": 1712 }, { "epoch": 2.229017566688354, "grad_norm": 1.0582541227340698, "learning_rate": 4.4528178243774576e-05, "loss": 1.0489, "step": 1713 }, { "epoch": 2.2303188028627194, "grad_norm": 1.0055848360061646, "learning_rate": 4.44954128440367e-05, "loss": 1.5889, "step": 1714 }, { "epoch": 2.231620039037085, "grad_norm": 0.929370105266571, "learning_rate": 4.446264744429882e-05, "loss": 1.1535, "step": 1715 }, { "epoch": 2.2329212752114507, "grad_norm": 1.3841906785964966, "learning_rate": 4.4429882044560943e-05, "loss": 0.8177, "step": 1716 }, { "epoch": 2.2342225113858163, "grad_norm": 1.0122038125991821, "learning_rate": 4.439711664482307e-05, "loss": 1.5165, "step": 1717 }, { "epoch": 2.2355237475601824, "grad_norm": 1.3264724016189575, "learning_rate": 4.4364351245085186e-05, "loss": 0.8092, "step": 1718 }, { "epoch": 2.236824983734548, "grad_norm": 1.2572050094604492, "learning_rate": 4.433158584534732e-05, "loss": 1.212, "step": 1719 }, { "epoch": 2.2381262199089136, "grad_norm": 1.025423288345337, "learning_rate": 4.429882044560944e-05, "loss": 1.1433, "step": 1720 }, { "epoch": 2.239427456083279, "grad_norm": 1.1811976432800293, "learning_rate": 4.426605504587156e-05, "loss": 1.1129, "step": 1721 }, { "epoch": 2.240728692257645, "grad_norm": 0.9251207113265991, "learning_rate": 4.4233289646133685e-05, "loss": 1.7148, "step": 1722 }, { "epoch": 2.2420299284320104, "grad_norm": 1.2217295169830322, "learning_rate": 4.420052424639581e-05, "loss": 1.7615, "step": 1723 }, { "epoch": 2.243331164606376, "grad_norm": 0.9518020153045654, "learning_rate": 4.416775884665793e-05, "loss": 1.1577, "step": 1724 }, { "epoch": 2.2446324007807417, "grad_norm": 0.8267199397087097, "learning_rate": 4.413499344692005e-05, "loss": 1.3874, "step": 1725 }, { "epoch": 2.2459336369551073, "grad_norm": 1.0343259572982788, "learning_rate": 4.410222804718218e-05, "loss": 0.7663, "step": 1726 }, { "epoch": 2.247234873129473, "grad_norm": 1.1362665891647339, "learning_rate": 4.4069462647444295e-05, "loss": 1.2597, "step": 1727 }, { "epoch": 2.2485361093038385, "grad_norm": 1.2902472019195557, "learning_rate": 4.403669724770643e-05, "loss": 0.952, "step": 1728 }, { "epoch": 2.249837345478204, "grad_norm": 0.8744992017745972, "learning_rate": 4.400393184796855e-05, "loss": 1.1243, "step": 1729 }, { "epoch": 2.2511385816525697, "grad_norm": 1.1168807744979858, "learning_rate": 4.397116644823067e-05, "loss": 0.9125, "step": 1730 }, { "epoch": 2.252439817826936, "grad_norm": 1.1996843814849854, "learning_rate": 4.3938401048492794e-05, "loss": 1.0911, "step": 1731 }, { "epoch": 2.2537410540013014, "grad_norm": 1.2427400350570679, "learning_rate": 4.390563564875492e-05, "loss": 0.6501, "step": 1732 }, { "epoch": 2.255042290175667, "grad_norm": 1.0133510828018188, "learning_rate": 4.387287024901704e-05, "loss": 1.6842, "step": 1733 }, { "epoch": 2.2563435263500327, "grad_norm": 1.4306493997573853, "learning_rate": 4.384010484927916e-05, "loss": 0.8814, "step": 1734 }, { "epoch": 2.2576447625243983, "grad_norm": 1.364565134048462, "learning_rate": 4.3807339449541286e-05, "loss": 1.0699, "step": 1735 }, { "epoch": 2.258945998698764, "grad_norm": 1.3825478553771973, "learning_rate": 4.3774574049803404e-05, "loss": 0.7547, "step": 1736 }, { "epoch": 2.2602472348731295, "grad_norm": 1.158313512802124, "learning_rate": 4.3741808650065536e-05, "loss": 1.0676, "step": 1737 }, { "epoch": 2.261548471047495, "grad_norm": 1.2459672689437866, "learning_rate": 4.370904325032766e-05, "loss": 1.1959, "step": 1738 }, { "epoch": 2.2628497072218607, "grad_norm": 1.2115509510040283, "learning_rate": 4.367627785058978e-05, "loss": 1.088, "step": 1739 }, { "epoch": 2.2641509433962264, "grad_norm": 1.426567792892456, "learning_rate": 4.36435124508519e-05, "loss": 0.9607, "step": 1740 }, { "epoch": 2.265452179570592, "grad_norm": 1.2074908018112183, "learning_rate": 4.361074705111403e-05, "loss": 1.1531, "step": 1741 }, { "epoch": 2.2667534157449576, "grad_norm": 1.2420618534088135, "learning_rate": 4.3577981651376146e-05, "loss": 1.4849, "step": 1742 }, { "epoch": 2.268054651919323, "grad_norm": 1.0459879636764526, "learning_rate": 4.354521625163827e-05, "loss": 1.2139, "step": 1743 }, { "epoch": 2.269355888093689, "grad_norm": 1.0822988748550415, "learning_rate": 4.3512450851900395e-05, "loss": 0.9427, "step": 1744 }, { "epoch": 2.2706571242680544, "grad_norm": 1.5372992753982544, "learning_rate": 4.3479685452162513e-05, "loss": 1.0869, "step": 1745 }, { "epoch": 2.2719583604424205, "grad_norm": 0.9946892857551575, "learning_rate": 4.344692005242464e-05, "loss": 1.0119, "step": 1746 }, { "epoch": 2.273259596616786, "grad_norm": 1.1908199787139893, "learning_rate": 4.341415465268677e-05, "loss": 0.8478, "step": 1747 }, { "epoch": 2.2745608327911517, "grad_norm": 1.1663750410079956, "learning_rate": 4.338138925294889e-05, "loss": 1.0484, "step": 1748 }, { "epoch": 2.2758620689655173, "grad_norm": 1.4847761392593384, "learning_rate": 4.334862385321101e-05, "loss": 1.1735, "step": 1749 }, { "epoch": 2.277163305139883, "grad_norm": 1.280993938446045, "learning_rate": 4.331585845347314e-05, "loss": 0.8166, "step": 1750 }, { "epoch": 2.2784645413142486, "grad_norm": 0.8333806395530701, "learning_rate": 4.3283093053735255e-05, "loss": 1.725, "step": 1751 }, { "epoch": 2.279765777488614, "grad_norm": 1.1756069660186768, "learning_rate": 4.325032765399738e-05, "loss": 1.1058, "step": 1752 }, { "epoch": 2.28106701366298, "grad_norm": 1.2033580541610718, "learning_rate": 4.3217562254259505e-05, "loss": 1.1624, "step": 1753 }, { "epoch": 2.2823682498373454, "grad_norm": 0.9953135848045349, "learning_rate": 4.318479685452162e-05, "loss": 1.0262, "step": 1754 }, { "epoch": 2.283669486011711, "grad_norm": 1.0724680423736572, "learning_rate": 4.315203145478375e-05, "loss": 1.0755, "step": 1755 }, { "epoch": 2.2849707221860767, "grad_norm": 1.062392234802246, "learning_rate": 4.311926605504588e-05, "loss": 1.3207, "step": 1756 }, { "epoch": 2.2862719583604423, "grad_norm": 1.1308764219284058, "learning_rate": 4.3086500655308e-05, "loss": 0.6564, "step": 1757 }, { "epoch": 2.287573194534808, "grad_norm": 1.4325194358825684, "learning_rate": 4.305373525557012e-05, "loss": 0.6789, "step": 1758 }, { "epoch": 2.288874430709174, "grad_norm": 0.8219449520111084, "learning_rate": 4.3020969855832246e-05, "loss": 1.2548, "step": 1759 }, { "epoch": 2.290175666883539, "grad_norm": 1.3199156522750854, "learning_rate": 4.2988204456094364e-05, "loss": 1.059, "step": 1760 }, { "epoch": 2.291476903057905, "grad_norm": 1.4385019540786743, "learning_rate": 4.295543905635649e-05, "loss": 0.5338, "step": 1761 }, { "epoch": 2.292778139232271, "grad_norm": 0.9361728429794312, "learning_rate": 4.2922673656618614e-05, "loss": 0.869, "step": 1762 }, { "epoch": 2.2940793754066364, "grad_norm": 1.0752485990524292, "learning_rate": 4.288990825688073e-05, "loss": 1.1769, "step": 1763 }, { "epoch": 2.295380611581002, "grad_norm": 1.5554713010787964, "learning_rate": 4.2857142857142856e-05, "loss": 0.8322, "step": 1764 }, { "epoch": 2.2966818477553677, "grad_norm": 0.6535924077033997, "learning_rate": 4.282437745740499e-05, "loss": 1.409, "step": 1765 }, { "epoch": 2.2979830839297333, "grad_norm": 1.1598364114761353, "learning_rate": 4.2791612057667106e-05, "loss": 1.0473, "step": 1766 }, { "epoch": 2.299284320104099, "grad_norm": 0.95011967420578, "learning_rate": 4.275884665792923e-05, "loss": 1.2994, "step": 1767 }, { "epoch": 2.3005855562784645, "grad_norm": 1.0208624601364136, "learning_rate": 4.2726081258191355e-05, "loss": 1.107, "step": 1768 }, { "epoch": 2.30188679245283, "grad_norm": 0.943922758102417, "learning_rate": 4.269331585845347e-05, "loss": 0.7643, "step": 1769 }, { "epoch": 2.3031880286271957, "grad_norm": 1.1610279083251953, "learning_rate": 4.26605504587156e-05, "loss": 1.01, "step": 1770 }, { "epoch": 2.3044892648015614, "grad_norm": 0.9191147089004517, "learning_rate": 4.262778505897772e-05, "loss": 1.2232, "step": 1771 }, { "epoch": 2.305790500975927, "grad_norm": 1.3710424900054932, "learning_rate": 4.259501965923984e-05, "loss": 1.1655, "step": 1772 }, { "epoch": 2.3070917371502926, "grad_norm": 1.0046590566635132, "learning_rate": 4.2562254259501965e-05, "loss": 1.2191, "step": 1773 }, { "epoch": 2.3083929733246586, "grad_norm": 1.3920799493789673, "learning_rate": 4.25294888597641e-05, "loss": 1.4904, "step": 1774 }, { "epoch": 2.3096942094990243, "grad_norm": 1.8670727014541626, "learning_rate": 4.2496723460026215e-05, "loss": 1.0104, "step": 1775 }, { "epoch": 2.31099544567339, "grad_norm": 1.4510011672973633, "learning_rate": 4.246395806028834e-05, "loss": 0.7377, "step": 1776 }, { "epoch": 2.3122966818477555, "grad_norm": 0.9878943562507629, "learning_rate": 4.2431192660550464e-05, "loss": 1.4162, "step": 1777 }, { "epoch": 2.313597918022121, "grad_norm": 1.1513895988464355, "learning_rate": 4.239842726081258e-05, "loss": 1.5876, "step": 1778 }, { "epoch": 2.3148991541964867, "grad_norm": 0.9727165699005127, "learning_rate": 4.236566186107471e-05, "loss": 0.7518, "step": 1779 }, { "epoch": 2.3162003903708523, "grad_norm": 1.0563483238220215, "learning_rate": 4.233289646133683e-05, "loss": 1.3615, "step": 1780 }, { "epoch": 2.317501626545218, "grad_norm": 1.3328278064727783, "learning_rate": 4.230013106159895e-05, "loss": 0.8787, "step": 1781 }, { "epoch": 2.3188028627195836, "grad_norm": 1.2878941297531128, "learning_rate": 4.2267365661861074e-05, "loss": 0.8516, "step": 1782 }, { "epoch": 2.320104098893949, "grad_norm": 1.108061671257019, "learning_rate": 4.22346002621232e-05, "loss": 1.2118, "step": 1783 }, { "epoch": 2.321405335068315, "grad_norm": 0.7487668991088867, "learning_rate": 4.2201834862385324e-05, "loss": 1.5677, "step": 1784 }, { "epoch": 2.3227065712426804, "grad_norm": 1.111276626586914, "learning_rate": 4.216906946264745e-05, "loss": 1.6706, "step": 1785 }, { "epoch": 2.324007807417046, "grad_norm": 1.079338788986206, "learning_rate": 4.2136304062909573e-05, "loss": 1.3414, "step": 1786 }, { "epoch": 2.3253090435914117, "grad_norm": 1.120708703994751, "learning_rate": 4.210353866317169e-05, "loss": 1.5325, "step": 1787 }, { "epoch": 2.3266102797657773, "grad_norm": 1.0741369724273682, "learning_rate": 4.2070773263433816e-05, "loss": 0.9172, "step": 1788 }, { "epoch": 2.3279115159401433, "grad_norm": 1.2654404640197754, "learning_rate": 4.203800786369594e-05, "loss": 1.4897, "step": 1789 }, { "epoch": 2.329212752114509, "grad_norm": 1.1906063556671143, "learning_rate": 4.200524246395806e-05, "loss": 1.5699, "step": 1790 }, { "epoch": 2.3305139882888746, "grad_norm": 1.0894522666931152, "learning_rate": 4.1972477064220184e-05, "loss": 0.7471, "step": 1791 }, { "epoch": 2.33181522446324, "grad_norm": 1.0740772485733032, "learning_rate": 4.193971166448231e-05, "loss": 0.7543, "step": 1792 }, { "epoch": 2.333116460637606, "grad_norm": 0.9592959880828857, "learning_rate": 4.190694626474443e-05, "loss": 0.9066, "step": 1793 }, { "epoch": 2.3344176968119714, "grad_norm": 1.020437240600586, "learning_rate": 4.187418086500656e-05, "loss": 1.1588, "step": 1794 }, { "epoch": 2.335718932986337, "grad_norm": 1.0770450830459595, "learning_rate": 4.184141546526868e-05, "loss": 0.9139, "step": 1795 }, { "epoch": 2.3370201691607027, "grad_norm": 1.8688472509384155, "learning_rate": 4.18086500655308e-05, "loss": 1.167, "step": 1796 }, { "epoch": 2.3383214053350683, "grad_norm": 1.1522212028503418, "learning_rate": 4.1775884665792925e-05, "loss": 1.225, "step": 1797 }, { "epoch": 2.339622641509434, "grad_norm": 1.122177004814148, "learning_rate": 4.174311926605505e-05, "loss": 1.1363, "step": 1798 }, { "epoch": 2.3409238776837995, "grad_norm": 1.9162256717681885, "learning_rate": 4.171035386631717e-05, "loss": 0.6031, "step": 1799 }, { "epoch": 2.342225113858165, "grad_norm": 1.3179223537445068, "learning_rate": 4.167758846657929e-05, "loss": 0.7592, "step": 1800 }, { "epoch": 2.3435263500325307, "grad_norm": 1.1932363510131836, "learning_rate": 4.164482306684142e-05, "loss": 1.4141, "step": 1801 }, { "epoch": 2.344827586206897, "grad_norm": 1.2240961790084839, "learning_rate": 4.161205766710354e-05, "loss": 1.1922, "step": 1802 }, { "epoch": 2.346128822381262, "grad_norm": 1.2638601064682007, "learning_rate": 4.157929226736567e-05, "loss": 1.5585, "step": 1803 }, { "epoch": 2.347430058555628, "grad_norm": 1.1544265747070312, "learning_rate": 4.154652686762779e-05, "loss": 1.1551, "step": 1804 }, { "epoch": 2.3487312947299936, "grad_norm": 1.0253955125808716, "learning_rate": 4.151376146788991e-05, "loss": 1.1151, "step": 1805 }, { "epoch": 2.3500325309043593, "grad_norm": 1.0229535102844238, "learning_rate": 4.1480996068152034e-05, "loss": 0.7328, "step": 1806 }, { "epoch": 2.351333767078725, "grad_norm": 1.2302680015563965, "learning_rate": 4.144823066841416e-05, "loss": 1.1665, "step": 1807 }, { "epoch": 2.3526350032530905, "grad_norm": 1.133195400238037, "learning_rate": 4.141546526867628e-05, "loss": 1.1034, "step": 1808 }, { "epoch": 2.353936239427456, "grad_norm": 1.7061669826507568, "learning_rate": 4.13826998689384e-05, "loss": 0.5894, "step": 1809 }, { "epoch": 2.3552374756018217, "grad_norm": 1.0402756929397583, "learning_rate": 4.1349934469200526e-05, "loss": 1.2451, "step": 1810 }, { "epoch": 2.3565387117761873, "grad_norm": 1.1561073064804077, "learning_rate": 4.1317169069462644e-05, "loss": 0.9042, "step": 1811 }, { "epoch": 2.357839947950553, "grad_norm": 1.5316799879074097, "learning_rate": 4.1284403669724776e-05, "loss": 0.9225, "step": 1812 }, { "epoch": 2.3591411841249186, "grad_norm": 1.3471319675445557, "learning_rate": 4.12516382699869e-05, "loss": 1.0551, "step": 1813 }, { "epoch": 2.360442420299284, "grad_norm": 1.084571361541748, "learning_rate": 4.121887287024902e-05, "loss": 0.7908, "step": 1814 }, { "epoch": 2.36174365647365, "grad_norm": 1.2899832725524902, "learning_rate": 4.118610747051114e-05, "loss": 0.8273, "step": 1815 }, { "epoch": 2.3630448926480154, "grad_norm": 1.5558810234069824, "learning_rate": 4.115334207077327e-05, "loss": 0.8402, "step": 1816 }, { "epoch": 2.3643461288223815, "grad_norm": 0.9348979592323303, "learning_rate": 4.1120576671035386e-05, "loss": 1.673, "step": 1817 }, { "epoch": 2.3656473649967467, "grad_norm": 1.0286372900009155, "learning_rate": 4.108781127129751e-05, "loss": 1.3562, "step": 1818 }, { "epoch": 2.3669486011711127, "grad_norm": 1.750483751296997, "learning_rate": 4.1055045871559636e-05, "loss": 0.8061, "step": 1819 }, { "epoch": 2.3682498373454783, "grad_norm": 1.068734049797058, "learning_rate": 4.1022280471821753e-05, "loss": 0.8352, "step": 1820 }, { "epoch": 2.369551073519844, "grad_norm": 0.9504371285438538, "learning_rate": 4.0989515072083885e-05, "loss": 1.3832, "step": 1821 }, { "epoch": 2.3708523096942096, "grad_norm": 1.0386601686477661, "learning_rate": 4.095674967234601e-05, "loss": 1.2927, "step": 1822 }, { "epoch": 2.372153545868575, "grad_norm": 1.1253865957260132, "learning_rate": 4.092398427260813e-05, "loss": 1.2352, "step": 1823 }, { "epoch": 2.373454782042941, "grad_norm": 1.002652645111084, "learning_rate": 4.089121887287025e-05, "loss": 1.4261, "step": 1824 }, { "epoch": 2.3747560182173064, "grad_norm": 0.80739426612854, "learning_rate": 4.085845347313238e-05, "loss": 1.0633, "step": 1825 }, { "epoch": 2.376057254391672, "grad_norm": 1.1164456605911255, "learning_rate": 4.0825688073394495e-05, "loss": 0.7312, "step": 1826 }, { "epoch": 2.3773584905660377, "grad_norm": 1.2889410257339478, "learning_rate": 4.079292267365662e-05, "loss": 0.9099, "step": 1827 }, { "epoch": 2.3786597267404033, "grad_norm": 1.1555187702178955, "learning_rate": 4.0760157273918745e-05, "loss": 0.7491, "step": 1828 }, { "epoch": 2.379960962914769, "grad_norm": 1.1549766063690186, "learning_rate": 4.072739187418086e-05, "loss": 1.0859, "step": 1829 }, { "epoch": 2.3812621990891345, "grad_norm": 1.4932876825332642, "learning_rate": 4.0694626474442994e-05, "loss": 0.9967, "step": 1830 }, { "epoch": 2.3825634352635, "grad_norm": 1.2883204221725464, "learning_rate": 4.066186107470511e-05, "loss": 1.0655, "step": 1831 }, { "epoch": 2.383864671437866, "grad_norm": 1.3616849184036255, "learning_rate": 4.062909567496724e-05, "loss": 0.8826, "step": 1832 }, { "epoch": 2.385165907612232, "grad_norm": 1.1421334743499756, "learning_rate": 4.059633027522936e-05, "loss": 1.3362, "step": 1833 }, { "epoch": 2.3864671437865974, "grad_norm": 1.158759355545044, "learning_rate": 4.0563564875491486e-05, "loss": 1.1242, "step": 1834 }, { "epoch": 2.387768379960963, "grad_norm": 1.0731501579284668, "learning_rate": 4.0530799475753604e-05, "loss": 1.3013, "step": 1835 }, { "epoch": 2.3890696161353286, "grad_norm": 1.1395175457000732, "learning_rate": 4.049803407601573e-05, "loss": 1.3764, "step": 1836 }, { "epoch": 2.3903708523096943, "grad_norm": 0.9518351554870605, "learning_rate": 4.0465268676277854e-05, "loss": 1.1715, "step": 1837 }, { "epoch": 2.39167208848406, "grad_norm": 0.8976961970329285, "learning_rate": 4.043250327653997e-05, "loss": 1.4038, "step": 1838 }, { "epoch": 2.3929733246584255, "grad_norm": 1.272120714187622, "learning_rate": 4.0399737876802096e-05, "loss": 1.1653, "step": 1839 }, { "epoch": 2.394274560832791, "grad_norm": 1.177999496459961, "learning_rate": 4.036697247706422e-05, "loss": 1.0589, "step": 1840 }, { "epoch": 2.3955757970071567, "grad_norm": 1.3012887239456177, "learning_rate": 4.0334207077326346e-05, "loss": 1.1362, "step": 1841 }, { "epoch": 2.3968770331815223, "grad_norm": 1.0056843757629395, "learning_rate": 4.030144167758847e-05, "loss": 1.3754, "step": 1842 }, { "epoch": 2.398178269355888, "grad_norm": 0.9843177795410156, "learning_rate": 4.0268676277850595e-05, "loss": 1.2776, "step": 1843 }, { "epoch": 2.3994795055302536, "grad_norm": 1.0729244947433472, "learning_rate": 4.023591087811271e-05, "loss": 1.0969, "step": 1844 }, { "epoch": 2.4007807417046196, "grad_norm": 0.7645602226257324, "learning_rate": 4.020314547837484e-05, "loss": 0.7659, "step": 1845 }, { "epoch": 2.402081977878985, "grad_norm": 1.212659478187561, "learning_rate": 4.017038007863696e-05, "loss": 0.9054, "step": 1846 }, { "epoch": 2.403383214053351, "grad_norm": 1.37046217918396, "learning_rate": 4.013761467889908e-05, "loss": 0.7727, "step": 1847 }, { "epoch": 2.4046844502277165, "grad_norm": 1.4522672891616821, "learning_rate": 4.0104849279161205e-05, "loss": 0.5741, "step": 1848 }, { "epoch": 2.405985686402082, "grad_norm": 1.285499095916748, "learning_rate": 4.007208387942333e-05, "loss": 1.0297, "step": 1849 }, { "epoch": 2.4072869225764477, "grad_norm": 1.553513526916504, "learning_rate": 4.0039318479685455e-05, "loss": 0.9725, "step": 1850 }, { "epoch": 2.4085881587508133, "grad_norm": 1.356145977973938, "learning_rate": 4.000655307994758e-05, "loss": 1.0951, "step": 1851 }, { "epoch": 2.409889394925179, "grad_norm": 1.2054119110107422, "learning_rate": 3.9973787680209704e-05, "loss": 1.3735, "step": 1852 }, { "epoch": 2.4111906310995446, "grad_norm": 1.2151093482971191, "learning_rate": 3.994102228047182e-05, "loss": 1.1236, "step": 1853 }, { "epoch": 2.41249186727391, "grad_norm": 1.1917223930358887, "learning_rate": 3.990825688073395e-05, "loss": 0.8762, "step": 1854 }, { "epoch": 2.413793103448276, "grad_norm": 1.2401689291000366, "learning_rate": 3.987549148099607e-05, "loss": 1.1045, "step": 1855 }, { "epoch": 2.4150943396226414, "grad_norm": 1.0950860977172852, "learning_rate": 3.984272608125819e-05, "loss": 1.0831, "step": 1856 }, { "epoch": 2.416395575797007, "grad_norm": 1.3500149250030518, "learning_rate": 3.9809960681520315e-05, "loss": 1.3168, "step": 1857 }, { "epoch": 2.4176968119713727, "grad_norm": 1.0486741065979004, "learning_rate": 3.977719528178244e-05, "loss": 1.4456, "step": 1858 }, { "epoch": 2.4189980481457383, "grad_norm": 1.1053255796432495, "learning_rate": 3.9744429882044564e-05, "loss": 1.0917, "step": 1859 }, { "epoch": 2.4202992843201043, "grad_norm": 1.2364349365234375, "learning_rate": 3.971166448230669e-05, "loss": 0.9363, "step": 1860 }, { "epoch": 2.4216005204944695, "grad_norm": 1.1455919742584229, "learning_rate": 3.967889908256881e-05, "loss": 1.0368, "step": 1861 }, { "epoch": 2.4229017566688356, "grad_norm": 1.0618627071380615, "learning_rate": 3.964613368283093e-05, "loss": 0.8434, "step": 1862 }, { "epoch": 2.424202992843201, "grad_norm": 0.9991618394851685, "learning_rate": 3.9613368283093056e-05, "loss": 0.8725, "step": 1863 }, { "epoch": 2.425504229017567, "grad_norm": 1.205911636352539, "learning_rate": 3.958060288335518e-05, "loss": 0.8699, "step": 1864 }, { "epoch": 2.4268054651919324, "grad_norm": 1.1527252197265625, "learning_rate": 3.95478374836173e-05, "loss": 1.1017, "step": 1865 }, { "epoch": 2.428106701366298, "grad_norm": 1.1253108978271484, "learning_rate": 3.9515072083879424e-05, "loss": 1.2178, "step": 1866 }, { "epoch": 2.4294079375406636, "grad_norm": 1.3172224760055542, "learning_rate": 3.948230668414155e-05, "loss": 1.0907, "step": 1867 }, { "epoch": 2.4307091737150293, "grad_norm": 1.3096762895584106, "learning_rate": 3.944954128440367e-05, "loss": 0.5436, "step": 1868 }, { "epoch": 2.432010409889395, "grad_norm": 1.0773050785064697, "learning_rate": 3.94167758846658e-05, "loss": 0.8623, "step": 1869 }, { "epoch": 2.4333116460637605, "grad_norm": 0.8582805395126343, "learning_rate": 3.9384010484927916e-05, "loss": 0.9384, "step": 1870 }, { "epoch": 2.434612882238126, "grad_norm": 1.045179009437561, "learning_rate": 3.935124508519004e-05, "loss": 0.9971, "step": 1871 }, { "epoch": 2.4359141184124917, "grad_norm": 1.3450506925582886, "learning_rate": 3.9318479685452165e-05, "loss": 1.3378, "step": 1872 }, { "epoch": 2.4372153545868573, "grad_norm": 1.1127147674560547, "learning_rate": 3.928571428571429e-05, "loss": 0.9364, "step": 1873 }, { "epoch": 2.438516590761223, "grad_norm": 1.1713413000106812, "learning_rate": 3.925294888597641e-05, "loss": 1.1938, "step": 1874 }, { "epoch": 2.439817826935589, "grad_norm": 1.4744749069213867, "learning_rate": 3.922018348623853e-05, "loss": 0.9216, "step": 1875 }, { "epoch": 2.4411190631099546, "grad_norm": 1.2448300123214722, "learning_rate": 3.918741808650066e-05, "loss": 0.9537, "step": 1876 }, { "epoch": 2.4424202992843203, "grad_norm": 1.0750465393066406, "learning_rate": 3.915465268676278e-05, "loss": 1.4554, "step": 1877 }, { "epoch": 2.443721535458686, "grad_norm": 0.9937528967857361, "learning_rate": 3.912188728702491e-05, "loss": 1.283, "step": 1878 }, { "epoch": 2.4450227716330515, "grad_norm": 0.7542230486869812, "learning_rate": 3.9089121887287025e-05, "loss": 2.0273, "step": 1879 }, { "epoch": 2.446324007807417, "grad_norm": 1.230617880821228, "learning_rate": 3.905635648754915e-05, "loss": 1.0741, "step": 1880 }, { "epoch": 2.4476252439817827, "grad_norm": 1.2924034595489502, "learning_rate": 3.9023591087811274e-05, "loss": 1.0489, "step": 1881 }, { "epoch": 2.4489264801561483, "grad_norm": 0.9558413028717041, "learning_rate": 3.89908256880734e-05, "loss": 1.5548, "step": 1882 }, { "epoch": 2.450227716330514, "grad_norm": 1.0637074708938599, "learning_rate": 3.895806028833552e-05, "loss": 1.3761, "step": 1883 }, { "epoch": 2.4515289525048796, "grad_norm": 1.3731704950332642, "learning_rate": 3.892529488859764e-05, "loss": 1.3637, "step": 1884 }, { "epoch": 2.452830188679245, "grad_norm": 1.3273123502731323, "learning_rate": 3.8892529488859767e-05, "loss": 1.2501, "step": 1885 }, { "epoch": 2.454131424853611, "grad_norm": 1.4137054681777954, "learning_rate": 3.885976408912189e-05, "loss": 0.5314, "step": 1886 }, { "epoch": 2.4554326610279764, "grad_norm": 0.9831037521362305, "learning_rate": 3.8826998689384016e-05, "loss": 0.6303, "step": 1887 }, { "epoch": 2.456733897202342, "grad_norm": 0.9975313544273376, "learning_rate": 3.8794233289646134e-05, "loss": 1.0512, "step": 1888 }, { "epoch": 2.4580351333767076, "grad_norm": 1.4761462211608887, "learning_rate": 3.876146788990826e-05, "loss": 1.0545, "step": 1889 }, { "epoch": 2.4593363695510737, "grad_norm": 1.167080044746399, "learning_rate": 3.8728702490170383e-05, "loss": 0.8901, "step": 1890 }, { "epoch": 2.4606376057254393, "grad_norm": 1.254652738571167, "learning_rate": 3.86959370904325e-05, "loss": 1.3594, "step": 1891 }, { "epoch": 2.461938841899805, "grad_norm": 1.136351227760315, "learning_rate": 3.8663171690694626e-05, "loss": 0.7384, "step": 1892 }, { "epoch": 2.4632400780741706, "grad_norm": 1.1121541261672974, "learning_rate": 3.863040629095675e-05, "loss": 0.7582, "step": 1893 }, { "epoch": 2.464541314248536, "grad_norm": 1.3438894748687744, "learning_rate": 3.8597640891218876e-05, "loss": 0.9329, "step": 1894 }, { "epoch": 2.465842550422902, "grad_norm": 1.1259146928787231, "learning_rate": 3.8564875491481e-05, "loss": 1.175, "step": 1895 }, { "epoch": 2.4671437865972674, "grad_norm": 1.1184132099151611, "learning_rate": 3.8532110091743125e-05, "loss": 1.1582, "step": 1896 }, { "epoch": 2.468445022771633, "grad_norm": 1.1729152202606201, "learning_rate": 3.849934469200524e-05, "loss": 1.0992, "step": 1897 }, { "epoch": 2.4697462589459986, "grad_norm": 0.9225536584854126, "learning_rate": 3.846657929226737e-05, "loss": 0.635, "step": 1898 }, { "epoch": 2.4710474951203643, "grad_norm": 1.5187641382217407, "learning_rate": 3.843381389252949e-05, "loss": 0.5542, "step": 1899 }, { "epoch": 2.47234873129473, "grad_norm": 1.497371792793274, "learning_rate": 3.840104849279161e-05, "loss": 0.8927, "step": 1900 }, { "epoch": 2.4736499674690955, "grad_norm": 1.2363369464874268, "learning_rate": 3.8368283093053735e-05, "loss": 0.8583, "step": 1901 }, { "epoch": 2.474951203643461, "grad_norm": 1.136533260345459, "learning_rate": 3.833551769331586e-05, "loss": 1.3436, "step": 1902 }, { "epoch": 2.476252439817827, "grad_norm": 1.2752931118011475, "learning_rate": 3.8302752293577985e-05, "loss": 0.8833, "step": 1903 }, { "epoch": 2.4775536759921923, "grad_norm": 1.1671394109725952, "learning_rate": 3.82699868938401e-05, "loss": 0.7022, "step": 1904 }, { "epoch": 2.4788549121665584, "grad_norm": 0.7025185823440552, "learning_rate": 3.8237221494102234e-05, "loss": 0.8721, "step": 1905 }, { "epoch": 2.480156148340924, "grad_norm": 1.128525972366333, "learning_rate": 3.820445609436435e-05, "loss": 0.9457, "step": 1906 }, { "epoch": 2.4814573845152896, "grad_norm": 1.0060700178146362, "learning_rate": 3.817169069462648e-05, "loss": 1.6641, "step": 1907 }, { "epoch": 2.4827586206896552, "grad_norm": 1.2221404314041138, "learning_rate": 3.81389252948886e-05, "loss": 1.2035, "step": 1908 }, { "epoch": 2.484059856864021, "grad_norm": 1.341734528541565, "learning_rate": 3.810615989515072e-05, "loss": 0.898, "step": 1909 }, { "epoch": 2.4853610930383865, "grad_norm": 1.221627950668335, "learning_rate": 3.8073394495412844e-05, "loss": 1.0811, "step": 1910 }, { "epoch": 2.486662329212752, "grad_norm": 1.7288252115249634, "learning_rate": 3.804062909567497e-05, "loss": 1.1079, "step": 1911 }, { "epoch": 2.4879635653871177, "grad_norm": 0.9846968650817871, "learning_rate": 3.8007863695937094e-05, "loss": 0.8314, "step": 1912 }, { "epoch": 2.4892648015614833, "grad_norm": 1.127299427986145, "learning_rate": 3.797509829619921e-05, "loss": 1.0824, "step": 1913 }, { "epoch": 2.490566037735849, "grad_norm": 1.376653790473938, "learning_rate": 3.794233289646134e-05, "loss": 0.6992, "step": 1914 }, { "epoch": 2.4918672739102146, "grad_norm": 1.1341685056686401, "learning_rate": 3.790956749672346e-05, "loss": 0.776, "step": 1915 }, { "epoch": 2.49316851008458, "grad_norm": 1.3726637363433838, "learning_rate": 3.7876802096985586e-05, "loss": 0.8626, "step": 1916 }, { "epoch": 2.494469746258946, "grad_norm": 0.9665217995643616, "learning_rate": 3.784403669724771e-05, "loss": 1.0408, "step": 1917 }, { "epoch": 2.495770982433312, "grad_norm": 1.0472183227539062, "learning_rate": 3.781127129750983e-05, "loss": 1.2511, "step": 1918 }, { "epoch": 2.4970722186076775, "grad_norm": 0.9349297285079956, "learning_rate": 3.7778505897771953e-05, "loss": 0.9029, "step": 1919 }, { "epoch": 2.498373454782043, "grad_norm": 1.110899806022644, "learning_rate": 3.774574049803408e-05, "loss": 0.7061, "step": 1920 }, { "epoch": 2.4996746909564087, "grad_norm": 1.1135655641555786, "learning_rate": 3.7712975098296196e-05, "loss": 0.579, "step": 1921 }, { "epoch": 2.5009759271307743, "grad_norm": 1.2051458358764648, "learning_rate": 3.768020969855832e-05, "loss": 1.3031, "step": 1922 }, { "epoch": 2.50227716330514, "grad_norm": 1.2863234281539917, "learning_rate": 3.764744429882045e-05, "loss": 0.8263, "step": 1923 }, { "epoch": 2.5035783994795056, "grad_norm": 1.3098443746566772, "learning_rate": 3.761467889908257e-05, "loss": 0.796, "step": 1924 }, { "epoch": 2.504879635653871, "grad_norm": 1.2931585311889648, "learning_rate": 3.7581913499344695e-05, "loss": 0.8598, "step": 1925 }, { "epoch": 2.506180871828237, "grad_norm": 1.065260648727417, "learning_rate": 3.754914809960682e-05, "loss": 1.0835, "step": 1926 }, { "epoch": 2.5074821080026024, "grad_norm": 1.0347198247909546, "learning_rate": 3.751638269986894e-05, "loss": 1.2784, "step": 1927 }, { "epoch": 2.508783344176968, "grad_norm": 1.0342434644699097, "learning_rate": 3.748361730013106e-05, "loss": 0.985, "step": 1928 }, { "epoch": 2.5100845803513336, "grad_norm": 1.4771910905838013, "learning_rate": 3.745085190039319e-05, "loss": 0.6737, "step": 1929 }, { "epoch": 2.5113858165256993, "grad_norm": 1.1593190431594849, "learning_rate": 3.7418086500655305e-05, "loss": 0.7206, "step": 1930 }, { "epoch": 2.5126870527000653, "grad_norm": 1.424752116203308, "learning_rate": 3.738532110091743e-05, "loss": 1.468, "step": 1931 }, { "epoch": 2.5139882888744305, "grad_norm": 1.6473464965820312, "learning_rate": 3.7352555701179555e-05, "loss": 0.7725, "step": 1932 }, { "epoch": 2.5152895250487965, "grad_norm": 1.433821439743042, "learning_rate": 3.731979030144168e-05, "loss": 0.6462, "step": 1933 }, { "epoch": 2.516590761223162, "grad_norm": 1.1487373113632202, "learning_rate": 3.7287024901703804e-05, "loss": 1.5275, "step": 1934 }, { "epoch": 2.517891997397528, "grad_norm": 1.3188120126724243, "learning_rate": 3.725425950196593e-05, "loss": 1.1728, "step": 1935 }, { "epoch": 2.5191932335718934, "grad_norm": 1.0525527000427246, "learning_rate": 3.722149410222805e-05, "loss": 1.4128, "step": 1936 }, { "epoch": 2.520494469746259, "grad_norm": 0.881142258644104, "learning_rate": 3.718872870249017e-05, "loss": 1.1011, "step": 1937 }, { "epoch": 2.5217957059206246, "grad_norm": 1.0992200374603271, "learning_rate": 3.7155963302752296e-05, "loss": 0.8625, "step": 1938 }, { "epoch": 2.5230969420949902, "grad_norm": 1.0898302793502808, "learning_rate": 3.7123197903014414e-05, "loss": 1.1249, "step": 1939 }, { "epoch": 2.524398178269356, "grad_norm": 1.0134360790252686, "learning_rate": 3.709043250327654e-05, "loss": 0.9354, "step": 1940 }, { "epoch": 2.5256994144437215, "grad_norm": 1.3872257471084595, "learning_rate": 3.7057667103538664e-05, "loss": 0.4306, "step": 1941 }, { "epoch": 2.527000650618087, "grad_norm": 0.9058359265327454, "learning_rate": 3.702490170380079e-05, "loss": 1.0873, "step": 1942 }, { "epoch": 2.5283018867924527, "grad_norm": 1.108252763748169, "learning_rate": 3.699213630406291e-05, "loss": 1.3743, "step": 1943 }, { "epoch": 2.5296031229668183, "grad_norm": 0.9381688833236694, "learning_rate": 3.695937090432504e-05, "loss": 0.8334, "step": 1944 }, { "epoch": 2.530904359141184, "grad_norm": 1.112039566040039, "learning_rate": 3.6926605504587156e-05, "loss": 1.0142, "step": 1945 }, { "epoch": 2.53220559531555, "grad_norm": 0.7390649318695068, "learning_rate": 3.689384010484928e-05, "loss": 1.0794, "step": 1946 }, { "epoch": 2.533506831489915, "grad_norm": 1.3483504056930542, "learning_rate": 3.6861074705111405e-05, "loss": 1.5479, "step": 1947 }, { "epoch": 2.5348080676642812, "grad_norm": 1.6963646411895752, "learning_rate": 3.682830930537352e-05, "loss": 1.3761, "step": 1948 }, { "epoch": 2.536109303838647, "grad_norm": 1.6242074966430664, "learning_rate": 3.679554390563565e-05, "loss": 0.9134, "step": 1949 }, { "epoch": 2.5374105400130125, "grad_norm": 1.3646211624145508, "learning_rate": 3.676277850589777e-05, "loss": 0.9607, "step": 1950 }, { "epoch": 2.538711776187378, "grad_norm": 1.1634751558303833, "learning_rate": 3.67300131061599e-05, "loss": 1.3144, "step": 1951 }, { "epoch": 2.5400130123617437, "grad_norm": 1.404241681098938, "learning_rate": 3.669724770642202e-05, "loss": 0.8019, "step": 1952 }, { "epoch": 2.5413142485361093, "grad_norm": 1.0854400396347046, "learning_rate": 3.666448230668415e-05, "loss": 1.0863, "step": 1953 }, { "epoch": 2.542615484710475, "grad_norm": 1.0845489501953125, "learning_rate": 3.6631716906946265e-05, "loss": 1.2351, "step": 1954 }, { "epoch": 2.5439167208848406, "grad_norm": 1.0844857692718506, "learning_rate": 3.659895150720839e-05, "loss": 0.6959, "step": 1955 }, { "epoch": 2.545217957059206, "grad_norm": 1.157378077507019, "learning_rate": 3.6566186107470514e-05, "loss": 0.7122, "step": 1956 }, { "epoch": 2.546519193233572, "grad_norm": 1.123293161392212, "learning_rate": 3.653342070773263e-05, "loss": 1.1736, "step": 1957 }, { "epoch": 2.5478204294079374, "grad_norm": 1.3712437152862549, "learning_rate": 3.650065530799476e-05, "loss": 1.0848, "step": 1958 }, { "epoch": 2.5491216655823035, "grad_norm": 1.1131659746170044, "learning_rate": 3.646788990825688e-05, "loss": 1.0405, "step": 1959 }, { "epoch": 2.5504229017566686, "grad_norm": 1.405646562576294, "learning_rate": 3.6435124508519e-05, "loss": 0.7179, "step": 1960 }, { "epoch": 2.5517241379310347, "grad_norm": 1.285691738128662, "learning_rate": 3.640235910878113e-05, "loss": 1.265, "step": 1961 }, { "epoch": 2.5530253741054, "grad_norm": 0.8229387998580933, "learning_rate": 3.6369593709043256e-05, "loss": 0.904, "step": 1962 }, { "epoch": 2.554326610279766, "grad_norm": 0.9654499292373657, "learning_rate": 3.6336828309305374e-05, "loss": 1.7129, "step": 1963 }, { "epoch": 2.5556278464541315, "grad_norm": 1.060847520828247, "learning_rate": 3.63040629095675e-05, "loss": 0.9079, "step": 1964 }, { "epoch": 2.556929082628497, "grad_norm": 1.2439122200012207, "learning_rate": 3.6271297509829624e-05, "loss": 1.3905, "step": 1965 }, { "epoch": 2.558230318802863, "grad_norm": 0.9342418909072876, "learning_rate": 3.623853211009174e-05, "loss": 1.3283, "step": 1966 }, { "epoch": 2.5595315549772284, "grad_norm": 1.129940152168274, "learning_rate": 3.6205766710353866e-05, "loss": 1.1314, "step": 1967 }, { "epoch": 2.560832791151594, "grad_norm": 1.1463640928268433, "learning_rate": 3.617300131061599e-05, "loss": 1.4915, "step": 1968 }, { "epoch": 2.5621340273259596, "grad_norm": 0.8832682371139526, "learning_rate": 3.614023591087811e-05, "loss": 0.9415, "step": 1969 }, { "epoch": 2.5634352635003252, "grad_norm": 0.8938408493995667, "learning_rate": 3.610747051114024e-05, "loss": 1.3953, "step": 1970 }, { "epoch": 2.564736499674691, "grad_norm": 1.257143259048462, "learning_rate": 3.6074705111402365e-05, "loss": 1.0238, "step": 1971 }, { "epoch": 2.5660377358490565, "grad_norm": 1.504628300666809, "learning_rate": 3.604193971166448e-05, "loss": 0.7041, "step": 1972 }, { "epoch": 2.567338972023422, "grad_norm": 1.3086888790130615, "learning_rate": 3.600917431192661e-05, "loss": 1.018, "step": 1973 }, { "epoch": 2.568640208197788, "grad_norm": 1.0367687940597534, "learning_rate": 3.597640891218873e-05, "loss": 1.7309, "step": 1974 }, { "epoch": 2.5699414443721533, "grad_norm": 1.5148067474365234, "learning_rate": 3.594364351245085e-05, "loss": 0.6701, "step": 1975 }, { "epoch": 2.5712426805465194, "grad_norm": 1.0701147317886353, "learning_rate": 3.5910878112712975e-05, "loss": 0.52, "step": 1976 }, { "epoch": 2.572543916720885, "grad_norm": 1.1730605363845825, "learning_rate": 3.58781127129751e-05, "loss": 0.9746, "step": 1977 }, { "epoch": 2.5738451528952506, "grad_norm": 0.9536253809928894, "learning_rate": 3.584534731323722e-05, "loss": 1.1239, "step": 1978 }, { "epoch": 2.5751463890696162, "grad_norm": 1.1808866262435913, "learning_rate": 3.581258191349935e-05, "loss": 0.5456, "step": 1979 }, { "epoch": 2.576447625243982, "grad_norm": 1.18390691280365, "learning_rate": 3.5779816513761474e-05, "loss": 1.0348, "step": 1980 }, { "epoch": 2.5777488614183475, "grad_norm": 1.222611665725708, "learning_rate": 3.574705111402359e-05, "loss": 0.9112, "step": 1981 }, { "epoch": 2.579050097592713, "grad_norm": 1.1183300018310547, "learning_rate": 3.571428571428572e-05, "loss": 1.5222, "step": 1982 }, { "epoch": 2.5803513337670787, "grad_norm": 1.2456930875778198, "learning_rate": 3.568152031454784e-05, "loss": 1.2757, "step": 1983 }, { "epoch": 2.5816525699414443, "grad_norm": 1.1919801235198975, "learning_rate": 3.564875491480996e-05, "loss": 0.7546, "step": 1984 }, { "epoch": 2.58295380611581, "grad_norm": 1.1399741172790527, "learning_rate": 3.5615989515072084e-05, "loss": 0.5745, "step": 1985 }, { "epoch": 2.5842550422901756, "grad_norm": 1.1720855236053467, "learning_rate": 3.558322411533421e-05, "loss": 1.4397, "step": 1986 }, { "epoch": 2.585556278464541, "grad_norm": 1.0394260883331299, "learning_rate": 3.555045871559633e-05, "loss": 1.0015, "step": 1987 }, { "epoch": 2.586857514638907, "grad_norm": 1.0666550397872925, "learning_rate": 3.551769331585846e-05, "loss": 1.1074, "step": 1988 }, { "epoch": 2.588158750813273, "grad_norm": 1.2836148738861084, "learning_rate": 3.548492791612058e-05, "loss": 0.8126, "step": 1989 }, { "epoch": 2.589459986987638, "grad_norm": 1.0711325407028198, "learning_rate": 3.54521625163827e-05, "loss": 1.3418, "step": 1990 }, { "epoch": 2.590761223162004, "grad_norm": 1.346846103668213, "learning_rate": 3.5419397116644826e-05, "loss": 0.5795, "step": 1991 }, { "epoch": 2.5920624593363697, "grad_norm": 1.0863971710205078, "learning_rate": 3.538663171690695e-05, "loss": 1.3368, "step": 1992 }, { "epoch": 2.5933636955107353, "grad_norm": 1.2654175758361816, "learning_rate": 3.535386631716907e-05, "loss": 0.9125, "step": 1993 }, { "epoch": 2.594664931685101, "grad_norm": 1.1332913637161255, "learning_rate": 3.5321100917431193e-05, "loss": 1.0901, "step": 1994 }, { "epoch": 2.5959661678594665, "grad_norm": 1.3070154190063477, "learning_rate": 3.528833551769332e-05, "loss": 0.4847, "step": 1995 }, { "epoch": 2.597267404033832, "grad_norm": 1.1136353015899658, "learning_rate": 3.5255570117955436e-05, "loss": 1.4921, "step": 1996 }, { "epoch": 2.5985686402081978, "grad_norm": 1.247953176498413, "learning_rate": 3.522280471821756e-05, "loss": 1.0304, "step": 1997 }, { "epoch": 2.5998698763825634, "grad_norm": 1.032402753829956, "learning_rate": 3.519003931847969e-05, "loss": 1.0635, "step": 1998 }, { "epoch": 2.601171112556929, "grad_norm": 0.7618741989135742, "learning_rate": 3.515727391874181e-05, "loss": 1.7821, "step": 1999 }, { "epoch": 2.6024723487312946, "grad_norm": 1.0860483646392822, "learning_rate": 3.5124508519003935e-05, "loss": 0.6093, "step": 2000 }, { "epoch": 2.6037735849056602, "grad_norm": 1.2554075717926025, "learning_rate": 3.509174311926606e-05, "loss": 1.2915, "step": 2001 }, { "epoch": 2.605074821080026, "grad_norm": 0.898199737071991, "learning_rate": 3.505897771952818e-05, "loss": 1.3952, "step": 2002 }, { "epoch": 2.6063760572543915, "grad_norm": 1.2178093194961548, "learning_rate": 3.50262123197903e-05, "loss": 1.3336, "step": 2003 }, { "epoch": 2.6076772934287575, "grad_norm": 0.9599814414978027, "learning_rate": 3.499344692005243e-05, "loss": 1.2034, "step": 2004 }, { "epoch": 2.6089785296031227, "grad_norm": 1.4260685443878174, "learning_rate": 3.4960681520314545e-05, "loss": 0.8004, "step": 2005 }, { "epoch": 2.6102797657774888, "grad_norm": 1.177470088005066, "learning_rate": 3.492791612057667e-05, "loss": 0.8554, "step": 2006 }, { "epoch": 2.6115810019518544, "grad_norm": 1.2614834308624268, "learning_rate": 3.48951507208388e-05, "loss": 0.8947, "step": 2007 }, { "epoch": 2.61288223812622, "grad_norm": 1.2721621990203857, "learning_rate": 3.486238532110092e-05, "loss": 0.9719, "step": 2008 }, { "epoch": 2.6141834743005856, "grad_norm": 1.253879189491272, "learning_rate": 3.4829619921363044e-05, "loss": 0.7004, "step": 2009 }, { "epoch": 2.6154847104749512, "grad_norm": 1.340600848197937, "learning_rate": 3.479685452162517e-05, "loss": 1.0006, "step": 2010 }, { "epoch": 2.616785946649317, "grad_norm": 1.1811487674713135, "learning_rate": 3.476408912188729e-05, "loss": 1.3291, "step": 2011 }, { "epoch": 2.6180871828236825, "grad_norm": 1.086012840270996, "learning_rate": 3.473132372214941e-05, "loss": 0.8727, "step": 2012 }, { "epoch": 2.619388418998048, "grad_norm": 1.4314385652542114, "learning_rate": 3.4698558322411536e-05, "loss": 1.019, "step": 2013 }, { "epoch": 2.6206896551724137, "grad_norm": 1.397735595703125, "learning_rate": 3.4665792922673654e-05, "loss": 0.8527, "step": 2014 }, { "epoch": 2.6219908913467793, "grad_norm": 1.115736484527588, "learning_rate": 3.463302752293578e-05, "loss": 1.5409, "step": 2015 }, { "epoch": 2.623292127521145, "grad_norm": 0.8748232126235962, "learning_rate": 3.460026212319791e-05, "loss": 0.9285, "step": 2016 }, { "epoch": 2.624593363695511, "grad_norm": 1.115389347076416, "learning_rate": 3.456749672346003e-05, "loss": 1.4027, "step": 2017 }, { "epoch": 2.625894599869876, "grad_norm": 1.2905304431915283, "learning_rate": 3.453473132372215e-05, "loss": 1.1785, "step": 2018 }, { "epoch": 2.6271958360442422, "grad_norm": 0.9509730339050293, "learning_rate": 3.450196592398428e-05, "loss": 0.9639, "step": 2019 }, { "epoch": 2.6284970722186074, "grad_norm": 1.1004712581634521, "learning_rate": 3.4469200524246396e-05, "loss": 0.6393, "step": 2020 }, { "epoch": 2.6297983083929735, "grad_norm": 0.9533021450042725, "learning_rate": 3.443643512450852e-05, "loss": 1.4032, "step": 2021 }, { "epoch": 2.631099544567339, "grad_norm": 1.4927071332931519, "learning_rate": 3.4403669724770645e-05, "loss": 0.8365, "step": 2022 }, { "epoch": 2.6324007807417047, "grad_norm": 1.5409412384033203, "learning_rate": 3.4370904325032763e-05, "loss": 0.8869, "step": 2023 }, { "epoch": 2.6337020169160703, "grad_norm": 1.1253074407577515, "learning_rate": 3.433813892529489e-05, "loss": 1.1932, "step": 2024 }, { "epoch": 2.635003253090436, "grad_norm": 1.1176422834396362, "learning_rate": 3.430537352555701e-05, "loss": 0.9076, "step": 2025 }, { "epoch": 2.6363044892648015, "grad_norm": 0.8870770931243896, "learning_rate": 3.427260812581914e-05, "loss": 1.0753, "step": 2026 }, { "epoch": 2.637605725439167, "grad_norm": 1.2995634078979492, "learning_rate": 3.423984272608126e-05, "loss": 0.8025, "step": 2027 }, { "epoch": 2.6389069616135328, "grad_norm": 1.255460262298584, "learning_rate": 3.420707732634339e-05, "loss": 1.4171, "step": 2028 }, { "epoch": 2.6402081977878984, "grad_norm": 1.3300275802612305, "learning_rate": 3.4174311926605505e-05, "loss": 0.8482, "step": 2029 }, { "epoch": 2.641509433962264, "grad_norm": 1.4555188417434692, "learning_rate": 3.414154652686763e-05, "loss": 1.3181, "step": 2030 }, { "epoch": 2.6428106701366296, "grad_norm": 1.3153265714645386, "learning_rate": 3.4108781127129755e-05, "loss": 0.9528, "step": 2031 }, { "epoch": 2.6441119063109957, "grad_norm": 0.9900735020637512, "learning_rate": 3.407601572739187e-05, "loss": 0.8518, "step": 2032 }, { "epoch": 2.645413142485361, "grad_norm": 1.2589973211288452, "learning_rate": 3.4043250327654e-05, "loss": 1.1122, "step": 2033 }, { "epoch": 2.646714378659727, "grad_norm": 1.4206888675689697, "learning_rate": 3.401048492791612e-05, "loss": 0.7981, "step": 2034 }, { "epoch": 2.6480156148340925, "grad_norm": 1.01674222946167, "learning_rate": 3.397771952817825e-05, "loss": 1.3117, "step": 2035 }, { "epoch": 2.649316851008458, "grad_norm": 1.036507487297058, "learning_rate": 3.394495412844037e-05, "loss": 0.9932, "step": 2036 }, { "epoch": 2.6506180871828238, "grad_norm": 0.9165423512458801, "learning_rate": 3.3912188728702496e-05, "loss": 1.1513, "step": 2037 }, { "epoch": 2.6519193233571894, "grad_norm": 1.0603758096694946, "learning_rate": 3.3879423328964614e-05, "loss": 1.3645, "step": 2038 }, { "epoch": 2.653220559531555, "grad_norm": 1.0392175912857056, "learning_rate": 3.384665792922674e-05, "loss": 1.1642, "step": 2039 }, { "epoch": 2.6545217957059206, "grad_norm": 1.05869460105896, "learning_rate": 3.3813892529488864e-05, "loss": 1.1326, "step": 2040 }, { "epoch": 2.6558230318802862, "grad_norm": 0.5939114689826965, "learning_rate": 3.378112712975098e-05, "loss": 1.7274, "step": 2041 }, { "epoch": 2.657124268054652, "grad_norm": 1.2393815517425537, "learning_rate": 3.3748361730013106e-05, "loss": 1.2098, "step": 2042 }, { "epoch": 2.6584255042290175, "grad_norm": 0.8468376398086548, "learning_rate": 3.371559633027523e-05, "loss": 1.2331, "step": 2043 }, { "epoch": 2.659726740403383, "grad_norm": 1.3866099119186401, "learning_rate": 3.3682830930537356e-05, "loss": 0.7579, "step": 2044 }, { "epoch": 2.6610279765777487, "grad_norm": 0.8292623162269592, "learning_rate": 3.365006553079948e-05, "loss": 1.608, "step": 2045 }, { "epoch": 2.6623292127521143, "grad_norm": 1.079100489616394, "learning_rate": 3.3617300131061605e-05, "loss": 1.7062, "step": 2046 }, { "epoch": 2.6636304489264804, "grad_norm": 1.3594248294830322, "learning_rate": 3.358453473132372e-05, "loss": 1.0145, "step": 2047 }, { "epoch": 2.6649316851008455, "grad_norm": 1.0338618755340576, "learning_rate": 3.355176933158585e-05, "loss": 1.2691, "step": 2048 }, { "epoch": 2.6662329212752116, "grad_norm": 0.8165931701660156, "learning_rate": 3.351900393184797e-05, "loss": 0.4958, "step": 2049 }, { "epoch": 2.6675341574495772, "grad_norm": 1.3062494993209839, "learning_rate": 3.348623853211009e-05, "loss": 0.6997, "step": 2050 }, { "epoch": 2.668835393623943, "grad_norm": 1.0314134359359741, "learning_rate": 3.3453473132372215e-05, "loss": 0.923, "step": 2051 }, { "epoch": 2.6701366297983085, "grad_norm": 1.3506455421447754, "learning_rate": 3.342070773263434e-05, "loss": 1.2811, "step": 2052 }, { "epoch": 2.671437865972674, "grad_norm": 2.278610944747925, "learning_rate": 3.338794233289646e-05, "loss": 1.3188, "step": 2053 }, { "epoch": 2.6727391021470397, "grad_norm": 1.4417905807495117, "learning_rate": 3.335517693315859e-05, "loss": 0.844, "step": 2054 }, { "epoch": 2.6740403383214053, "grad_norm": 1.0965591669082642, "learning_rate": 3.3322411533420714e-05, "loss": 1.2033, "step": 2055 }, { "epoch": 2.675341574495771, "grad_norm": 1.3358672857284546, "learning_rate": 3.328964613368283e-05, "loss": 0.5347, "step": 2056 }, { "epoch": 2.6766428106701365, "grad_norm": 1.097642183303833, "learning_rate": 3.325688073394496e-05, "loss": 0.3328, "step": 2057 }, { "epoch": 2.677944046844502, "grad_norm": 1.049255609512329, "learning_rate": 3.322411533420708e-05, "loss": 0.7359, "step": 2058 }, { "epoch": 2.6792452830188678, "grad_norm": 1.3848555088043213, "learning_rate": 3.31913499344692e-05, "loss": 0.9534, "step": 2059 }, { "epoch": 2.680546519193234, "grad_norm": 1.589851975440979, "learning_rate": 3.3158584534731324e-05, "loss": 0.995, "step": 2060 }, { "epoch": 2.681847755367599, "grad_norm": 0.9914082288742065, "learning_rate": 3.312581913499345e-05, "loss": 1.4618, "step": 2061 }, { "epoch": 2.683148991541965, "grad_norm": 1.07041597366333, "learning_rate": 3.309305373525557e-05, "loss": 1.2787, "step": 2062 }, { "epoch": 2.6844502277163302, "grad_norm": 1.1813546419143677, "learning_rate": 3.30602883355177e-05, "loss": 1.3277, "step": 2063 }, { "epoch": 2.6857514638906963, "grad_norm": 1.3390944004058838, "learning_rate": 3.302752293577982e-05, "loss": 0.8391, "step": 2064 }, { "epoch": 2.687052700065062, "grad_norm": 0.9106911420822144, "learning_rate": 3.299475753604194e-05, "loss": 1.4433, "step": 2065 }, { "epoch": 2.6883539362394275, "grad_norm": 0.9453976154327393, "learning_rate": 3.2961992136304066e-05, "loss": 1.0904, "step": 2066 }, { "epoch": 2.689655172413793, "grad_norm": 0.9442291259765625, "learning_rate": 3.292922673656619e-05, "loss": 1.1438, "step": 2067 }, { "epoch": 2.6909564085881588, "grad_norm": 1.119746446609497, "learning_rate": 3.289646133682831e-05, "loss": 1.2716, "step": 2068 }, { "epoch": 2.6922576447625244, "grad_norm": 1.316210389137268, "learning_rate": 3.2863695937090434e-05, "loss": 0.647, "step": 2069 }, { "epoch": 2.69355888093689, "grad_norm": 1.2120968103408813, "learning_rate": 3.283093053735256e-05, "loss": 1.3589, "step": 2070 }, { "epoch": 2.6948601171112556, "grad_norm": 1.1146553754806519, "learning_rate": 3.2798165137614676e-05, "loss": 1.2379, "step": 2071 }, { "epoch": 2.6961613532856212, "grad_norm": 1.0254977941513062, "learning_rate": 3.276539973787681e-05, "loss": 1.6688, "step": 2072 }, { "epoch": 2.697462589459987, "grad_norm": 1.7204437255859375, "learning_rate": 3.2732634338138926e-05, "loss": 1.0969, "step": 2073 }, { "epoch": 2.6987638256343525, "grad_norm": 1.0769907236099243, "learning_rate": 3.269986893840105e-05, "loss": 1.0053, "step": 2074 }, { "epoch": 2.7000650618087185, "grad_norm": 1.0076085329055786, "learning_rate": 3.2667103538663175e-05, "loss": 1.3096, "step": 2075 }, { "epoch": 2.7013662979830837, "grad_norm": 1.1863735914230347, "learning_rate": 3.26343381389253e-05, "loss": 1.3027, "step": 2076 }, { "epoch": 2.7026675341574498, "grad_norm": 0.8250691294670105, "learning_rate": 3.260157273918742e-05, "loss": 1.0881, "step": 2077 }, { "epoch": 2.7039687703318154, "grad_norm": 1.449862003326416, "learning_rate": 3.256880733944954e-05, "loss": 0.7462, "step": 2078 }, { "epoch": 2.705270006506181, "grad_norm": 1.3153328895568848, "learning_rate": 3.253604193971167e-05, "loss": 1.5066, "step": 2079 }, { "epoch": 2.7065712426805466, "grad_norm": 1.3603192567825317, "learning_rate": 3.2503276539973785e-05, "loss": 0.9127, "step": 2080 }, { "epoch": 2.7078724788549122, "grad_norm": 1.0865672826766968, "learning_rate": 3.247051114023591e-05, "loss": 0.9097, "step": 2081 }, { "epoch": 2.709173715029278, "grad_norm": 1.6353363990783691, "learning_rate": 3.2437745740498035e-05, "loss": 1.0396, "step": 2082 }, { "epoch": 2.7104749512036435, "grad_norm": 1.168495774269104, "learning_rate": 3.240498034076016e-05, "loss": 0.8275, "step": 2083 }, { "epoch": 2.711776187378009, "grad_norm": 1.1853296756744385, "learning_rate": 3.2372214941022284e-05, "loss": 0.7737, "step": 2084 }, { "epoch": 2.7130774235523747, "grad_norm": 0.9267755150794983, "learning_rate": 3.23394495412844e-05, "loss": 1.1082, "step": 2085 }, { "epoch": 2.7143786597267403, "grad_norm": 0.8487802147865295, "learning_rate": 3.230668414154653e-05, "loss": 1.4655, "step": 2086 }, { "epoch": 2.715679895901106, "grad_norm": 1.1424003839492798, "learning_rate": 3.227391874180865e-05, "loss": 1.5648, "step": 2087 }, { "epoch": 2.7169811320754715, "grad_norm": 1.3779935836791992, "learning_rate": 3.2241153342070776e-05, "loss": 1.3213, "step": 2088 }, { "epoch": 2.718282368249837, "grad_norm": 1.3659793138504028, "learning_rate": 3.2208387942332894e-05, "loss": 1.2318, "step": 2089 }, { "epoch": 2.719583604424203, "grad_norm": 1.318729043006897, "learning_rate": 3.217562254259502e-05, "loss": 1.2164, "step": 2090 }, { "epoch": 2.7208848405985684, "grad_norm": 1.0752902030944824, "learning_rate": 3.2142857142857144e-05, "loss": 0.5349, "step": 2091 }, { "epoch": 2.7221860767729344, "grad_norm": 1.1986547708511353, "learning_rate": 3.211009174311927e-05, "loss": 1.0266, "step": 2092 }, { "epoch": 2.7234873129473, "grad_norm": 0.9414269924163818, "learning_rate": 3.207732634338139e-05, "loss": 1.668, "step": 2093 }, { "epoch": 2.7247885491216657, "grad_norm": 0.9947839379310608, "learning_rate": 3.204456094364351e-05, "loss": 1.6631, "step": 2094 }, { "epoch": 2.7260897852960313, "grad_norm": 2.825197458267212, "learning_rate": 3.2011795543905636e-05, "loss": 1.0529, "step": 2095 }, { "epoch": 2.727391021470397, "grad_norm": 0.9812507033348083, "learning_rate": 3.197903014416776e-05, "loss": 0.6838, "step": 2096 }, { "epoch": 2.7286922576447625, "grad_norm": 0.9684117436408997, "learning_rate": 3.1946264744429886e-05, "loss": 1.3735, "step": 2097 }, { "epoch": 2.729993493819128, "grad_norm": 1.3705201148986816, "learning_rate": 3.1913499344692004e-05, "loss": 1.018, "step": 2098 }, { "epoch": 2.7312947299934938, "grad_norm": 22.000720977783203, "learning_rate": 3.188073394495413e-05, "loss": 1.1236, "step": 2099 }, { "epoch": 2.7325959661678594, "grad_norm": 0.9967165589332581, "learning_rate": 3.184796854521625e-05, "loss": 0.7974, "step": 2100 }, { "epoch": 2.733897202342225, "grad_norm": 1.3412355184555054, "learning_rate": 3.181520314547838e-05, "loss": 1.2159, "step": 2101 }, { "epoch": 2.7351984385165906, "grad_norm": 1.487433671951294, "learning_rate": 3.17824377457405e-05, "loss": 0.6038, "step": 2102 }, { "epoch": 2.7364996746909567, "grad_norm": 1.0215046405792236, "learning_rate": 3.174967234600262e-05, "loss": 0.5847, "step": 2103 }, { "epoch": 2.737800910865322, "grad_norm": 1.0318578481674194, "learning_rate": 3.1716906946264745e-05, "loss": 1.1456, "step": 2104 }, { "epoch": 2.739102147039688, "grad_norm": 0.9738665819168091, "learning_rate": 3.168414154652687e-05, "loss": 0.7979, "step": 2105 }, { "epoch": 2.740403383214053, "grad_norm": 1.084647536277771, "learning_rate": 3.1651376146788995e-05, "loss": 1.558, "step": 2106 }, { "epoch": 2.741704619388419, "grad_norm": 1.2681946754455566, "learning_rate": 3.161861074705111e-05, "loss": 1.4728, "step": 2107 }, { "epoch": 2.7430058555627848, "grad_norm": 1.2492679357528687, "learning_rate": 3.158584534731324e-05, "loss": 0.9641, "step": 2108 }, { "epoch": 2.7443070917371504, "grad_norm": 1.2403067350387573, "learning_rate": 3.155307994757536e-05, "loss": 1.4203, "step": 2109 }, { "epoch": 2.745608327911516, "grad_norm": 1.2350013256072998, "learning_rate": 3.152031454783749e-05, "loss": 0.9768, "step": 2110 }, { "epoch": 2.7469095640858816, "grad_norm": 1.2828670740127563, "learning_rate": 3.148754914809961e-05, "loss": 0.9177, "step": 2111 }, { "epoch": 2.748210800260247, "grad_norm": 1.2588303089141846, "learning_rate": 3.145478374836173e-05, "loss": 1.1193, "step": 2112 }, { "epoch": 2.749512036434613, "grad_norm": 0.7041882276535034, "learning_rate": 3.1422018348623854e-05, "loss": 1.4572, "step": 2113 }, { "epoch": 2.7508132726089785, "grad_norm": 1.2522560358047485, "learning_rate": 3.138925294888598e-05, "loss": 1.1045, "step": 2114 }, { "epoch": 2.752114508783344, "grad_norm": 0.8506551384925842, "learning_rate": 3.13564875491481e-05, "loss": 1.4405, "step": 2115 }, { "epoch": 2.7534157449577097, "grad_norm": 1.2827317714691162, "learning_rate": 3.132372214941022e-05, "loss": 0.5784, "step": 2116 }, { "epoch": 2.7547169811320753, "grad_norm": 1.143186330795288, "learning_rate": 3.1290956749672346e-05, "loss": 0.5592, "step": 2117 }, { "epoch": 2.7560182173064414, "grad_norm": 1.0596147775650024, "learning_rate": 3.125819134993447e-05, "loss": 1.3074, "step": 2118 }, { "epoch": 2.7573194534808065, "grad_norm": 1.1027448177337646, "learning_rate": 3.1225425950196596e-05, "loss": 1.3912, "step": 2119 }, { "epoch": 2.7586206896551726, "grad_norm": 1.2030410766601562, "learning_rate": 3.119266055045872e-05, "loss": 1.1173, "step": 2120 }, { "epoch": 2.759921925829538, "grad_norm": 1.0351167917251587, "learning_rate": 3.115989515072084e-05, "loss": 1.4383, "step": 2121 }, { "epoch": 2.761223162003904, "grad_norm": 0.6431595683097839, "learning_rate": 3.112712975098296e-05, "loss": 1.3832, "step": 2122 }, { "epoch": 2.7625243981782694, "grad_norm": 1.073502540588379, "learning_rate": 3.109436435124509e-05, "loss": 1.5708, "step": 2123 }, { "epoch": 2.763825634352635, "grad_norm": 1.1776831150054932, "learning_rate": 3.1061598951507206e-05, "loss": 0.9931, "step": 2124 }, { "epoch": 2.7651268705270007, "grad_norm": 1.3575359582901, "learning_rate": 3.102883355176933e-05, "loss": 0.9766, "step": 2125 }, { "epoch": 2.7664281067013663, "grad_norm": 1.5416173934936523, "learning_rate": 3.0996068152031456e-05, "loss": 1.322, "step": 2126 }, { "epoch": 2.767729342875732, "grad_norm": 1.4217301607131958, "learning_rate": 3.096330275229358e-05, "loss": 0.8876, "step": 2127 }, { "epoch": 2.7690305790500975, "grad_norm": 1.163795828819275, "learning_rate": 3.0930537352555705e-05, "loss": 0.9688, "step": 2128 }, { "epoch": 2.770331815224463, "grad_norm": 1.3624329566955566, "learning_rate": 3.089777195281783e-05, "loss": 1.1231, "step": 2129 }, { "epoch": 2.7716330513988288, "grad_norm": 1.0671261548995972, "learning_rate": 3.086500655307995e-05, "loss": 1.0276, "step": 2130 }, { "epoch": 2.7729342875731944, "grad_norm": 1.2339879274368286, "learning_rate": 3.083224115334207e-05, "loss": 0.8704, "step": 2131 }, { "epoch": 2.77423552374756, "grad_norm": 1.3349480628967285, "learning_rate": 3.07994757536042e-05, "loss": 1.1364, "step": 2132 }, { "epoch": 2.775536759921926, "grad_norm": 1.2595629692077637, "learning_rate": 3.0766710353866315e-05, "loss": 1.298, "step": 2133 }, { "epoch": 2.7768379960962912, "grad_norm": 0.9153382182121277, "learning_rate": 3.073394495412844e-05, "loss": 1.2045, "step": 2134 }, { "epoch": 2.7781392322706573, "grad_norm": 1.2573519945144653, "learning_rate": 3.0701179554390565e-05, "loss": 0.8953, "step": 2135 }, { "epoch": 2.779440468445023, "grad_norm": 1.032522201538086, "learning_rate": 3.066841415465269e-05, "loss": 1.3571, "step": 2136 }, { "epoch": 2.7807417046193885, "grad_norm": 1.4318081140518188, "learning_rate": 3.0635648754914814e-05, "loss": 1.0917, "step": 2137 }, { "epoch": 2.782042940793754, "grad_norm": 1.2354663610458374, "learning_rate": 3.060288335517694e-05, "loss": 0.9652, "step": 2138 }, { "epoch": 2.7833441769681198, "grad_norm": 1.3827632665634155, "learning_rate": 3.057011795543906e-05, "loss": 0.7666, "step": 2139 }, { "epoch": 2.7846454131424854, "grad_norm": 1.4858529567718506, "learning_rate": 3.053735255570118e-05, "loss": 0.7365, "step": 2140 }, { "epoch": 2.785946649316851, "grad_norm": 1.1190621852874756, "learning_rate": 3.0504587155963303e-05, "loss": 1.248, "step": 2141 }, { "epoch": 2.7872478854912166, "grad_norm": 1.3180001974105835, "learning_rate": 3.0471821756225428e-05, "loss": 0.8775, "step": 2142 }, { "epoch": 2.788549121665582, "grad_norm": 1.2037584781646729, "learning_rate": 3.043905635648755e-05, "loss": 0.9113, "step": 2143 }, { "epoch": 2.789850357839948, "grad_norm": 1.2026875019073486, "learning_rate": 3.0406290956749674e-05, "loss": 0.9495, "step": 2144 }, { "epoch": 2.7911515940143135, "grad_norm": 0.9929836392402649, "learning_rate": 3.0373525557011795e-05, "loss": 0.8219, "step": 2145 }, { "epoch": 2.7924528301886795, "grad_norm": 1.1025787591934204, "learning_rate": 3.0340760157273916e-05, "loss": 1.0569, "step": 2146 }, { "epoch": 2.7937540663630447, "grad_norm": 1.3486242294311523, "learning_rate": 3.0307994757536044e-05, "loss": 1.1153, "step": 2147 }, { "epoch": 2.7950553025374107, "grad_norm": 0.9651402831077576, "learning_rate": 3.027522935779817e-05, "loss": 1.4791, "step": 2148 }, { "epoch": 2.796356538711776, "grad_norm": 1.017928957939148, "learning_rate": 3.024246395806029e-05, "loss": 1.451, "step": 2149 }, { "epoch": 2.797657774886142, "grad_norm": 1.3178097009658813, "learning_rate": 3.0209698558322412e-05, "loss": 0.8144, "step": 2150 }, { "epoch": 2.7989590110605076, "grad_norm": 1.6603940725326538, "learning_rate": 3.0176933158584537e-05, "loss": 1.6467, "step": 2151 }, { "epoch": 2.800260247234873, "grad_norm": 1.1972813606262207, "learning_rate": 3.0144167758846658e-05, "loss": 1.2701, "step": 2152 }, { "epoch": 2.801561483409239, "grad_norm": 1.5856012105941772, "learning_rate": 3.0111402359108783e-05, "loss": 0.382, "step": 2153 }, { "epoch": 2.8028627195836044, "grad_norm": 1.2603814601898193, "learning_rate": 3.0078636959370904e-05, "loss": 0.7677, "step": 2154 }, { "epoch": 2.80416395575797, "grad_norm": 1.219555377960205, "learning_rate": 3.0045871559633025e-05, "loss": 1.173, "step": 2155 }, { "epoch": 2.8054651919323357, "grad_norm": 1.304913878440857, "learning_rate": 3.0013106159895154e-05, "loss": 1.2803, "step": 2156 }, { "epoch": 2.8067664281067013, "grad_norm": 1.2624419927597046, "learning_rate": 2.998034076015728e-05, "loss": 1.4121, "step": 2157 }, { "epoch": 2.808067664281067, "grad_norm": 1.2204453945159912, "learning_rate": 2.99475753604194e-05, "loss": 0.8301, "step": 2158 }, { "epoch": 2.8093689004554325, "grad_norm": 1.4434590339660645, "learning_rate": 2.991480996068152e-05, "loss": 1.1258, "step": 2159 }, { "epoch": 2.810670136629798, "grad_norm": 0.9585099816322327, "learning_rate": 2.9882044560943646e-05, "loss": 1.0241, "step": 2160 }, { "epoch": 2.811971372804164, "grad_norm": 1.0932941436767578, "learning_rate": 2.9849279161205767e-05, "loss": 1.1139, "step": 2161 }, { "epoch": 2.8132726089785294, "grad_norm": 1.2006969451904297, "learning_rate": 2.9816513761467892e-05, "loss": 1.1594, "step": 2162 }, { "epoch": 2.8145738451528954, "grad_norm": 1.1747583150863647, "learning_rate": 2.9783748361730013e-05, "loss": 1.3315, "step": 2163 }, { "epoch": 2.815875081327261, "grad_norm": 1.2951160669326782, "learning_rate": 2.9750982961992135e-05, "loss": 0.8188, "step": 2164 }, { "epoch": 2.8171763175016267, "grad_norm": 0.9896459579467773, "learning_rate": 2.9718217562254263e-05, "loss": 1.697, "step": 2165 }, { "epoch": 2.8184775536759923, "grad_norm": 1.151100993156433, "learning_rate": 2.9685452162516387e-05, "loss": 0.8963, "step": 2166 }, { "epoch": 2.819778789850358, "grad_norm": 1.144951581954956, "learning_rate": 2.965268676277851e-05, "loss": 1.3036, "step": 2167 }, { "epoch": 2.8210800260247235, "grad_norm": 0.6802929639816284, "learning_rate": 2.961992136304063e-05, "loss": 0.8536, "step": 2168 }, { "epoch": 2.822381262199089, "grad_norm": 1.2014013528823853, "learning_rate": 2.9587155963302755e-05, "loss": 1.6074, "step": 2169 }, { "epoch": 2.8236824983734548, "grad_norm": 1.3735086917877197, "learning_rate": 2.9554390563564876e-05, "loss": 0.6898, "step": 2170 }, { "epoch": 2.8249837345478204, "grad_norm": 0.9849275946617126, "learning_rate": 2.9521625163826998e-05, "loss": 0.9255, "step": 2171 }, { "epoch": 2.826284970722186, "grad_norm": 1.6306239366531372, "learning_rate": 2.9488859764089122e-05, "loss": 1.0081, "step": 2172 }, { "epoch": 2.8275862068965516, "grad_norm": 1.303929328918457, "learning_rate": 2.9456094364351244e-05, "loss": 1.2051, "step": 2173 }, { "epoch": 2.828887443070917, "grad_norm": 1.1789219379425049, "learning_rate": 2.942332896461337e-05, "loss": 1.2094, "step": 2174 }, { "epoch": 2.830188679245283, "grad_norm": 1.8218307495117188, "learning_rate": 2.9390563564875496e-05, "loss": 1.3632, "step": 2175 }, { "epoch": 2.831489915419649, "grad_norm": 1.2730393409729004, "learning_rate": 2.9357798165137618e-05, "loss": 0.6987, "step": 2176 }, { "epoch": 2.832791151594014, "grad_norm": 1.0916818380355835, "learning_rate": 2.932503276539974e-05, "loss": 1.1944, "step": 2177 }, { "epoch": 2.83409238776838, "grad_norm": 1.5729656219482422, "learning_rate": 2.9292267365661864e-05, "loss": 0.6302, "step": 2178 }, { "epoch": 2.8353936239427457, "grad_norm": 1.2921148538589478, "learning_rate": 2.9259501965923985e-05, "loss": 1.3523, "step": 2179 }, { "epoch": 2.8366948601171114, "grad_norm": 1.448190450668335, "learning_rate": 2.9226736566186107e-05, "loss": 0.7113, "step": 2180 }, { "epoch": 2.837996096291477, "grad_norm": 1.536225438117981, "learning_rate": 2.919397116644823e-05, "loss": 1.6323, "step": 2181 }, { "epoch": 2.8392973324658426, "grad_norm": 1.1478601694107056, "learning_rate": 2.9161205766710353e-05, "loss": 1.041, "step": 2182 }, { "epoch": 2.840598568640208, "grad_norm": 0.9600987434387207, "learning_rate": 2.9128440366972477e-05, "loss": 1.3763, "step": 2183 }, { "epoch": 2.841899804814574, "grad_norm": 1.2167528867721558, "learning_rate": 2.9095674967234606e-05, "loss": 0.6386, "step": 2184 }, { "epoch": 2.8432010409889394, "grad_norm": 1.1971583366394043, "learning_rate": 2.9062909567496727e-05, "loss": 0.4167, "step": 2185 }, { "epoch": 2.844502277163305, "grad_norm": 1.3129202127456665, "learning_rate": 2.9030144167758848e-05, "loss": 1.1579, "step": 2186 }, { "epoch": 2.8458035133376707, "grad_norm": 1.2901173830032349, "learning_rate": 2.8997378768020973e-05, "loss": 0.7616, "step": 2187 }, { "epoch": 2.8471047495120363, "grad_norm": 1.0085517168045044, "learning_rate": 2.8964613368283094e-05, "loss": 1.0578, "step": 2188 }, { "epoch": 2.8484059856864024, "grad_norm": 0.8473798036575317, "learning_rate": 2.8931847968545216e-05, "loss": 1.7301, "step": 2189 }, { "epoch": 2.8497072218607675, "grad_norm": 1.190201997756958, "learning_rate": 2.889908256880734e-05, "loss": 1.2881, "step": 2190 }, { "epoch": 2.8510084580351336, "grad_norm": 1.0192360877990723, "learning_rate": 2.8866317169069462e-05, "loss": 1.5129, "step": 2191 }, { "epoch": 2.8523096942094988, "grad_norm": 1.3346720933914185, "learning_rate": 2.8833551769331587e-05, "loss": 0.7821, "step": 2192 }, { "epoch": 2.853610930383865, "grad_norm": 1.147904396057129, "learning_rate": 2.8800786369593715e-05, "loss": 1.1992, "step": 2193 }, { "epoch": 2.8549121665582304, "grad_norm": 1.0410293340682983, "learning_rate": 2.8768020969855836e-05, "loss": 1.0086, "step": 2194 }, { "epoch": 2.856213402732596, "grad_norm": 1.1989842653274536, "learning_rate": 2.8735255570117957e-05, "loss": 0.9306, "step": 2195 }, { "epoch": 2.8575146389069617, "grad_norm": 1.3104974031448364, "learning_rate": 2.8702490170380082e-05, "loss": 1.0646, "step": 2196 }, { "epoch": 2.8588158750813273, "grad_norm": 0.979457437992096, "learning_rate": 2.8669724770642203e-05, "loss": 1.3918, "step": 2197 }, { "epoch": 2.860117111255693, "grad_norm": 1.3049026727676392, "learning_rate": 2.8636959370904325e-05, "loss": 1.1586, "step": 2198 }, { "epoch": 2.8614183474300585, "grad_norm": 1.414306879043579, "learning_rate": 2.860419397116645e-05, "loss": 0.6975, "step": 2199 }, { "epoch": 2.862719583604424, "grad_norm": 1.3470391035079956, "learning_rate": 2.857142857142857e-05, "loss": 0.8671, "step": 2200 }, { "epoch": 2.8640208197787898, "grad_norm": 0.8922164440155029, "learning_rate": 2.8538663171690692e-05, "loss": 1.1627, "step": 2201 }, { "epoch": 2.8653220559531554, "grad_norm": 1.1149957180023193, "learning_rate": 2.850589777195282e-05, "loss": 1.6821, "step": 2202 }, { "epoch": 2.866623292127521, "grad_norm": 1.3545335531234741, "learning_rate": 2.8473132372214945e-05, "loss": 0.9824, "step": 2203 }, { "epoch": 2.867924528301887, "grad_norm": 1.3498330116271973, "learning_rate": 2.8440366972477066e-05, "loss": 0.8355, "step": 2204 }, { "epoch": 2.869225764476252, "grad_norm": 1.2998915910720825, "learning_rate": 2.840760157273919e-05, "loss": 0.9405, "step": 2205 }, { "epoch": 2.8705270006506183, "grad_norm": 1.2146217823028564, "learning_rate": 2.8374836173001313e-05, "loss": 1.2031, "step": 2206 }, { "epoch": 2.871828236824984, "grad_norm": 0.9784196019172668, "learning_rate": 2.8342070773263434e-05, "loss": 0.7906, "step": 2207 }, { "epoch": 2.8731294729993495, "grad_norm": 0.8168174028396606, "learning_rate": 2.830930537352556e-05, "loss": 0.8991, "step": 2208 }, { "epoch": 2.874430709173715, "grad_norm": 0.9746270775794983, "learning_rate": 2.827653997378768e-05, "loss": 1.6094, "step": 2209 }, { "epoch": 2.8757319453480807, "grad_norm": 1.2500298023223877, "learning_rate": 2.82437745740498e-05, "loss": 0.9773, "step": 2210 }, { "epoch": 2.8770331815224464, "grad_norm": 1.034111499786377, "learning_rate": 2.8211009174311926e-05, "loss": 1.1561, "step": 2211 }, { "epoch": 2.878334417696812, "grad_norm": 1.0458588600158691, "learning_rate": 2.8178243774574054e-05, "loss": 0.8726, "step": 2212 }, { "epoch": 2.8796356538711776, "grad_norm": 0.9169819951057434, "learning_rate": 2.8145478374836176e-05, "loss": 1.2198, "step": 2213 }, { "epoch": 2.880936890045543, "grad_norm": 1.1924256086349487, "learning_rate": 2.81127129750983e-05, "loss": 0.9715, "step": 2214 }, { "epoch": 2.882238126219909, "grad_norm": 1.0221024751663208, "learning_rate": 2.807994757536042e-05, "loss": 1.1256, "step": 2215 }, { "epoch": 2.8835393623942744, "grad_norm": 1.0046567916870117, "learning_rate": 2.8047182175622543e-05, "loss": 1.5351, "step": 2216 }, { "epoch": 2.88484059856864, "grad_norm": 1.0506677627563477, "learning_rate": 2.8014416775884668e-05, "loss": 0.7806, "step": 2217 }, { "epoch": 2.8861418347430057, "grad_norm": 0.9440841674804688, "learning_rate": 2.798165137614679e-05, "loss": 1.7092, "step": 2218 }, { "epoch": 2.8874430709173717, "grad_norm": 1.2449657917022705, "learning_rate": 2.794888597640891e-05, "loss": 1.0276, "step": 2219 }, { "epoch": 2.888744307091737, "grad_norm": 1.7142256498336792, "learning_rate": 2.7916120576671035e-05, "loss": 0.9612, "step": 2220 }, { "epoch": 2.890045543266103, "grad_norm": 1.3049640655517578, "learning_rate": 2.7883355176933163e-05, "loss": 1.4016, "step": 2221 }, { "epoch": 2.8913467794404686, "grad_norm": 1.2219700813293457, "learning_rate": 2.7850589777195285e-05, "loss": 1.2044, "step": 2222 }, { "epoch": 2.892648015614834, "grad_norm": 1.3041671514511108, "learning_rate": 2.781782437745741e-05, "loss": 1.2348, "step": 2223 }, { "epoch": 2.8939492517892, "grad_norm": 1.2235584259033203, "learning_rate": 2.778505897771953e-05, "loss": 0.9834, "step": 2224 }, { "epoch": 2.8952504879635654, "grad_norm": 1.3997191190719604, "learning_rate": 2.7752293577981652e-05, "loss": 1.2259, "step": 2225 }, { "epoch": 2.896551724137931, "grad_norm": 1.3120765686035156, "learning_rate": 2.7719528178243777e-05, "loss": 1.1808, "step": 2226 }, { "epoch": 2.8978529603122967, "grad_norm": 1.0775145292282104, "learning_rate": 2.7686762778505898e-05, "loss": 1.2241, "step": 2227 }, { "epoch": 2.8991541964866623, "grad_norm": 0.9246907234191895, "learning_rate": 2.765399737876802e-05, "loss": 0.9304, "step": 2228 }, { "epoch": 2.900455432661028, "grad_norm": 1.0337483882904053, "learning_rate": 2.7621231979030144e-05, "loss": 1.3097, "step": 2229 }, { "epoch": 2.9017566688353935, "grad_norm": 0.9489741921424866, "learning_rate": 2.7588466579292272e-05, "loss": 1.2563, "step": 2230 }, { "epoch": 2.903057905009759, "grad_norm": 0.9083828330039978, "learning_rate": 2.7555701179554394e-05, "loss": 0.5496, "step": 2231 }, { "epoch": 2.904359141184125, "grad_norm": 1.234840750694275, "learning_rate": 2.7522935779816515e-05, "loss": 1.3153, "step": 2232 }, { "epoch": 2.9056603773584904, "grad_norm": 1.0396250486373901, "learning_rate": 2.749017038007864e-05, "loss": 0.8848, "step": 2233 }, { "epoch": 2.9069616135328564, "grad_norm": 1.457045555114746, "learning_rate": 2.745740498034076e-05, "loss": 0.7831, "step": 2234 }, { "epoch": 2.9082628497072216, "grad_norm": 1.0620949268341064, "learning_rate": 2.7424639580602886e-05, "loss": 0.8973, "step": 2235 }, { "epoch": 2.9095640858815877, "grad_norm": 1.2896455526351929, "learning_rate": 2.7391874180865007e-05, "loss": 1.5504, "step": 2236 }, { "epoch": 2.9108653220559533, "grad_norm": 1.231295108795166, "learning_rate": 2.735910878112713e-05, "loss": 0.8293, "step": 2237 }, { "epoch": 2.912166558230319, "grad_norm": 1.736441969871521, "learning_rate": 2.7326343381389253e-05, "loss": 0.8922, "step": 2238 }, { "epoch": 2.9134677944046845, "grad_norm": 1.1010597944259644, "learning_rate": 2.7293577981651375e-05, "loss": 1.3022, "step": 2239 }, { "epoch": 2.91476903057905, "grad_norm": 1.1967710256576538, "learning_rate": 2.7260812581913503e-05, "loss": 1.3213, "step": 2240 }, { "epoch": 2.9160702667534157, "grad_norm": 1.186230182647705, "learning_rate": 2.7228047182175624e-05, "loss": 1.2443, "step": 2241 }, { "epoch": 2.9173715029277814, "grad_norm": 1.2016786336898804, "learning_rate": 2.719528178243775e-05, "loss": 0.758, "step": 2242 }, { "epoch": 2.918672739102147, "grad_norm": 1.2476412057876587, "learning_rate": 2.716251638269987e-05, "loss": 1.196, "step": 2243 }, { "epoch": 2.9199739752765126, "grad_norm": 1.0142265558242798, "learning_rate": 2.7129750982961995e-05, "loss": 1.7651, "step": 2244 }, { "epoch": 2.921275211450878, "grad_norm": 1.145908236503601, "learning_rate": 2.7096985583224116e-05, "loss": 1.096, "step": 2245 }, { "epoch": 2.922576447625244, "grad_norm": 1.1751781702041626, "learning_rate": 2.7064220183486238e-05, "loss": 1.147, "step": 2246 }, { "epoch": 2.92387768379961, "grad_norm": 1.215545415878296, "learning_rate": 2.7031454783748362e-05, "loss": 1.0557, "step": 2247 }, { "epoch": 2.925178919973975, "grad_norm": 1.1052942276000977, "learning_rate": 2.6998689384010484e-05, "loss": 1.0826, "step": 2248 }, { "epoch": 2.926480156148341, "grad_norm": 0.8546999096870422, "learning_rate": 2.6965923984272612e-05, "loss": 1.6184, "step": 2249 }, { "epoch": 2.9277813923227067, "grad_norm": 1.23273766040802, "learning_rate": 2.6933158584534733e-05, "loss": 1.1452, "step": 2250 }, { "epoch": 2.9290826284970723, "grad_norm": 1.0549744367599487, "learning_rate": 2.6900393184796858e-05, "loss": 0.774, "step": 2251 }, { "epoch": 2.930383864671438, "grad_norm": 0.7598772644996643, "learning_rate": 2.686762778505898e-05, "loss": 1.0509, "step": 2252 }, { "epoch": 2.9316851008458036, "grad_norm": 1.0779083967208862, "learning_rate": 2.68348623853211e-05, "loss": 0.7886, "step": 2253 }, { "epoch": 2.932986337020169, "grad_norm": 1.4423840045928955, "learning_rate": 2.6802096985583225e-05, "loss": 1.1845, "step": 2254 }, { "epoch": 2.934287573194535, "grad_norm": 1.0684314966201782, "learning_rate": 2.6769331585845347e-05, "loss": 1.2008, "step": 2255 }, { "epoch": 2.9355888093689004, "grad_norm": 1.174725890159607, "learning_rate": 2.673656618610747e-05, "loss": 1.3211, "step": 2256 }, { "epoch": 2.936890045543266, "grad_norm": 1.1522530317306519, "learning_rate": 2.6703800786369593e-05, "loss": 0.8613, "step": 2257 }, { "epoch": 2.9381912817176317, "grad_norm": 0.8441818356513977, "learning_rate": 2.667103538663172e-05, "loss": 1.1383, "step": 2258 }, { "epoch": 2.9394925178919973, "grad_norm": 1.1148555278778076, "learning_rate": 2.6638269986893842e-05, "loss": 0.8313, "step": 2259 }, { "epoch": 2.940793754066363, "grad_norm": 1.273287057876587, "learning_rate": 2.6605504587155967e-05, "loss": 0.9341, "step": 2260 }, { "epoch": 2.9420949902407285, "grad_norm": 1.2231539487838745, "learning_rate": 2.657273918741809e-05, "loss": 1.2795, "step": 2261 }, { "epoch": 2.9433962264150946, "grad_norm": 0.8901998996734619, "learning_rate": 2.653997378768021e-05, "loss": 1.4039, "step": 2262 }, { "epoch": 2.9446974625894597, "grad_norm": 0.9062055349349976, "learning_rate": 2.6507208387942334e-05, "loss": 1.0196, "step": 2263 }, { "epoch": 2.945998698763826, "grad_norm": 0.9170067310333252, "learning_rate": 2.6474442988204456e-05, "loss": 1.2378, "step": 2264 }, { "epoch": 2.9472999349381914, "grad_norm": 1.1292369365692139, "learning_rate": 2.644167758846658e-05, "loss": 0.9002, "step": 2265 }, { "epoch": 2.948601171112557, "grad_norm": 1.3564952611923218, "learning_rate": 2.6408912188728702e-05, "loss": 0.9315, "step": 2266 }, { "epoch": 2.9499024072869227, "grad_norm": 1.243621587753296, "learning_rate": 2.6376146788990823e-05, "loss": 1.1323, "step": 2267 }, { "epoch": 2.9512036434612883, "grad_norm": 1.6271698474884033, "learning_rate": 2.634338138925295e-05, "loss": 0.6339, "step": 2268 }, { "epoch": 2.952504879635654, "grad_norm": 1.6169137954711914, "learning_rate": 2.6310615989515076e-05, "loss": 0.9482, "step": 2269 }, { "epoch": 2.9538061158100195, "grad_norm": 1.1517918109893799, "learning_rate": 2.6277850589777197e-05, "loss": 1.0022, "step": 2270 }, { "epoch": 2.955107351984385, "grad_norm": 0.9843864440917969, "learning_rate": 2.624508519003932e-05, "loss": 1.6133, "step": 2271 }, { "epoch": 2.9564085881587507, "grad_norm": 1.3900986909866333, "learning_rate": 2.6212319790301444e-05, "loss": 0.776, "step": 2272 }, { "epoch": 2.9577098243331164, "grad_norm": 1.2190008163452148, "learning_rate": 2.6179554390563565e-05, "loss": 0.7874, "step": 2273 }, { "epoch": 2.959011060507482, "grad_norm": 1.1685391664505005, "learning_rate": 2.614678899082569e-05, "loss": 1.1888, "step": 2274 }, { "epoch": 2.960312296681848, "grad_norm": 1.32119619846344, "learning_rate": 2.611402359108781e-05, "loss": 1.0145, "step": 2275 }, { "epoch": 2.961613532856213, "grad_norm": 1.2174742221832275, "learning_rate": 2.6081258191349932e-05, "loss": 0.9686, "step": 2276 }, { "epoch": 2.9629147690305793, "grad_norm": 0.8733085989952087, "learning_rate": 2.604849279161206e-05, "loss": 0.6863, "step": 2277 }, { "epoch": 2.9642160052049444, "grad_norm": 1.198016881942749, "learning_rate": 2.6015727391874185e-05, "loss": 0.8855, "step": 2278 }, { "epoch": 2.9655172413793105, "grad_norm": 1.013948678970337, "learning_rate": 2.5982961992136307e-05, "loss": 1.7331, "step": 2279 }, { "epoch": 2.966818477553676, "grad_norm": 1.0356991291046143, "learning_rate": 2.5950196592398428e-05, "loss": 0.6316, "step": 2280 }, { "epoch": 2.9681197137280417, "grad_norm": 1.1468837261199951, "learning_rate": 2.5917431192660553e-05, "loss": 1.0131, "step": 2281 }, { "epoch": 2.9694209499024073, "grad_norm": 1.1900533437728882, "learning_rate": 2.5884665792922674e-05, "loss": 1.1554, "step": 2282 }, { "epoch": 2.970722186076773, "grad_norm": 1.110066294670105, "learning_rate": 2.5851900393184795e-05, "loss": 0.9149, "step": 2283 }, { "epoch": 2.9720234222511386, "grad_norm": 1.0681748390197754, "learning_rate": 2.581913499344692e-05, "loss": 0.9113, "step": 2284 }, { "epoch": 2.973324658425504, "grad_norm": 1.128723382949829, "learning_rate": 2.578636959370904e-05, "loss": 0.6589, "step": 2285 }, { "epoch": 2.97462589459987, "grad_norm": 1.1354516744613647, "learning_rate": 2.575360419397117e-05, "loss": 1.6002, "step": 2286 }, { "epoch": 2.9759271307742354, "grad_norm": 0.9696306586265564, "learning_rate": 2.5720838794233294e-05, "loss": 1.3813, "step": 2287 }, { "epoch": 2.977228366948601, "grad_norm": 1.0159891843795776, "learning_rate": 2.5688073394495416e-05, "loss": 1.0861, "step": 2288 }, { "epoch": 2.9785296031229667, "grad_norm": 1.3373439311981201, "learning_rate": 2.5655307994757537e-05, "loss": 1.1507, "step": 2289 }, { "epoch": 2.9798308392973327, "grad_norm": 1.393628716468811, "learning_rate": 2.562254259501966e-05, "loss": 1.0542, "step": 2290 }, { "epoch": 2.981132075471698, "grad_norm": 1.4429336786270142, "learning_rate": 2.5589777195281783e-05, "loss": 1.2881, "step": 2291 }, { "epoch": 2.982433311646064, "grad_norm": 1.0309065580368042, "learning_rate": 2.5557011795543904e-05, "loss": 1.6443, "step": 2292 }, { "epoch": 2.9837345478204296, "grad_norm": 1.6249536275863647, "learning_rate": 2.552424639580603e-05, "loss": 0.9083, "step": 2293 }, { "epoch": 2.985035783994795, "grad_norm": 1.281638264656067, "learning_rate": 2.549148099606815e-05, "loss": 1.1705, "step": 2294 }, { "epoch": 2.986337020169161, "grad_norm": 1.242109775543213, "learning_rate": 2.5458715596330275e-05, "loss": 1.1428, "step": 2295 }, { "epoch": 2.9876382563435264, "grad_norm": 1.209904670715332, "learning_rate": 2.5425950196592403e-05, "loss": 0.7206, "step": 2296 }, { "epoch": 2.988939492517892, "grad_norm": 1.259116530418396, "learning_rate": 2.5393184796854525e-05, "loss": 1.1772, "step": 2297 }, { "epoch": 2.9902407286922577, "grad_norm": 1.2644240856170654, "learning_rate": 2.5360419397116646e-05, "loss": 1.2691, "step": 2298 }, { "epoch": 2.9915419648666233, "grad_norm": 1.2261430025100708, "learning_rate": 2.532765399737877e-05, "loss": 1.0461, "step": 2299 }, { "epoch": 2.992843201040989, "grad_norm": 1.407590389251709, "learning_rate": 2.5294888597640892e-05, "loss": 1.1613, "step": 2300 }, { "epoch": 2.9941444372153545, "grad_norm": 1.517208456993103, "learning_rate": 2.5262123197903013e-05, "loss": 0.5945, "step": 2301 }, { "epoch": 2.99544567338972, "grad_norm": 0.9011630415916443, "learning_rate": 2.5229357798165138e-05, "loss": 1.2943, "step": 2302 }, { "epoch": 2.9967469095640857, "grad_norm": 2.1898372173309326, "learning_rate": 2.519659239842726e-05, "loss": 0.7765, "step": 2303 }, { "epoch": 2.9980481457384514, "grad_norm": 1.5480730533599854, "learning_rate": 2.5163826998689384e-05, "loss": 1.4673, "step": 2304 }, { "epoch": 2.9993493819128174, "grad_norm": 1.1190319061279297, "learning_rate": 2.5131061598951512e-05, "loss": 1.6022, "step": 2305 }, { "epoch": 3.000650618087183, "grad_norm": 1.0218429565429688, "learning_rate": 2.5098296199213634e-05, "loss": 0.8462, "step": 2306 }, { "epoch": 3.0019518542615486, "grad_norm": 1.0551577806472778, "learning_rate": 2.5065530799475755e-05, "loss": 0.9863, "step": 2307 }, { "epoch": 3.0032530904359143, "grad_norm": 1.0490330457687378, "learning_rate": 2.503276539973788e-05, "loss": 0.3867, "step": 2308 }, { "epoch": 3.00455432661028, "grad_norm": 1.2152104377746582, "learning_rate": 2.5e-05, "loss": 1.3825, "step": 2309 }, { "epoch": 3.0058555627846455, "grad_norm": 0.8942937254905701, "learning_rate": 2.4967234600262123e-05, "loss": 0.2526, "step": 2310 }, { "epoch": 3.007156798959011, "grad_norm": 0.9736802577972412, "learning_rate": 2.4934469200524247e-05, "loss": 0.3423, "step": 2311 }, { "epoch": 3.0084580351333767, "grad_norm": 1.355350136756897, "learning_rate": 2.4901703800786372e-05, "loss": 0.3299, "step": 2312 }, { "epoch": 3.0097592713077423, "grad_norm": 1.2429851293563843, "learning_rate": 2.4868938401048493e-05, "loss": 0.4052, "step": 2313 }, { "epoch": 3.011060507482108, "grad_norm": 2.001514434814453, "learning_rate": 2.4836173001310618e-05, "loss": 0.5762, "step": 2314 }, { "epoch": 3.0123617436564736, "grad_norm": 2.030777931213379, "learning_rate": 2.480340760157274e-05, "loss": 0.5933, "step": 2315 }, { "epoch": 3.013662979830839, "grad_norm": 1.5050444602966309, "learning_rate": 2.4770642201834864e-05, "loss": 0.8015, "step": 2316 }, { "epoch": 3.014964216005205, "grad_norm": 1.3923333883285522, "learning_rate": 2.473787680209699e-05, "loss": 0.3834, "step": 2317 }, { "epoch": 3.0162654521795704, "grad_norm": 1.394416093826294, "learning_rate": 2.470511140235911e-05, "loss": 0.3856, "step": 2318 }, { "epoch": 3.017566688353936, "grad_norm": 1.2047126293182373, "learning_rate": 2.467234600262123e-05, "loss": 0.4118, "step": 2319 }, { "epoch": 3.018867924528302, "grad_norm": 1.4159356355667114, "learning_rate": 2.4639580602883356e-05, "loss": 0.5029, "step": 2320 }, { "epoch": 3.0201691607026677, "grad_norm": 0.9608712792396545, "learning_rate": 2.460681520314548e-05, "loss": 0.6767, "step": 2321 }, { "epoch": 3.0214703968770333, "grad_norm": 1.3776761293411255, "learning_rate": 2.4574049803407602e-05, "loss": 0.1621, "step": 2322 }, { "epoch": 3.022771633051399, "grad_norm": 0.9410603642463684, "learning_rate": 2.4541284403669727e-05, "loss": 0.1768, "step": 2323 }, { "epoch": 3.0240728692257646, "grad_norm": 1.0585469007492065, "learning_rate": 2.450851900393185e-05, "loss": 0.6193, "step": 2324 }, { "epoch": 3.02537410540013, "grad_norm": 1.0434179306030273, "learning_rate": 2.447575360419397e-05, "loss": 0.9859, "step": 2325 }, { "epoch": 3.026675341574496, "grad_norm": 1.2792407274246216, "learning_rate": 2.4442988204456098e-05, "loss": 0.4621, "step": 2326 }, { "epoch": 3.0279765777488614, "grad_norm": 1.2210074663162231, "learning_rate": 2.441022280471822e-05, "loss": 0.5126, "step": 2327 }, { "epoch": 3.029277813923227, "grad_norm": 1.012719988822937, "learning_rate": 2.437745740498034e-05, "loss": 0.5337, "step": 2328 }, { "epoch": 3.0305790500975927, "grad_norm": 1.1410574913024902, "learning_rate": 2.4344692005242465e-05, "loss": 0.3441, "step": 2329 }, { "epoch": 3.0318802862719583, "grad_norm": 1.4326224327087402, "learning_rate": 2.431192660550459e-05, "loss": 1.0692, "step": 2330 }, { "epoch": 3.033181522446324, "grad_norm": 1.1703680753707886, "learning_rate": 2.427916120576671e-05, "loss": 0.4311, "step": 2331 }, { "epoch": 3.0344827586206895, "grad_norm": 1.2722362279891968, "learning_rate": 2.4246395806028836e-05, "loss": 0.3743, "step": 2332 }, { "epoch": 3.035783994795055, "grad_norm": 1.1848348379135132, "learning_rate": 2.4213630406290958e-05, "loss": 0.2698, "step": 2333 }, { "epoch": 3.0370852309694207, "grad_norm": 0.9460013508796692, "learning_rate": 2.418086500655308e-05, "loss": 0.5992, "step": 2334 }, { "epoch": 3.038386467143787, "grad_norm": 1.6224673986434937, "learning_rate": 2.4148099606815207e-05, "loss": 0.53, "step": 2335 }, { "epoch": 3.0396877033181524, "grad_norm": 1.882067322731018, "learning_rate": 2.411533420707733e-05, "loss": 0.6241, "step": 2336 }, { "epoch": 3.040988939492518, "grad_norm": 1.715589165687561, "learning_rate": 2.408256880733945e-05, "loss": 0.4812, "step": 2337 }, { "epoch": 3.0422901756668836, "grad_norm": 1.907049536705017, "learning_rate": 2.4049803407601575e-05, "loss": 0.2522, "step": 2338 }, { "epoch": 3.0435914118412493, "grad_norm": 1.817122220993042, "learning_rate": 2.4017038007863696e-05, "loss": 0.3918, "step": 2339 }, { "epoch": 3.044892648015615, "grad_norm": 1.264664649963379, "learning_rate": 2.398427260812582e-05, "loss": 0.5102, "step": 2340 }, { "epoch": 3.0461938841899805, "grad_norm": 1.4708404541015625, "learning_rate": 2.3951507208387945e-05, "loss": 0.4686, "step": 2341 }, { "epoch": 3.047495120364346, "grad_norm": 1.0696125030517578, "learning_rate": 2.3918741808650067e-05, "loss": 0.3282, "step": 2342 }, { "epoch": 3.0487963565387117, "grad_norm": 1.412275791168213, "learning_rate": 2.3885976408912188e-05, "loss": 0.8637, "step": 2343 }, { "epoch": 3.0500975927130773, "grad_norm": 0.8984838724136353, "learning_rate": 2.3853211009174313e-05, "loss": 0.6922, "step": 2344 }, { "epoch": 3.051398828887443, "grad_norm": 0.9333692193031311, "learning_rate": 2.3820445609436438e-05, "loss": 0.3739, "step": 2345 }, { "epoch": 3.0527000650618086, "grad_norm": 0.9598501920700073, "learning_rate": 2.378768020969856e-05, "loss": 0.4723, "step": 2346 }, { "epoch": 3.054001301236174, "grad_norm": 1.5608861446380615, "learning_rate": 2.3754914809960684e-05, "loss": 0.4652, "step": 2347 }, { "epoch": 3.05530253741054, "grad_norm": 1.1710401773452759, "learning_rate": 2.3722149410222805e-05, "loss": 0.6016, "step": 2348 }, { "epoch": 3.056603773584906, "grad_norm": 1.0170390605926514, "learning_rate": 2.368938401048493e-05, "loss": 0.4295, "step": 2349 }, { "epoch": 3.0579050097592715, "grad_norm": 1.1102229356765747, "learning_rate": 2.3656618610747054e-05, "loss": 0.2697, "step": 2350 }, { "epoch": 3.059206245933637, "grad_norm": 0.9093301296234131, "learning_rate": 2.3623853211009176e-05, "loss": 1.2772, "step": 2351 }, { "epoch": 3.0605074821080027, "grad_norm": 1.850153923034668, "learning_rate": 2.3591087811271297e-05, "loss": 0.8068, "step": 2352 }, { "epoch": 3.0618087182823683, "grad_norm": 1.5126678943634033, "learning_rate": 2.3558322411533422e-05, "loss": 0.7464, "step": 2353 }, { "epoch": 3.063109954456734, "grad_norm": 1.443954348564148, "learning_rate": 2.3525557011795547e-05, "loss": 0.853, "step": 2354 }, { "epoch": 3.0644111906310996, "grad_norm": 1.623030424118042, "learning_rate": 2.3492791612057668e-05, "loss": 0.7586, "step": 2355 }, { "epoch": 3.065712426805465, "grad_norm": 1.4001879692077637, "learning_rate": 2.3460026212319793e-05, "loss": 0.4191, "step": 2356 }, { "epoch": 3.067013662979831, "grad_norm": 1.3383530378341675, "learning_rate": 2.3427260812581914e-05, "loss": 0.5133, "step": 2357 }, { "epoch": 3.0683148991541964, "grad_norm": 1.1870273351669312, "learning_rate": 2.339449541284404e-05, "loss": 0.3261, "step": 2358 }, { "epoch": 3.069616135328562, "grad_norm": 1.2247710227966309, "learning_rate": 2.336173001310616e-05, "loss": 0.215, "step": 2359 }, { "epoch": 3.0709173715029277, "grad_norm": 0.8856111764907837, "learning_rate": 2.3328964613368285e-05, "loss": 0.8637, "step": 2360 }, { "epoch": 3.0722186076772933, "grad_norm": 0.9899976849555969, "learning_rate": 2.3296199213630406e-05, "loss": 0.7184, "step": 2361 }, { "epoch": 3.073519843851659, "grad_norm": 1.3027199506759644, "learning_rate": 2.326343381389253e-05, "loss": 0.7382, "step": 2362 }, { "epoch": 3.074821080026025, "grad_norm": 1.0329011678695679, "learning_rate": 2.3230668414154656e-05, "loss": 0.9397, "step": 2363 }, { "epoch": 3.0761223162003906, "grad_norm": 1.1049106121063232, "learning_rate": 2.3197903014416777e-05, "loss": 0.806, "step": 2364 }, { "epoch": 3.077423552374756, "grad_norm": 0.9721315503120422, "learning_rate": 2.3165137614678902e-05, "loss": 0.1866, "step": 2365 }, { "epoch": 3.078724788549122, "grad_norm": 1.1062901020050049, "learning_rate": 2.3132372214941023e-05, "loss": 0.5795, "step": 2366 }, { "epoch": 3.0800260247234874, "grad_norm": 1.325032353401184, "learning_rate": 2.3099606815203144e-05, "loss": 0.7496, "step": 2367 }, { "epoch": 3.081327260897853, "grad_norm": 1.2286983728408813, "learning_rate": 2.306684141546527e-05, "loss": 0.5368, "step": 2368 }, { "epoch": 3.0826284970722186, "grad_norm": 1.2688992023468018, "learning_rate": 2.3034076015727394e-05, "loss": 0.9626, "step": 2369 }, { "epoch": 3.0839297332465843, "grad_norm": 0.7892445921897888, "learning_rate": 2.3001310615989515e-05, "loss": 0.2566, "step": 2370 }, { "epoch": 3.08523096942095, "grad_norm": 1.3920766115188599, "learning_rate": 2.296854521625164e-05, "loss": 0.625, "step": 2371 }, { "epoch": 3.0865322055953155, "grad_norm": 1.5427662134170532, "learning_rate": 2.2935779816513765e-05, "loss": 0.4718, "step": 2372 }, { "epoch": 3.087833441769681, "grad_norm": 1.2104597091674805, "learning_rate": 2.2903014416775886e-05, "loss": 0.4955, "step": 2373 }, { "epoch": 3.0891346779440467, "grad_norm": 1.2865535020828247, "learning_rate": 2.2870249017038007e-05, "loss": 0.41, "step": 2374 }, { "epoch": 3.0904359141184123, "grad_norm": 1.0505013465881348, "learning_rate": 2.2837483617300132e-05, "loss": 0.9959, "step": 2375 }, { "epoch": 3.091737150292778, "grad_norm": 0.9925892353057861, "learning_rate": 2.2804718217562254e-05, "loss": 0.549, "step": 2376 }, { "epoch": 3.0930383864671436, "grad_norm": 1.3497719764709473, "learning_rate": 2.2771952817824378e-05, "loss": 1.0141, "step": 2377 }, { "epoch": 3.0943396226415096, "grad_norm": 1.344407081604004, "learning_rate": 2.2739187418086503e-05, "loss": 0.9446, "step": 2378 }, { "epoch": 3.0956408588158753, "grad_norm": 1.2035760879516602, "learning_rate": 2.2706422018348624e-05, "loss": 0.443, "step": 2379 }, { "epoch": 3.096942094990241, "grad_norm": 1.2615950107574463, "learning_rate": 2.2673656618610746e-05, "loss": 0.5651, "step": 2380 }, { "epoch": 3.0982433311646065, "grad_norm": 1.4722880125045776, "learning_rate": 2.264089121887287e-05, "loss": 0.3824, "step": 2381 }, { "epoch": 3.099544567338972, "grad_norm": 1.4240970611572266, "learning_rate": 2.2608125819134995e-05, "loss": 0.2055, "step": 2382 }, { "epoch": 3.1008458035133377, "grad_norm": 1.2488726377487183, "learning_rate": 2.2575360419397117e-05, "loss": 0.3188, "step": 2383 }, { "epoch": 3.1021470396877033, "grad_norm": 1.365740418434143, "learning_rate": 2.254259501965924e-05, "loss": 0.3763, "step": 2384 }, { "epoch": 3.103448275862069, "grad_norm": 1.5159258842468262, "learning_rate": 2.2509829619921363e-05, "loss": 0.9403, "step": 2385 }, { "epoch": 3.1047495120364346, "grad_norm": 1.080348253250122, "learning_rate": 2.2477064220183487e-05, "loss": 0.1251, "step": 2386 }, { "epoch": 3.1060507482108, "grad_norm": 1.2740956544876099, "learning_rate": 2.2444298820445612e-05, "loss": 0.8551, "step": 2387 }, { "epoch": 3.107351984385166, "grad_norm": 1.1411464214324951, "learning_rate": 2.2411533420707733e-05, "loss": 0.8866, "step": 2388 }, { "epoch": 3.1086532205595314, "grad_norm": 1.137598991394043, "learning_rate": 2.2378768020969855e-05, "loss": 0.2939, "step": 2389 }, { "epoch": 3.109954456733897, "grad_norm": 1.0940495729446411, "learning_rate": 2.234600262123198e-05, "loss": 0.5865, "step": 2390 }, { "epoch": 3.1112556929082626, "grad_norm": 1.1396749019622803, "learning_rate": 2.2313237221494104e-05, "loss": 0.8383, "step": 2391 }, { "epoch": 3.1125569290826287, "grad_norm": 1.325338363647461, "learning_rate": 2.2280471821756226e-05, "loss": 0.7271, "step": 2392 }, { "epoch": 3.1138581652569943, "grad_norm": 1.3903186321258545, "learning_rate": 2.224770642201835e-05, "loss": 0.8074, "step": 2393 }, { "epoch": 3.11515940143136, "grad_norm": 1.4454991817474365, "learning_rate": 2.2214941022280472e-05, "loss": 0.9784, "step": 2394 }, { "epoch": 3.1164606376057256, "grad_norm": 1.3565646409988403, "learning_rate": 2.2182175622542593e-05, "loss": 0.6341, "step": 2395 }, { "epoch": 3.117761873780091, "grad_norm": 1.1957907676696777, "learning_rate": 2.214941022280472e-05, "loss": 0.5977, "step": 2396 }, { "epoch": 3.119063109954457, "grad_norm": 1.2001465559005737, "learning_rate": 2.2116644823066843e-05, "loss": 0.4884, "step": 2397 }, { "epoch": 3.1203643461288224, "grad_norm": 1.6271768808364868, "learning_rate": 2.2083879423328964e-05, "loss": 0.5956, "step": 2398 }, { "epoch": 3.121665582303188, "grad_norm": 1.6868019104003906, "learning_rate": 2.205111402359109e-05, "loss": 0.7694, "step": 2399 }, { "epoch": 3.1229668184775536, "grad_norm": 1.2876911163330078, "learning_rate": 2.2018348623853213e-05, "loss": 0.3078, "step": 2400 }, { "epoch": 3.1242680546519193, "grad_norm": 1.2186596393585205, "learning_rate": 2.1985583224115335e-05, "loss": 0.6786, "step": 2401 }, { "epoch": 3.125569290826285, "grad_norm": 1.4328972101211548, "learning_rate": 2.195281782437746e-05, "loss": 0.9041, "step": 2402 }, { "epoch": 3.1268705270006505, "grad_norm": 1.2887179851531982, "learning_rate": 2.192005242463958e-05, "loss": 0.8552, "step": 2403 }, { "epoch": 3.128171763175016, "grad_norm": 0.8244163393974304, "learning_rate": 2.1887287024901702e-05, "loss": 0.4936, "step": 2404 }, { "epoch": 3.1294729993493817, "grad_norm": 1.0007126331329346, "learning_rate": 2.185452162516383e-05, "loss": 0.7674, "step": 2405 }, { "epoch": 3.130774235523748, "grad_norm": 1.3016561269760132, "learning_rate": 2.182175622542595e-05, "loss": 0.4431, "step": 2406 }, { "epoch": 3.1320754716981134, "grad_norm": 0.9661433696746826, "learning_rate": 2.1788990825688073e-05, "loss": 0.1068, "step": 2407 }, { "epoch": 3.133376707872479, "grad_norm": 1.1545472145080566, "learning_rate": 2.1756225425950198e-05, "loss": 0.8085, "step": 2408 }, { "epoch": 3.1346779440468446, "grad_norm": 1.1240206956863403, "learning_rate": 2.172346002621232e-05, "loss": 1.4929, "step": 2409 }, { "epoch": 3.1359791802212102, "grad_norm": 1.0363967418670654, "learning_rate": 2.1690694626474444e-05, "loss": 0.3529, "step": 2410 }, { "epoch": 3.137280416395576, "grad_norm": 1.3930280208587646, "learning_rate": 2.165792922673657e-05, "loss": 0.3261, "step": 2411 }, { "epoch": 3.1385816525699415, "grad_norm": 1.2244817018508911, "learning_rate": 2.162516382699869e-05, "loss": 0.2691, "step": 2412 }, { "epoch": 3.139882888744307, "grad_norm": 1.2625492811203003, "learning_rate": 2.159239842726081e-05, "loss": 0.7914, "step": 2413 }, { "epoch": 3.1411841249186727, "grad_norm": 0.9568580389022827, "learning_rate": 2.155963302752294e-05, "loss": 0.7166, "step": 2414 }, { "epoch": 3.1424853610930383, "grad_norm": 1.243276834487915, "learning_rate": 2.152686762778506e-05, "loss": 0.3808, "step": 2415 }, { "epoch": 3.143786597267404, "grad_norm": 1.1889079809188843, "learning_rate": 2.1494102228047182e-05, "loss": 0.5657, "step": 2416 }, { "epoch": 3.1450878334417696, "grad_norm": 1.0551782846450806, "learning_rate": 2.1461336828309307e-05, "loss": 1.1918, "step": 2417 }, { "epoch": 3.146389069616135, "grad_norm": 1.1028246879577637, "learning_rate": 2.1428571428571428e-05, "loss": 0.8764, "step": 2418 }, { "epoch": 3.147690305790501, "grad_norm": 1.2319958209991455, "learning_rate": 2.1395806028833553e-05, "loss": 0.6994, "step": 2419 }, { "epoch": 3.1489915419648664, "grad_norm": 1.1947638988494873, "learning_rate": 2.1363040629095678e-05, "loss": 0.6532, "step": 2420 }, { "epoch": 3.1502927781392325, "grad_norm": 1.0889928340911865, "learning_rate": 2.13302752293578e-05, "loss": 0.6956, "step": 2421 }, { "epoch": 3.151594014313598, "grad_norm": 1.1046183109283447, "learning_rate": 2.129750982961992e-05, "loss": 0.7278, "step": 2422 }, { "epoch": 3.1528952504879637, "grad_norm": 1.107011318206787, "learning_rate": 2.126474442988205e-05, "loss": 1.0231, "step": 2423 }, { "epoch": 3.1541964866623293, "grad_norm": 1.2895599603652954, "learning_rate": 2.123197903014417e-05, "loss": 0.3571, "step": 2424 }, { "epoch": 3.155497722836695, "grad_norm": 0.9432070851325989, "learning_rate": 2.119921363040629e-05, "loss": 1.2632, "step": 2425 }, { "epoch": 3.1567989590110606, "grad_norm": 1.4628534317016602, "learning_rate": 2.1166448230668416e-05, "loss": 1.0779, "step": 2426 }, { "epoch": 3.158100195185426, "grad_norm": 1.1061464548110962, "learning_rate": 2.1133682830930537e-05, "loss": 0.3083, "step": 2427 }, { "epoch": 3.159401431359792, "grad_norm": 1.2631947994232178, "learning_rate": 2.1100917431192662e-05, "loss": 0.649, "step": 2428 }, { "epoch": 3.1607026675341574, "grad_norm": 0.8899189829826355, "learning_rate": 2.1068152031454787e-05, "loss": 0.6189, "step": 2429 }, { "epoch": 3.162003903708523, "grad_norm": 1.0370689630508423, "learning_rate": 2.1035386631716908e-05, "loss": 0.7834, "step": 2430 }, { "epoch": 3.1633051398828886, "grad_norm": 1.383809208869934, "learning_rate": 2.100262123197903e-05, "loss": 0.5654, "step": 2431 }, { "epoch": 3.1646063760572543, "grad_norm": 1.4112236499786377, "learning_rate": 2.0969855832241154e-05, "loss": 0.4304, "step": 2432 }, { "epoch": 3.16590761223162, "grad_norm": 1.1036996841430664, "learning_rate": 2.093709043250328e-05, "loss": 0.7034, "step": 2433 }, { "epoch": 3.1672088484059855, "grad_norm": 1.1233173608779907, "learning_rate": 2.09043250327654e-05, "loss": 1.0658, "step": 2434 }, { "epoch": 3.168510084580351, "grad_norm": 0.9838454127311707, "learning_rate": 2.0871559633027525e-05, "loss": 0.5658, "step": 2435 }, { "epoch": 3.169811320754717, "grad_norm": 1.461228847503662, "learning_rate": 2.0838794233289646e-05, "loss": 0.3547, "step": 2436 }, { "epoch": 3.171112556929083, "grad_norm": 1.0328741073608398, "learning_rate": 2.080602883355177e-05, "loss": 0.4595, "step": 2437 }, { "epoch": 3.1724137931034484, "grad_norm": 1.0163720846176147, "learning_rate": 2.0773263433813896e-05, "loss": 1.1421, "step": 2438 }, { "epoch": 3.173715029277814, "grad_norm": 1.3500219583511353, "learning_rate": 2.0740498034076017e-05, "loss": 0.8865, "step": 2439 }, { "epoch": 3.1750162654521796, "grad_norm": 1.3257821798324585, "learning_rate": 2.070773263433814e-05, "loss": 0.4079, "step": 2440 }, { "epoch": 3.1763175016265452, "grad_norm": 0.983088493347168, "learning_rate": 2.0674967234600263e-05, "loss": 1.586, "step": 2441 }, { "epoch": 3.177618737800911, "grad_norm": 1.4123256206512451, "learning_rate": 2.0642201834862388e-05, "loss": 0.597, "step": 2442 }, { "epoch": 3.1789199739752765, "grad_norm": 2.1609888076782227, "learning_rate": 2.060943643512451e-05, "loss": 0.2242, "step": 2443 }, { "epoch": 3.180221210149642, "grad_norm": 1.2107707262039185, "learning_rate": 2.0576671035386634e-05, "loss": 0.5022, "step": 2444 }, { "epoch": 3.1815224463240077, "grad_norm": 1.09236741065979, "learning_rate": 2.0543905635648755e-05, "loss": 0.6928, "step": 2445 }, { "epoch": 3.1828236824983733, "grad_norm": 1.256871223449707, "learning_rate": 2.0511140235910877e-05, "loss": 0.6438, "step": 2446 }, { "epoch": 3.184124918672739, "grad_norm": 1.1375051736831665, "learning_rate": 2.0478374836173005e-05, "loss": 1.0029, "step": 2447 }, { "epoch": 3.1854261548471046, "grad_norm": 1.3802268505096436, "learning_rate": 2.0445609436435126e-05, "loss": 0.833, "step": 2448 }, { "epoch": 3.1867273910214706, "grad_norm": 1.5490180253982544, "learning_rate": 2.0412844036697248e-05, "loss": 0.7223, "step": 2449 }, { "epoch": 3.1880286271958362, "grad_norm": 1.4468603134155273, "learning_rate": 2.0380078636959372e-05, "loss": 0.5993, "step": 2450 }, { "epoch": 3.189329863370202, "grad_norm": 1.2750144004821777, "learning_rate": 2.0347313237221497e-05, "loss": 0.4895, "step": 2451 }, { "epoch": 3.1906310995445675, "grad_norm": 1.0780305862426758, "learning_rate": 2.031454783748362e-05, "loss": 0.4956, "step": 2452 }, { "epoch": 3.191932335718933, "grad_norm": 1.1072388887405396, "learning_rate": 2.0281782437745743e-05, "loss": 0.2765, "step": 2453 }, { "epoch": 3.1932335718932987, "grad_norm": 1.2294474840164185, "learning_rate": 2.0249017038007864e-05, "loss": 0.4122, "step": 2454 }, { "epoch": 3.1945348080676643, "grad_norm": 1.2333407402038574, "learning_rate": 2.0216251638269986e-05, "loss": 0.514, "step": 2455 }, { "epoch": 3.19583604424203, "grad_norm": 1.2324448823928833, "learning_rate": 2.018348623853211e-05, "loss": 0.5063, "step": 2456 }, { "epoch": 3.1971372804163956, "grad_norm": 1.118080973625183, "learning_rate": 2.0150720838794235e-05, "loss": 0.2871, "step": 2457 }, { "epoch": 3.198438516590761, "grad_norm": 1.1127344369888306, "learning_rate": 2.0117955439056357e-05, "loss": 0.4595, "step": 2458 }, { "epoch": 3.199739752765127, "grad_norm": 0.9592211246490479, "learning_rate": 2.008519003931848e-05, "loss": 0.8092, "step": 2459 }, { "epoch": 3.2010409889394924, "grad_norm": 1.1019593477249146, "learning_rate": 2.0052424639580603e-05, "loss": 0.6058, "step": 2460 }, { "epoch": 3.202342225113858, "grad_norm": 0.7721280455589294, "learning_rate": 2.0019659239842727e-05, "loss": 0.4256, "step": 2461 }, { "epoch": 3.2036434612882236, "grad_norm": 1.2579761743545532, "learning_rate": 1.9986893840104852e-05, "loss": 0.9112, "step": 2462 }, { "epoch": 3.2049446974625893, "grad_norm": 1.505386233329773, "learning_rate": 1.9954128440366974e-05, "loss": 0.7185, "step": 2463 }, { "epoch": 3.2062459336369553, "grad_norm": 1.4977048635482788, "learning_rate": 1.9921363040629095e-05, "loss": 0.3884, "step": 2464 }, { "epoch": 3.207547169811321, "grad_norm": 1.4203286170959473, "learning_rate": 1.988859764089122e-05, "loss": 0.8514, "step": 2465 }, { "epoch": 3.2088484059856865, "grad_norm": 1.218504548072815, "learning_rate": 1.9855832241153344e-05, "loss": 0.4762, "step": 2466 }, { "epoch": 3.210149642160052, "grad_norm": 0.908968448638916, "learning_rate": 1.9823066841415466e-05, "loss": 1.0471, "step": 2467 }, { "epoch": 3.211450878334418, "grad_norm": 1.2638041973114014, "learning_rate": 1.979030144167759e-05, "loss": 0.258, "step": 2468 }, { "epoch": 3.2127521145087834, "grad_norm": 1.327499508857727, "learning_rate": 1.9757536041939712e-05, "loss": 0.9271, "step": 2469 }, { "epoch": 3.214053350683149, "grad_norm": 1.6284979581832886, "learning_rate": 1.9724770642201837e-05, "loss": 0.6937, "step": 2470 }, { "epoch": 3.2153545868575146, "grad_norm": 1.3890019655227661, "learning_rate": 1.9692005242463958e-05, "loss": 0.6177, "step": 2471 }, { "epoch": 3.2166558230318802, "grad_norm": 1.6823475360870361, "learning_rate": 1.9659239842726083e-05, "loss": 0.7332, "step": 2472 }, { "epoch": 3.217957059206246, "grad_norm": 1.2302037477493286, "learning_rate": 1.9626474442988204e-05, "loss": 0.3353, "step": 2473 }, { "epoch": 3.2192582953806115, "grad_norm": 1.2247365713119507, "learning_rate": 1.959370904325033e-05, "loss": 0.8997, "step": 2474 }, { "epoch": 3.220559531554977, "grad_norm": 1.3507542610168457, "learning_rate": 1.9560943643512453e-05, "loss": 0.7347, "step": 2475 }, { "epoch": 3.2218607677293427, "grad_norm": 1.0540045499801636, "learning_rate": 1.9528178243774575e-05, "loss": 0.2373, "step": 2476 }, { "epoch": 3.2231620039037083, "grad_norm": 1.123458981513977, "learning_rate": 1.94954128440367e-05, "loss": 0.3842, "step": 2477 }, { "epoch": 3.224463240078074, "grad_norm": 1.2988686561584473, "learning_rate": 1.946264744429882e-05, "loss": 0.8592, "step": 2478 }, { "epoch": 3.22576447625244, "grad_norm": 0.9898180961608887, "learning_rate": 1.9429882044560946e-05, "loss": 0.9577, "step": 2479 }, { "epoch": 3.2270657124268056, "grad_norm": 0.9997983574867249, "learning_rate": 1.9397116644823067e-05, "loss": 0.3827, "step": 2480 }, { "epoch": 3.2283669486011712, "grad_norm": 1.523417353630066, "learning_rate": 1.9364351245085192e-05, "loss": 0.583, "step": 2481 }, { "epoch": 3.229668184775537, "grad_norm": 1.1093852519989014, "learning_rate": 1.9331585845347313e-05, "loss": 0.8762, "step": 2482 }, { "epoch": 3.2309694209499025, "grad_norm": 1.4472417831420898, "learning_rate": 1.9298820445609438e-05, "loss": 0.5769, "step": 2483 }, { "epoch": 3.232270657124268, "grad_norm": 1.2662935256958008, "learning_rate": 1.9266055045871563e-05, "loss": 0.4873, "step": 2484 }, { "epoch": 3.2335718932986337, "grad_norm": 1.0346791744232178, "learning_rate": 1.9233289646133684e-05, "loss": 1.1118, "step": 2485 }, { "epoch": 3.2348731294729993, "grad_norm": 1.4210716485977173, "learning_rate": 1.9200524246395805e-05, "loss": 0.2987, "step": 2486 }, { "epoch": 3.236174365647365, "grad_norm": 1.032605528831482, "learning_rate": 1.916775884665793e-05, "loss": 0.3136, "step": 2487 }, { "epoch": 3.2374756018217306, "grad_norm": 1.0953253507614136, "learning_rate": 1.913499344692005e-05, "loss": 0.5687, "step": 2488 }, { "epoch": 3.238776837996096, "grad_norm": 1.2218925952911377, "learning_rate": 1.9102228047182176e-05, "loss": 0.8138, "step": 2489 }, { "epoch": 3.240078074170462, "grad_norm": 1.368369221687317, "learning_rate": 1.90694626474443e-05, "loss": 0.7313, "step": 2490 }, { "epoch": 3.2413793103448274, "grad_norm": 1.2327080965042114, "learning_rate": 1.9036697247706422e-05, "loss": 0.928, "step": 2491 }, { "epoch": 3.2426805465191935, "grad_norm": 1.476767659187317, "learning_rate": 1.9003931847968547e-05, "loss": 0.8462, "step": 2492 }, { "epoch": 3.243981782693559, "grad_norm": 1.492253065109253, "learning_rate": 1.897116644823067e-05, "loss": 0.2966, "step": 2493 }, { "epoch": 3.2452830188679247, "grad_norm": 1.6782037019729614, "learning_rate": 1.8938401048492793e-05, "loss": 0.5705, "step": 2494 }, { "epoch": 3.2465842550422903, "grad_norm": 1.2197721004486084, "learning_rate": 1.8905635648754914e-05, "loss": 0.3263, "step": 2495 }, { "epoch": 3.247885491216656, "grad_norm": 1.371960997581482, "learning_rate": 1.887287024901704e-05, "loss": 0.4376, "step": 2496 }, { "epoch": 3.2491867273910215, "grad_norm": 1.1334913969039917, "learning_rate": 1.884010484927916e-05, "loss": 0.6292, "step": 2497 }, { "epoch": 3.250487963565387, "grad_norm": 1.093039631843567, "learning_rate": 1.8807339449541285e-05, "loss": 1.011, "step": 2498 }, { "epoch": 3.251789199739753, "grad_norm": 1.1340581178665161, "learning_rate": 1.877457404980341e-05, "loss": 0.9706, "step": 2499 }, { "epoch": 3.2530904359141184, "grad_norm": 0.9342528581619263, "learning_rate": 1.874180865006553e-05, "loss": 0.9951, "step": 2500 }, { "epoch": 3.254391672088484, "grad_norm": 1.5238959789276123, "learning_rate": 1.8709043250327653e-05, "loss": 0.4155, "step": 2501 }, { "epoch": 3.2556929082628496, "grad_norm": 1.2588598728179932, "learning_rate": 1.8676277850589777e-05, "loss": 0.5185, "step": 2502 }, { "epoch": 3.2569941444372152, "grad_norm": 1.252254605293274, "learning_rate": 1.8643512450851902e-05, "loss": 0.4589, "step": 2503 }, { "epoch": 3.258295380611581, "grad_norm": 1.1743361949920654, "learning_rate": 1.8610747051114023e-05, "loss": 0.6107, "step": 2504 }, { "epoch": 3.2595966167859465, "grad_norm": 1.133967638015747, "learning_rate": 1.8577981651376148e-05, "loss": 0.1374, "step": 2505 }, { "epoch": 3.260897852960312, "grad_norm": 1.4746960401535034, "learning_rate": 1.854521625163827e-05, "loss": 0.7327, "step": 2506 }, { "epoch": 3.262199089134678, "grad_norm": 1.271582841873169, "learning_rate": 1.8512450851900394e-05, "loss": 0.4025, "step": 2507 }, { "epoch": 3.2635003253090438, "grad_norm": 1.4512951374053955, "learning_rate": 1.847968545216252e-05, "loss": 0.7559, "step": 2508 }, { "epoch": 3.2648015614834094, "grad_norm": 1.018992304801941, "learning_rate": 1.844692005242464e-05, "loss": 1.0631, "step": 2509 }, { "epoch": 3.266102797657775, "grad_norm": 1.0581026077270508, "learning_rate": 1.841415465268676e-05, "loss": 0.8774, "step": 2510 }, { "epoch": 3.2674040338321406, "grad_norm": 1.1539779901504517, "learning_rate": 1.8381389252948886e-05, "loss": 0.8892, "step": 2511 }, { "epoch": 3.2687052700065062, "grad_norm": 1.0932186841964722, "learning_rate": 1.834862385321101e-05, "loss": 0.7724, "step": 2512 }, { "epoch": 3.270006506180872, "grad_norm": 1.2045902013778687, "learning_rate": 1.8315858453473132e-05, "loss": 0.6792, "step": 2513 }, { "epoch": 3.2713077423552375, "grad_norm": 1.4324183464050293, "learning_rate": 1.8283093053735257e-05, "loss": 0.3974, "step": 2514 }, { "epoch": 3.272608978529603, "grad_norm": 0.9114199280738831, "learning_rate": 1.825032765399738e-05, "loss": 0.7686, "step": 2515 }, { "epoch": 3.2739102147039687, "grad_norm": 1.4731467962265015, "learning_rate": 1.82175622542595e-05, "loss": 0.3729, "step": 2516 }, { "epoch": 3.2752114508783343, "grad_norm": 1.354844570159912, "learning_rate": 1.8184796854521628e-05, "loss": 0.9278, "step": 2517 }, { "epoch": 3.2765126870527, "grad_norm": 1.3490159511566162, "learning_rate": 1.815203145478375e-05, "loss": 0.9728, "step": 2518 }, { "epoch": 3.2778139232270656, "grad_norm": 1.235819935798645, "learning_rate": 1.811926605504587e-05, "loss": 0.6212, "step": 2519 }, { "epoch": 3.279115159401431, "grad_norm": 1.9447908401489258, "learning_rate": 1.8086500655307995e-05, "loss": 0.4987, "step": 2520 }, { "epoch": 3.280416395575797, "grad_norm": 1.2526427507400513, "learning_rate": 1.805373525557012e-05, "loss": 0.2358, "step": 2521 }, { "epoch": 3.281717631750163, "grad_norm": 1.4363075494766235, "learning_rate": 1.802096985583224e-05, "loss": 0.7388, "step": 2522 }, { "epoch": 3.2830188679245285, "grad_norm": 1.30640709400177, "learning_rate": 1.7988204456094366e-05, "loss": 0.5853, "step": 2523 }, { "epoch": 3.284320104098894, "grad_norm": 1.0671273469924927, "learning_rate": 1.7955439056356488e-05, "loss": 0.5414, "step": 2524 }, { "epoch": 3.2856213402732597, "grad_norm": 1.2096980810165405, "learning_rate": 1.792267365661861e-05, "loss": 0.2981, "step": 2525 }, { "epoch": 3.2869225764476253, "grad_norm": 1.3022063970565796, "learning_rate": 1.7889908256880737e-05, "loss": 0.7254, "step": 2526 }, { "epoch": 3.288223812621991, "grad_norm": 1.1199564933776855, "learning_rate": 1.785714285714286e-05, "loss": 0.9733, "step": 2527 }, { "epoch": 3.2895250487963565, "grad_norm": 1.1671794652938843, "learning_rate": 1.782437745740498e-05, "loss": 0.4495, "step": 2528 }, { "epoch": 3.290826284970722, "grad_norm": 1.1408313512802124, "learning_rate": 1.7791612057667105e-05, "loss": 1.0642, "step": 2529 }, { "epoch": 3.2921275211450878, "grad_norm": 1.2137606143951416, "learning_rate": 1.775884665792923e-05, "loss": 0.59, "step": 2530 }, { "epoch": 3.2934287573194534, "grad_norm": 1.3647676706314087, "learning_rate": 1.772608125819135e-05, "loss": 1.0874, "step": 2531 }, { "epoch": 3.294729993493819, "grad_norm": 1.2265632152557373, "learning_rate": 1.7693315858453475e-05, "loss": 0.9848, "step": 2532 }, { "epoch": 3.2960312296681846, "grad_norm": 1.0388823747634888, "learning_rate": 1.7660550458715597e-05, "loss": 1.5054, "step": 2533 }, { "epoch": 3.2973324658425502, "grad_norm": 1.3064407110214233, "learning_rate": 1.7627785058977718e-05, "loss": 0.9716, "step": 2534 }, { "epoch": 3.2986337020169163, "grad_norm": 1.13992178440094, "learning_rate": 1.7595019659239846e-05, "loss": 1.0017, "step": 2535 }, { "epoch": 3.2999349381912815, "grad_norm": 0.8696495890617371, "learning_rate": 1.7562254259501968e-05, "loss": 0.5519, "step": 2536 }, { "epoch": 3.3012361743656475, "grad_norm": 1.4377583265304565, "learning_rate": 1.752948885976409e-05, "loss": 0.445, "step": 2537 }, { "epoch": 3.302537410540013, "grad_norm": 1.2635341882705688, "learning_rate": 1.7496723460026214e-05, "loss": 1.0404, "step": 2538 }, { "epoch": 3.3038386467143788, "grad_norm": 1.3187878131866455, "learning_rate": 1.7463958060288335e-05, "loss": 0.3426, "step": 2539 }, { "epoch": 3.3051398828887444, "grad_norm": 0.9361656904220581, "learning_rate": 1.743119266055046e-05, "loss": 0.7809, "step": 2540 }, { "epoch": 3.30644111906311, "grad_norm": 1.1163005828857422, "learning_rate": 1.7398427260812584e-05, "loss": 0.5151, "step": 2541 }, { "epoch": 3.3077423552374756, "grad_norm": 1.2410955429077148, "learning_rate": 1.7365661861074706e-05, "loss": 1.1771, "step": 2542 }, { "epoch": 3.3090435914118412, "grad_norm": 1.2477842569351196, "learning_rate": 1.7332896461336827e-05, "loss": 0.9611, "step": 2543 }, { "epoch": 3.310344827586207, "grad_norm": 1.4302699565887451, "learning_rate": 1.7300131061598955e-05, "loss": 0.3223, "step": 2544 }, { "epoch": 3.3116460637605725, "grad_norm": 1.1427030563354492, "learning_rate": 1.7267365661861077e-05, "loss": 0.7456, "step": 2545 }, { "epoch": 3.312947299934938, "grad_norm": 1.0666667222976685, "learning_rate": 1.7234600262123198e-05, "loss": 0.6221, "step": 2546 }, { "epoch": 3.3142485361093037, "grad_norm": 1.0812749862670898, "learning_rate": 1.7201834862385323e-05, "loss": 0.9032, "step": 2547 }, { "epoch": 3.3155497722836693, "grad_norm": 1.0736185312271118, "learning_rate": 1.7169069462647444e-05, "loss": 0.8589, "step": 2548 }, { "epoch": 3.316851008458035, "grad_norm": 0.9936012625694275, "learning_rate": 1.713630406290957e-05, "loss": 0.9049, "step": 2549 }, { "epoch": 3.318152244632401, "grad_norm": 1.1488243341445923, "learning_rate": 1.7103538663171694e-05, "loss": 0.7236, "step": 2550 }, { "epoch": 3.3194534808067666, "grad_norm": 1.0767332315444946, "learning_rate": 1.7070773263433815e-05, "loss": 0.7912, "step": 2551 }, { "epoch": 3.3207547169811322, "grad_norm": 1.2073698043823242, "learning_rate": 1.7038007863695936e-05, "loss": 0.865, "step": 2552 }, { "epoch": 3.322055953155498, "grad_norm": 1.3459885120391846, "learning_rate": 1.700524246395806e-05, "loss": 0.404, "step": 2553 }, { "epoch": 3.3233571893298635, "grad_norm": 1.3921170234680176, "learning_rate": 1.6972477064220186e-05, "loss": 0.3559, "step": 2554 }, { "epoch": 3.324658425504229, "grad_norm": 1.3103896379470825, "learning_rate": 1.6939711664482307e-05, "loss": 0.3179, "step": 2555 }, { "epoch": 3.3259596616785947, "grad_norm": 0.9860579967498779, "learning_rate": 1.6906946264744432e-05, "loss": 0.4479, "step": 2556 }, { "epoch": 3.3272608978529603, "grad_norm": 1.2772053480148315, "learning_rate": 1.6874180865006553e-05, "loss": 0.6463, "step": 2557 }, { "epoch": 3.328562134027326, "grad_norm": 0.9959273338317871, "learning_rate": 1.6841415465268678e-05, "loss": 0.3281, "step": 2558 }, { "epoch": 3.3298633702016915, "grad_norm": 1.143431305885315, "learning_rate": 1.6808650065530803e-05, "loss": 0.3296, "step": 2559 }, { "epoch": 3.331164606376057, "grad_norm": 1.436946153640747, "learning_rate": 1.6775884665792924e-05, "loss": 0.4189, "step": 2560 }, { "epoch": 3.3324658425504228, "grad_norm": 1.0519565343856812, "learning_rate": 1.6743119266055045e-05, "loss": 0.7346, "step": 2561 }, { "epoch": 3.3337670787247884, "grad_norm": 1.7606933116912842, "learning_rate": 1.671035386631717e-05, "loss": 0.6751, "step": 2562 }, { "epoch": 3.335068314899154, "grad_norm": 1.1092280149459839, "learning_rate": 1.6677588466579295e-05, "loss": 0.2799, "step": 2563 }, { "epoch": 3.3363695510735196, "grad_norm": 1.263614535331726, "learning_rate": 1.6644823066841416e-05, "loss": 0.2134, "step": 2564 }, { "epoch": 3.3376707872478857, "grad_norm": 1.0859283208847046, "learning_rate": 1.661205766710354e-05, "loss": 0.5046, "step": 2565 }, { "epoch": 3.3389720234222513, "grad_norm": 1.0192055702209473, "learning_rate": 1.6579292267365662e-05, "loss": 0.3892, "step": 2566 }, { "epoch": 3.340273259596617, "grad_norm": 1.018629789352417, "learning_rate": 1.6546526867627784e-05, "loss": 1.1975, "step": 2567 }, { "epoch": 3.3415744957709825, "grad_norm": 1.3225228786468506, "learning_rate": 1.651376146788991e-05, "loss": 0.4916, "step": 2568 }, { "epoch": 3.342875731945348, "grad_norm": 1.4714657068252563, "learning_rate": 1.6480996068152033e-05, "loss": 0.3906, "step": 2569 }, { "epoch": 3.3441769681197138, "grad_norm": 0.9159764051437378, "learning_rate": 1.6448230668414154e-05, "loss": 0.6775, "step": 2570 }, { "epoch": 3.3454782042940794, "grad_norm": 1.1224517822265625, "learning_rate": 1.641546526867628e-05, "loss": 0.8083, "step": 2571 }, { "epoch": 3.346779440468445, "grad_norm": 1.2159911394119263, "learning_rate": 1.6382699868938404e-05, "loss": 0.761, "step": 2572 }, { "epoch": 3.3480806766428106, "grad_norm": 1.507957935333252, "learning_rate": 1.6349934469200525e-05, "loss": 0.719, "step": 2573 }, { "epoch": 3.3493819128171762, "grad_norm": 1.1319146156311035, "learning_rate": 1.631716906946265e-05, "loss": 1.0241, "step": 2574 }, { "epoch": 3.350683148991542, "grad_norm": 1.1937400102615356, "learning_rate": 1.628440366972477e-05, "loss": 0.3934, "step": 2575 }, { "epoch": 3.3519843851659075, "grad_norm": 1.2957446575164795, "learning_rate": 1.6251638269986893e-05, "loss": 1.144, "step": 2576 }, { "epoch": 3.353285621340273, "grad_norm": 1.2963451147079468, "learning_rate": 1.6218872870249017e-05, "loss": 0.9285, "step": 2577 }, { "epoch": 3.354586857514639, "grad_norm": 0.9317249655723572, "learning_rate": 1.6186107470511142e-05, "loss": 0.9162, "step": 2578 }, { "epoch": 3.3558880936890043, "grad_norm": 1.1308162212371826, "learning_rate": 1.6153342070773264e-05, "loss": 0.8295, "step": 2579 }, { "epoch": 3.3571893298633704, "grad_norm": 0.8148242831230164, "learning_rate": 1.6120576671035388e-05, "loss": 0.4135, "step": 2580 }, { "epoch": 3.358490566037736, "grad_norm": 1.2193864583969116, "learning_rate": 1.608781127129751e-05, "loss": 0.9282, "step": 2581 }, { "epoch": 3.3597918022121016, "grad_norm": 1.2662286758422852, "learning_rate": 1.6055045871559634e-05, "loss": 0.4982, "step": 2582 }, { "epoch": 3.3610930383864672, "grad_norm": 1.0979644060134888, "learning_rate": 1.6022280471821756e-05, "loss": 0.7044, "step": 2583 }, { "epoch": 3.362394274560833, "grad_norm": 1.357429027557373, "learning_rate": 1.598951507208388e-05, "loss": 1.1756, "step": 2584 }, { "epoch": 3.3636955107351985, "grad_norm": 1.279432773590088, "learning_rate": 1.5956749672346002e-05, "loss": 0.2734, "step": 2585 }, { "epoch": 3.364996746909564, "grad_norm": 1.1774811744689941, "learning_rate": 1.5923984272608126e-05, "loss": 0.764, "step": 2586 }, { "epoch": 3.3662979830839297, "grad_norm": 1.3673694133758545, "learning_rate": 1.589121887287025e-05, "loss": 0.8495, "step": 2587 }, { "epoch": 3.3675992192582953, "grad_norm": 1.1940855979919434, "learning_rate": 1.5858453473132373e-05, "loss": 0.8073, "step": 2588 }, { "epoch": 3.368900455432661, "grad_norm": 1.0861462354660034, "learning_rate": 1.5825688073394497e-05, "loss": 0.7371, "step": 2589 }, { "epoch": 3.3702016916070265, "grad_norm": 1.2966219186782837, "learning_rate": 1.579292267365662e-05, "loss": 0.4964, "step": 2590 }, { "epoch": 3.371502927781392, "grad_norm": 1.6874120235443115, "learning_rate": 1.5760157273918743e-05, "loss": 0.3716, "step": 2591 }, { "epoch": 3.3728041639557578, "grad_norm": 0.9779640436172485, "learning_rate": 1.5727391874180865e-05, "loss": 1.0211, "step": 2592 }, { "epoch": 3.374105400130124, "grad_norm": 1.0483431816101074, "learning_rate": 1.569462647444299e-05, "loss": 0.5505, "step": 2593 }, { "epoch": 3.3754066363044894, "grad_norm": 0.9676886200904846, "learning_rate": 1.566186107470511e-05, "loss": 0.7513, "step": 2594 }, { "epoch": 3.376707872478855, "grad_norm": 1.2805075645446777, "learning_rate": 1.5629095674967236e-05, "loss": 0.5474, "step": 2595 }, { "epoch": 3.3780091086532207, "grad_norm": 1.4243063926696777, "learning_rate": 1.559633027522936e-05, "loss": 0.6011, "step": 2596 }, { "epoch": 3.3793103448275863, "grad_norm": 1.1852598190307617, "learning_rate": 1.556356487549148e-05, "loss": 0.4238, "step": 2597 }, { "epoch": 3.380611581001952, "grad_norm": 1.284533977508545, "learning_rate": 1.5530799475753603e-05, "loss": 0.4848, "step": 2598 }, { "epoch": 3.3819128171763175, "grad_norm": 1.384884238243103, "learning_rate": 1.5498034076015728e-05, "loss": 0.9388, "step": 2599 }, { "epoch": 3.383214053350683, "grad_norm": 1.4465221166610718, "learning_rate": 1.5465268676277852e-05, "loss": 0.5479, "step": 2600 }, { "epoch": 3.3845152895250488, "grad_norm": 1.1769533157348633, "learning_rate": 1.5432503276539974e-05, "loss": 0.3622, "step": 2601 }, { "epoch": 3.3858165256994144, "grad_norm": 1.1010891199111938, "learning_rate": 1.53997378768021e-05, "loss": 0.5423, "step": 2602 }, { "epoch": 3.38711776187378, "grad_norm": 1.4723896980285645, "learning_rate": 1.536697247706422e-05, "loss": 0.4111, "step": 2603 }, { "epoch": 3.3884189980481456, "grad_norm": 1.2880455255508423, "learning_rate": 1.5334207077326345e-05, "loss": 0.597, "step": 2604 }, { "epoch": 3.3897202342225112, "grad_norm": 1.0237464904785156, "learning_rate": 1.530144167758847e-05, "loss": 1.044, "step": 2605 }, { "epoch": 3.391021470396877, "grad_norm": 1.2046126127243042, "learning_rate": 1.526867627785059e-05, "loss": 0.5956, "step": 2606 }, { "epoch": 3.3923227065712425, "grad_norm": 1.5896433591842651, "learning_rate": 1.5235910878112714e-05, "loss": 0.7895, "step": 2607 }, { "epoch": 3.3936239427456085, "grad_norm": 1.1411288976669312, "learning_rate": 1.5203145478374837e-05, "loss": 0.5447, "step": 2608 }, { "epoch": 3.394925178919974, "grad_norm": 1.376236915588379, "learning_rate": 1.5170380078636958e-05, "loss": 0.4639, "step": 2609 }, { "epoch": 3.3962264150943398, "grad_norm": 1.2666608095169067, "learning_rate": 1.5137614678899085e-05, "loss": 0.5657, "step": 2610 }, { "epoch": 3.3975276512687054, "grad_norm": 0.8902050256729126, "learning_rate": 1.5104849279161206e-05, "loss": 0.6275, "step": 2611 }, { "epoch": 3.398828887443071, "grad_norm": 1.4317795038223267, "learning_rate": 1.5072083879423329e-05, "loss": 0.4337, "step": 2612 }, { "epoch": 3.4001301236174366, "grad_norm": 1.1046252250671387, "learning_rate": 1.5039318479685452e-05, "loss": 0.9253, "step": 2613 }, { "epoch": 3.4014313597918022, "grad_norm": 1.4763599634170532, "learning_rate": 1.5006553079947577e-05, "loss": 0.5731, "step": 2614 }, { "epoch": 3.402732595966168, "grad_norm": 1.3067066669464111, "learning_rate": 1.49737876802097e-05, "loss": 0.8946, "step": 2615 }, { "epoch": 3.4040338321405335, "grad_norm": 1.1767325401306152, "learning_rate": 1.4941022280471823e-05, "loss": 0.697, "step": 2616 }, { "epoch": 3.405335068314899, "grad_norm": 1.049501895904541, "learning_rate": 1.4908256880733946e-05, "loss": 0.2116, "step": 2617 }, { "epoch": 3.4066363044892647, "grad_norm": 1.3739365339279175, "learning_rate": 1.4875491480996067e-05, "loss": 0.2534, "step": 2618 }, { "epoch": 3.4079375406636303, "grad_norm": 1.4168038368225098, "learning_rate": 1.4842726081258194e-05, "loss": 0.6319, "step": 2619 }, { "epoch": 3.409238776837996, "grad_norm": 1.3066325187683105, "learning_rate": 1.4809960681520315e-05, "loss": 0.6185, "step": 2620 }, { "epoch": 3.410540013012362, "grad_norm": 1.496492862701416, "learning_rate": 1.4777195281782438e-05, "loss": 0.3486, "step": 2621 }, { "epoch": 3.411841249186727, "grad_norm": 0.8840242028236389, "learning_rate": 1.4744429882044561e-05, "loss": 0.9859, "step": 2622 }, { "epoch": 3.413142485361093, "grad_norm": 1.5772520303726196, "learning_rate": 1.4711664482306684e-05, "loss": 0.435, "step": 2623 }, { "epoch": 3.414443721535459, "grad_norm": 1.132530689239502, "learning_rate": 1.4678899082568809e-05, "loss": 0.7255, "step": 2624 }, { "epoch": 3.4157449577098244, "grad_norm": 1.7339543104171753, "learning_rate": 1.4646133682830932e-05, "loss": 0.6717, "step": 2625 }, { "epoch": 3.41704619388419, "grad_norm": 1.042712688446045, "learning_rate": 1.4613368283093053e-05, "loss": 0.2252, "step": 2626 }, { "epoch": 3.4183474300585557, "grad_norm": 1.3277219533920288, "learning_rate": 1.4580602883355176e-05, "loss": 0.7489, "step": 2627 }, { "epoch": 3.4196486662329213, "grad_norm": 1.354589581489563, "learning_rate": 1.4547837483617303e-05, "loss": 0.3864, "step": 2628 }, { "epoch": 3.420949902407287, "grad_norm": 1.2601925134658813, "learning_rate": 1.4515072083879424e-05, "loss": 1.2089, "step": 2629 }, { "epoch": 3.4222511385816525, "grad_norm": 1.3952313661575317, "learning_rate": 1.4482306684141547e-05, "loss": 0.5376, "step": 2630 }, { "epoch": 3.423552374756018, "grad_norm": 1.0787806510925293, "learning_rate": 1.444954128440367e-05, "loss": 0.4972, "step": 2631 }, { "epoch": 3.4248536109303838, "grad_norm": 1.2269136905670166, "learning_rate": 1.4416775884665793e-05, "loss": 0.2114, "step": 2632 }, { "epoch": 3.4261548471047494, "grad_norm": 1.2916500568389893, "learning_rate": 1.4384010484927918e-05, "loss": 0.8367, "step": 2633 }, { "epoch": 3.427456083279115, "grad_norm": 1.018850564956665, "learning_rate": 1.4351245085190041e-05, "loss": 0.7633, "step": 2634 }, { "epoch": 3.4287573194534806, "grad_norm": 1.2440516948699951, "learning_rate": 1.4318479685452162e-05, "loss": 0.9372, "step": 2635 }, { "epoch": 3.4300585556278467, "grad_norm": 1.2787904739379883, "learning_rate": 1.4285714285714285e-05, "loss": 0.355, "step": 2636 }, { "epoch": 3.4313597918022123, "grad_norm": 1.3705238103866577, "learning_rate": 1.425294888597641e-05, "loss": 0.5337, "step": 2637 }, { "epoch": 3.432661027976578, "grad_norm": 1.058925747871399, "learning_rate": 1.4220183486238533e-05, "loss": 1.0506, "step": 2638 }, { "epoch": 3.4339622641509435, "grad_norm": 1.6658560037612915, "learning_rate": 1.4187418086500656e-05, "loss": 0.3877, "step": 2639 }, { "epoch": 3.435263500325309, "grad_norm": 1.5602723360061646, "learning_rate": 1.415465268676278e-05, "loss": 0.8654, "step": 2640 }, { "epoch": 3.4365647364996748, "grad_norm": 1.5793238878250122, "learning_rate": 1.41218872870249e-05, "loss": 0.9364, "step": 2641 }, { "epoch": 3.4378659726740404, "grad_norm": 1.451010823249817, "learning_rate": 1.4089121887287027e-05, "loss": 0.7556, "step": 2642 }, { "epoch": 3.439167208848406, "grad_norm": 1.037593960762024, "learning_rate": 1.405635648754915e-05, "loss": 0.8917, "step": 2643 }, { "epoch": 3.4404684450227716, "grad_norm": 1.1538320779800415, "learning_rate": 1.4023591087811271e-05, "loss": 1.1556, "step": 2644 }, { "epoch": 3.441769681197137, "grad_norm": 0.926427960395813, "learning_rate": 1.3990825688073395e-05, "loss": 0.6597, "step": 2645 }, { "epoch": 3.443070917371503, "grad_norm": 1.2711085081100464, "learning_rate": 1.3958060288335518e-05, "loss": 0.4171, "step": 2646 }, { "epoch": 3.4443721535458685, "grad_norm": 0.8655628561973572, "learning_rate": 1.3925294888597642e-05, "loss": 0.78, "step": 2647 }, { "epoch": 3.445673389720234, "grad_norm": 0.8572363257408142, "learning_rate": 1.3892529488859765e-05, "loss": 0.4992, "step": 2648 }, { "epoch": 3.4469746258945997, "grad_norm": 1.0560927391052246, "learning_rate": 1.3859764089121888e-05, "loss": 0.5729, "step": 2649 }, { "epoch": 3.4482758620689653, "grad_norm": 1.220673680305481, "learning_rate": 1.382699868938401e-05, "loss": 0.9718, "step": 2650 }, { "epoch": 3.4495770982433314, "grad_norm": 0.9774724841117859, "learning_rate": 1.3794233289646136e-05, "loss": 0.6243, "step": 2651 }, { "epoch": 3.450878334417697, "grad_norm": 0.8217886686325073, "learning_rate": 1.3761467889908258e-05, "loss": 1.1363, "step": 2652 }, { "epoch": 3.4521795705920626, "grad_norm": 1.3839573860168457, "learning_rate": 1.372870249017038e-05, "loss": 0.4344, "step": 2653 }, { "epoch": 3.453480806766428, "grad_norm": 1.2022353410720825, "learning_rate": 1.3695937090432504e-05, "loss": 0.7595, "step": 2654 }, { "epoch": 3.454782042940794, "grad_norm": 1.1528173685073853, "learning_rate": 1.3663171690694627e-05, "loss": 0.6471, "step": 2655 }, { "epoch": 3.4560832791151594, "grad_norm": 1.2933878898620605, "learning_rate": 1.3630406290956751e-05, "loss": 0.6884, "step": 2656 }, { "epoch": 3.457384515289525, "grad_norm": 0.9615000486373901, "learning_rate": 1.3597640891218874e-05, "loss": 1.0155, "step": 2657 }, { "epoch": 3.4586857514638907, "grad_norm": 1.1620204448699951, "learning_rate": 1.3564875491480997e-05, "loss": 0.3484, "step": 2658 }, { "epoch": 3.4599869876382563, "grad_norm": 1.1926933526992798, "learning_rate": 1.3532110091743119e-05, "loss": 0.2873, "step": 2659 }, { "epoch": 3.461288223812622, "grad_norm": 1.2384406328201294, "learning_rate": 1.3499344692005242e-05, "loss": 0.7218, "step": 2660 }, { "epoch": 3.4625894599869875, "grad_norm": 1.5945106744766235, "learning_rate": 1.3466579292267367e-05, "loss": 0.4893, "step": 2661 }, { "epoch": 3.463890696161353, "grad_norm": 1.405644178390503, "learning_rate": 1.343381389252949e-05, "loss": 0.5279, "step": 2662 }, { "epoch": 3.4651919323357188, "grad_norm": 1.0757676362991333, "learning_rate": 1.3401048492791613e-05, "loss": 0.3978, "step": 2663 }, { "epoch": 3.466493168510085, "grad_norm": 0.8614431023597717, "learning_rate": 1.3368283093053736e-05, "loss": 0.2559, "step": 2664 }, { "epoch": 3.46779440468445, "grad_norm": 1.413575291633606, "learning_rate": 1.333551769331586e-05, "loss": 0.4985, "step": 2665 }, { "epoch": 3.469095640858816, "grad_norm": 1.1096311807632446, "learning_rate": 1.3302752293577984e-05, "loss": 0.3016, "step": 2666 }, { "epoch": 3.4703968770331817, "grad_norm": 1.2793371677398682, "learning_rate": 1.3269986893840105e-05, "loss": 0.1994, "step": 2667 }, { "epoch": 3.4716981132075473, "grad_norm": 1.54989492893219, "learning_rate": 1.3237221494102228e-05, "loss": 0.4064, "step": 2668 }, { "epoch": 3.472999349381913, "grad_norm": 0.7178281545639038, "learning_rate": 1.3204456094364351e-05, "loss": 0.4918, "step": 2669 }, { "epoch": 3.4743005855562785, "grad_norm": 1.3376415967941284, "learning_rate": 1.3171690694626476e-05, "loss": 0.5251, "step": 2670 }, { "epoch": 3.475601821730644, "grad_norm": 0.9500694870948792, "learning_rate": 1.3138925294888599e-05, "loss": 1.1603, "step": 2671 }, { "epoch": 3.4769030579050098, "grad_norm": 1.1129080057144165, "learning_rate": 1.3106159895150722e-05, "loss": 0.3998, "step": 2672 }, { "epoch": 3.4782042940793754, "grad_norm": 1.0943604707717896, "learning_rate": 1.3073394495412845e-05, "loss": 1.0168, "step": 2673 }, { "epoch": 3.479505530253741, "grad_norm": 1.4157041311264038, "learning_rate": 1.3040629095674966e-05, "loss": 0.7765, "step": 2674 }, { "epoch": 3.4808067664281066, "grad_norm": 1.4426865577697754, "learning_rate": 1.3007863695937093e-05, "loss": 0.5166, "step": 2675 }, { "epoch": 3.482108002602472, "grad_norm": 0.8959588408470154, "learning_rate": 1.2975098296199214e-05, "loss": 0.5976, "step": 2676 }, { "epoch": 3.483409238776838, "grad_norm": 1.3297231197357178, "learning_rate": 1.2942332896461337e-05, "loss": 0.6723, "step": 2677 }, { "epoch": 3.4847104749512035, "grad_norm": 1.3978371620178223, "learning_rate": 1.290956749672346e-05, "loss": 0.4175, "step": 2678 }, { "epoch": 3.4860117111255695, "grad_norm": 1.0897712707519531, "learning_rate": 1.2876802096985585e-05, "loss": 1.1185, "step": 2679 }, { "epoch": 3.487312947299935, "grad_norm": 1.1075831651687622, "learning_rate": 1.2844036697247708e-05, "loss": 0.3232, "step": 2680 }, { "epoch": 3.4886141834743007, "grad_norm": 1.1132524013519287, "learning_rate": 1.281127129750983e-05, "loss": 0.2174, "step": 2681 }, { "epoch": 3.4899154196486664, "grad_norm": 1.1477892398834229, "learning_rate": 1.2778505897771952e-05, "loss": 0.1509, "step": 2682 }, { "epoch": 3.491216655823032, "grad_norm": 1.078464150428772, "learning_rate": 1.2745740498034075e-05, "loss": 0.7914, "step": 2683 }, { "epoch": 3.4925178919973976, "grad_norm": 1.2315900325775146, "learning_rate": 1.2712975098296202e-05, "loss": 0.8897, "step": 2684 }, { "epoch": 3.493819128171763, "grad_norm": 1.031937837600708, "learning_rate": 1.2680209698558323e-05, "loss": 0.4771, "step": 2685 }, { "epoch": 3.495120364346129, "grad_norm": 0.891343891620636, "learning_rate": 1.2647444298820446e-05, "loss": 0.7938, "step": 2686 }, { "epoch": 3.4964216005204944, "grad_norm": 1.2748843431472778, "learning_rate": 1.2614678899082569e-05, "loss": 0.5235, "step": 2687 }, { "epoch": 3.49772283669486, "grad_norm": 1.3400328159332275, "learning_rate": 1.2581913499344692e-05, "loss": 0.7001, "step": 2688 }, { "epoch": 3.4990240728692257, "grad_norm": 1.4430938959121704, "learning_rate": 1.2549148099606817e-05, "loss": 1.0356, "step": 2689 }, { "epoch": 3.5003253090435913, "grad_norm": 1.1347033977508545, "learning_rate": 1.251638269986894e-05, "loss": 0.9399, "step": 2690 }, { "epoch": 3.501626545217957, "grad_norm": 1.2736738920211792, "learning_rate": 1.2483617300131061e-05, "loss": 0.463, "step": 2691 }, { "epoch": 3.502927781392323, "grad_norm": 0.8204697370529175, "learning_rate": 1.2450851900393186e-05, "loss": 1.131, "step": 2692 }, { "epoch": 3.504229017566688, "grad_norm": 1.145477056503296, "learning_rate": 1.2418086500655309e-05, "loss": 0.4059, "step": 2693 }, { "epoch": 3.505530253741054, "grad_norm": 1.1626449823379517, "learning_rate": 1.2385321100917432e-05, "loss": 0.7623, "step": 2694 }, { "epoch": 3.5068314899154194, "grad_norm": 1.3860214948654175, "learning_rate": 1.2352555701179555e-05, "loss": 0.6879, "step": 2695 }, { "epoch": 3.5081327260897854, "grad_norm": 1.021337866783142, "learning_rate": 1.2319790301441678e-05, "loss": 1.0076, "step": 2696 }, { "epoch": 3.509433962264151, "grad_norm": 1.3309950828552246, "learning_rate": 1.2287024901703801e-05, "loss": 0.7295, "step": 2697 }, { "epoch": 3.5107351984385167, "grad_norm": 1.5034615993499756, "learning_rate": 1.2254259501965924e-05, "loss": 0.2659, "step": 2698 }, { "epoch": 3.5120364346128823, "grad_norm": 1.5256434679031372, "learning_rate": 1.2221494102228049e-05, "loss": 0.6192, "step": 2699 }, { "epoch": 3.513337670787248, "grad_norm": 1.8670177459716797, "learning_rate": 1.218872870249017e-05, "loss": 1.3199, "step": 2700 }, { "epoch": 3.5146389069616135, "grad_norm": 1.248712420463562, "learning_rate": 1.2155963302752295e-05, "loss": 0.5926, "step": 2701 }, { "epoch": 3.515940143135979, "grad_norm": 1.4853813648223877, "learning_rate": 1.2123197903014418e-05, "loss": 0.7716, "step": 2702 }, { "epoch": 3.5172413793103448, "grad_norm": 0.7749918103218079, "learning_rate": 1.209043250327654e-05, "loss": 0.3941, "step": 2703 }, { "epoch": 3.5185426154847104, "grad_norm": 1.0598622560501099, "learning_rate": 1.2057667103538664e-05, "loss": 0.9365, "step": 2704 }, { "epoch": 3.519843851659076, "grad_norm": 1.5593008995056152, "learning_rate": 1.2024901703800787e-05, "loss": 0.4021, "step": 2705 }, { "epoch": 3.5211450878334416, "grad_norm": 1.212952733039856, "learning_rate": 1.199213630406291e-05, "loss": 0.565, "step": 2706 }, { "epoch": 3.5224463240078077, "grad_norm": 1.1329755783081055, "learning_rate": 1.1959370904325033e-05, "loss": 0.4937, "step": 2707 }, { "epoch": 3.523747560182173, "grad_norm": 1.3280813694000244, "learning_rate": 1.1926605504587156e-05, "loss": 0.446, "step": 2708 }, { "epoch": 3.525048796356539, "grad_norm": 1.3388231992721558, "learning_rate": 1.189384010484928e-05, "loss": 1.427, "step": 2709 }, { "epoch": 3.5263500325309045, "grad_norm": 1.4600673913955688, "learning_rate": 1.1861074705111402e-05, "loss": 0.4805, "step": 2710 }, { "epoch": 3.52765126870527, "grad_norm": 1.3730820417404175, "learning_rate": 1.1828309305373527e-05, "loss": 0.3557, "step": 2711 }, { "epoch": 3.5289525048796357, "grad_norm": 1.1716935634613037, "learning_rate": 1.1795543905635649e-05, "loss": 0.4827, "step": 2712 }, { "epoch": 3.5302537410540014, "grad_norm": 1.1674566268920898, "learning_rate": 1.1762778505897773e-05, "loss": 0.4195, "step": 2713 }, { "epoch": 3.531554977228367, "grad_norm": 0.9875827431678772, "learning_rate": 1.1730013106159896e-05, "loss": 0.1562, "step": 2714 }, { "epoch": 3.5328562134027326, "grad_norm": 1.0601670742034912, "learning_rate": 1.169724770642202e-05, "loss": 0.8407, "step": 2715 }, { "epoch": 3.534157449577098, "grad_norm": 1.0620843172073364, "learning_rate": 1.1664482306684142e-05, "loss": 0.744, "step": 2716 }, { "epoch": 3.535458685751464, "grad_norm": 1.3268284797668457, "learning_rate": 1.1631716906946265e-05, "loss": 0.627, "step": 2717 }, { "epoch": 3.5367599219258294, "grad_norm": 0.8451982736587524, "learning_rate": 1.1598951507208389e-05, "loss": 0.0807, "step": 2718 }, { "epoch": 3.538061158100195, "grad_norm": 1.207477331161499, "learning_rate": 1.1566186107470512e-05, "loss": 0.4644, "step": 2719 }, { "epoch": 3.5393623942745607, "grad_norm": 1.2847217321395874, "learning_rate": 1.1533420707732635e-05, "loss": 0.2735, "step": 2720 }, { "epoch": 3.5406636304489263, "grad_norm": 1.0521581172943115, "learning_rate": 1.1500655307994758e-05, "loss": 0.1685, "step": 2721 }, { "epoch": 3.5419648666232924, "grad_norm": 1.223474383354187, "learning_rate": 1.1467889908256882e-05, "loss": 0.4824, "step": 2722 }, { "epoch": 3.5432661027976575, "grad_norm": 0.8339914083480835, "learning_rate": 1.1435124508519004e-05, "loss": 0.8004, "step": 2723 }, { "epoch": 3.5445673389720236, "grad_norm": 1.3238886594772339, "learning_rate": 1.1402359108781127e-05, "loss": 0.9312, "step": 2724 }, { "epoch": 3.545868575146389, "grad_norm": 1.1145790815353394, "learning_rate": 1.1369593709043252e-05, "loss": 0.4901, "step": 2725 }, { "epoch": 3.547169811320755, "grad_norm": 1.3922890424728394, "learning_rate": 1.1336828309305373e-05, "loss": 0.4752, "step": 2726 }, { "epoch": 3.5484710474951204, "grad_norm": 1.215207815170288, "learning_rate": 1.1304062909567498e-05, "loss": 0.7249, "step": 2727 }, { "epoch": 3.549772283669486, "grad_norm": 0.8828149437904358, "learning_rate": 1.127129750982962e-05, "loss": 1.7801, "step": 2728 }, { "epoch": 3.5510735198438517, "grad_norm": 0.993402361869812, "learning_rate": 1.1238532110091744e-05, "loss": 0.813, "step": 2729 }, { "epoch": 3.5523747560182173, "grad_norm": 1.1275135278701782, "learning_rate": 1.1205766710353867e-05, "loss": 0.7487, "step": 2730 }, { "epoch": 3.553675992192583, "grad_norm": 1.3542726039886475, "learning_rate": 1.117300131061599e-05, "loss": 1.0879, "step": 2731 }, { "epoch": 3.5549772283669485, "grad_norm": 1.0169683694839478, "learning_rate": 1.1140235910878113e-05, "loss": 0.5915, "step": 2732 }, { "epoch": 3.556278464541314, "grad_norm": 1.068089485168457, "learning_rate": 1.1107470511140236e-05, "loss": 0.7802, "step": 2733 }, { "epoch": 3.5575797007156797, "grad_norm": 0.9731435179710388, "learning_rate": 1.107470511140236e-05, "loss": 1.3396, "step": 2734 }, { "epoch": 3.558880936890046, "grad_norm": 1.6179232597351074, "learning_rate": 1.1041939711664482e-05, "loss": 0.2437, "step": 2735 }, { "epoch": 3.560182173064411, "grad_norm": 1.1745191812515259, "learning_rate": 1.1009174311926607e-05, "loss": 0.8275, "step": 2736 }, { "epoch": 3.561483409238777, "grad_norm": 1.4166122674942017, "learning_rate": 1.097640891218873e-05, "loss": 0.4996, "step": 2737 }, { "epoch": 3.562784645413142, "grad_norm": 1.0852811336517334, "learning_rate": 1.0943643512450851e-05, "loss": 0.2237, "step": 2738 }, { "epoch": 3.5640858815875083, "grad_norm": 1.325425624847412, "learning_rate": 1.0910878112712976e-05, "loss": 0.4066, "step": 2739 }, { "epoch": 3.565387117761874, "grad_norm": 1.5981690883636475, "learning_rate": 1.0878112712975099e-05, "loss": 0.3608, "step": 2740 }, { "epoch": 3.5666883539362395, "grad_norm": 1.3713462352752686, "learning_rate": 1.0845347313237222e-05, "loss": 0.3951, "step": 2741 }, { "epoch": 3.567989590110605, "grad_norm": 1.026716947555542, "learning_rate": 1.0812581913499345e-05, "loss": 1.0448, "step": 2742 }, { "epoch": 3.5692908262849707, "grad_norm": 1.4420583248138428, "learning_rate": 1.077981651376147e-05, "loss": 0.9747, "step": 2743 }, { "epoch": 3.5705920624593364, "grad_norm": 1.3501439094543457, "learning_rate": 1.0747051114023591e-05, "loss": 0.2299, "step": 2744 }, { "epoch": 3.571893298633702, "grad_norm": 1.3581769466400146, "learning_rate": 1.0714285714285714e-05, "loss": 0.8554, "step": 2745 }, { "epoch": 3.5731945348080676, "grad_norm": 1.4549369812011719, "learning_rate": 1.0681520314547839e-05, "loss": 0.7786, "step": 2746 }, { "epoch": 3.574495770982433, "grad_norm": 1.1988143920898438, "learning_rate": 1.064875491480996e-05, "loss": 0.8193, "step": 2747 }, { "epoch": 3.575797007156799, "grad_norm": 1.1031153202056885, "learning_rate": 1.0615989515072085e-05, "loss": 0.4183, "step": 2748 }, { "epoch": 3.5770982433311644, "grad_norm": 1.4587297439575195, "learning_rate": 1.0583224115334208e-05, "loss": 0.5433, "step": 2749 }, { "epoch": 3.5783994795055305, "grad_norm": 1.3231396675109863, "learning_rate": 1.0550458715596331e-05, "loss": 0.2029, "step": 2750 }, { "epoch": 3.5797007156798957, "grad_norm": 1.1057839393615723, "learning_rate": 1.0517693315858454e-05, "loss": 0.992, "step": 2751 }, { "epoch": 3.5810019518542617, "grad_norm": 1.6316359043121338, "learning_rate": 1.0484927916120577e-05, "loss": 0.6218, "step": 2752 }, { "epoch": 3.5823031880286273, "grad_norm": 1.4686520099639893, "learning_rate": 1.04521625163827e-05, "loss": 0.3403, "step": 2753 }, { "epoch": 3.583604424202993, "grad_norm": 0.9131117463111877, "learning_rate": 1.0419397116644823e-05, "loss": 1.4131, "step": 2754 }, { "epoch": 3.5849056603773586, "grad_norm": 1.439653992652893, "learning_rate": 1.0386631716906948e-05, "loss": 0.363, "step": 2755 }, { "epoch": 3.586206896551724, "grad_norm": 1.5356016159057617, "learning_rate": 1.035386631716907e-05, "loss": 0.4408, "step": 2756 }, { "epoch": 3.58750813272609, "grad_norm": 1.3632124662399292, "learning_rate": 1.0321100917431194e-05, "loss": 0.7446, "step": 2757 }, { "epoch": 3.5888093689004554, "grad_norm": 1.2761905193328857, "learning_rate": 1.0288335517693317e-05, "loss": 0.4815, "step": 2758 }, { "epoch": 3.590110605074821, "grad_norm": 1.0872904062271118, "learning_rate": 1.0255570117955438e-05, "loss": 0.5773, "step": 2759 }, { "epoch": 3.5914118412491867, "grad_norm": 1.2615957260131836, "learning_rate": 1.0222804718217563e-05, "loss": 0.3192, "step": 2760 }, { "epoch": 3.5927130774235523, "grad_norm": 1.3516645431518555, "learning_rate": 1.0190039318479686e-05, "loss": 0.5203, "step": 2761 }, { "epoch": 3.594014313597918, "grad_norm": 1.4621561765670776, "learning_rate": 1.015727391874181e-05, "loss": 0.6675, "step": 2762 }, { "epoch": 3.5953155497722835, "grad_norm": 1.4858742952346802, "learning_rate": 1.0124508519003932e-05, "loss": 0.632, "step": 2763 }, { "epoch": 3.596616785946649, "grad_norm": 1.3683016300201416, "learning_rate": 1.0091743119266055e-05, "loss": 0.3701, "step": 2764 }, { "epoch": 3.597918022121015, "grad_norm": 1.187746286392212, "learning_rate": 1.0058977719528178e-05, "loss": 0.3668, "step": 2765 }, { "epoch": 3.5992192582953804, "grad_norm": 1.1019160747528076, "learning_rate": 1.0026212319790301e-05, "loss": 0.3361, "step": 2766 }, { "epoch": 3.6005204944697464, "grad_norm": 1.0344854593276978, "learning_rate": 9.993446920052426e-06, "loss": 0.5139, "step": 2767 }, { "epoch": 3.601821730644112, "grad_norm": 1.3702645301818848, "learning_rate": 9.960681520314547e-06, "loss": 0.7572, "step": 2768 }, { "epoch": 3.6031229668184777, "grad_norm": 0.9481873512268066, "learning_rate": 9.927916120576672e-06, "loss": 0.4652, "step": 2769 }, { "epoch": 3.6044242029928433, "grad_norm": 1.2302868366241455, "learning_rate": 9.895150720838795e-06, "loss": 1.1109, "step": 2770 }, { "epoch": 3.605725439167209, "grad_norm": 1.2274341583251953, "learning_rate": 9.862385321100918e-06, "loss": 0.7783, "step": 2771 }, { "epoch": 3.6070266753415745, "grad_norm": 0.8875812292098999, "learning_rate": 9.829619921363041e-06, "loss": 1.0303, "step": 2772 }, { "epoch": 3.60832791151594, "grad_norm": 1.0916876792907715, "learning_rate": 9.796854521625164e-06, "loss": 0.5557, "step": 2773 }, { "epoch": 3.6096291476903057, "grad_norm": 0.9609829187393188, "learning_rate": 9.764089121887287e-06, "loss": 0.9084, "step": 2774 }, { "epoch": 3.6109303838646714, "grad_norm": 1.2083057165145874, "learning_rate": 9.73132372214941e-06, "loss": 0.8956, "step": 2775 }, { "epoch": 3.612231620039037, "grad_norm": 1.077060341835022, "learning_rate": 9.698558322411533e-06, "loss": 1.0087, "step": 2776 }, { "epoch": 3.6135328562134026, "grad_norm": 1.2848398685455322, "learning_rate": 9.665792922673657e-06, "loss": 0.2678, "step": 2777 }, { "epoch": 3.6148340923877686, "grad_norm": 1.2483744621276855, "learning_rate": 9.633027522935781e-06, "loss": 0.372, "step": 2778 }, { "epoch": 3.616135328562134, "grad_norm": 1.0897185802459717, "learning_rate": 9.600262123197903e-06, "loss": 0.3073, "step": 2779 }, { "epoch": 3.6174365647365, "grad_norm": 1.085809588432312, "learning_rate": 9.567496723460026e-06, "loss": 0.2646, "step": 2780 }, { "epoch": 3.618737800910865, "grad_norm": 1.4805822372436523, "learning_rate": 9.53473132372215e-06, "loss": 0.5239, "step": 2781 }, { "epoch": 3.620039037085231, "grad_norm": 0.9890291094779968, "learning_rate": 9.501965923984273e-06, "loss": 0.6142, "step": 2782 }, { "epoch": 3.6213402732595967, "grad_norm": 1.4204612970352173, "learning_rate": 9.469200524246396e-06, "loss": 0.3553, "step": 2783 }, { "epoch": 3.6226415094339623, "grad_norm": 1.3294740915298462, "learning_rate": 9.43643512450852e-06, "loss": 0.3313, "step": 2784 }, { "epoch": 3.623942745608328, "grad_norm": 1.2093985080718994, "learning_rate": 9.403669724770643e-06, "loss": 0.9752, "step": 2785 }, { "epoch": 3.6252439817826936, "grad_norm": 1.2423747777938843, "learning_rate": 9.370904325032766e-06, "loss": 0.3858, "step": 2786 }, { "epoch": 3.626545217957059, "grad_norm": 1.1941457986831665, "learning_rate": 9.338138925294889e-06, "loss": 0.6017, "step": 2787 }, { "epoch": 3.627846454131425, "grad_norm": 1.3232699632644653, "learning_rate": 9.305373525557012e-06, "loss": 0.3687, "step": 2788 }, { "epoch": 3.6291476903057904, "grad_norm": 1.2541453838348389, "learning_rate": 9.272608125819135e-06, "loss": 0.4239, "step": 2789 }, { "epoch": 3.630448926480156, "grad_norm": 1.1581919193267822, "learning_rate": 9.23984272608126e-06, "loss": 1.0722, "step": 2790 }, { "epoch": 3.6317501626545217, "grad_norm": 1.2644391059875488, "learning_rate": 9.20707732634338e-06, "loss": 0.2958, "step": 2791 }, { "epoch": 3.6330513988288873, "grad_norm": 1.4206126928329468, "learning_rate": 9.174311926605506e-06, "loss": 0.6824, "step": 2792 }, { "epoch": 3.6343526350032533, "grad_norm": 1.2815566062927246, "learning_rate": 9.141546526867629e-06, "loss": 0.7139, "step": 2793 }, { "epoch": 3.6356538711776185, "grad_norm": 1.3940539360046387, "learning_rate": 9.10878112712975e-06, "loss": 0.5165, "step": 2794 }, { "epoch": 3.6369551073519846, "grad_norm": 1.3667677640914917, "learning_rate": 9.076015727391875e-06, "loss": 0.5325, "step": 2795 }, { "epoch": 3.63825634352635, "grad_norm": 0.758829653263092, "learning_rate": 9.043250327653998e-06, "loss": 1.0032, "step": 2796 }, { "epoch": 3.639557579700716, "grad_norm": 0.9183512926101685, "learning_rate": 9.01048492791612e-06, "loss": 0.6469, "step": 2797 }, { "epoch": 3.6408588158750814, "grad_norm": 1.220260739326477, "learning_rate": 8.977719528178244e-06, "loss": 0.3431, "step": 2798 }, { "epoch": 3.642160052049447, "grad_norm": 1.1976498365402222, "learning_rate": 8.944954128440369e-06, "loss": 0.2913, "step": 2799 }, { "epoch": 3.6434612882238127, "grad_norm": 0.8291826844215393, "learning_rate": 8.91218872870249e-06, "loss": 0.6133, "step": 2800 }, { "epoch": 3.6447625243981783, "grad_norm": 1.1449323892593384, "learning_rate": 8.879423328964615e-06, "loss": 0.7864, "step": 2801 }, { "epoch": 3.646063760572544, "grad_norm": 0.9846778512001038, "learning_rate": 8.846657929226738e-06, "loss": 0.3832, "step": 2802 }, { "epoch": 3.6473649967469095, "grad_norm": 1.3869627714157104, "learning_rate": 8.813892529488859e-06, "loss": 1.0456, "step": 2803 }, { "epoch": 3.648666232921275, "grad_norm": 1.2912912368774414, "learning_rate": 8.781127129750984e-06, "loss": 0.8242, "step": 2804 }, { "epoch": 3.6499674690956407, "grad_norm": 1.334542155265808, "learning_rate": 8.748361730013107e-06, "loss": 0.5624, "step": 2805 }, { "epoch": 3.6512687052700064, "grad_norm": 1.279402732849121, "learning_rate": 8.71559633027523e-06, "loss": 0.8531, "step": 2806 }, { "epoch": 3.652569941444372, "grad_norm": 1.0433616638183594, "learning_rate": 8.682830930537353e-06, "loss": 0.8139, "step": 2807 }, { "epoch": 3.653871177618738, "grad_norm": 1.0716416835784912, "learning_rate": 8.650065530799478e-06, "loss": 0.539, "step": 2808 }, { "epoch": 3.655172413793103, "grad_norm": 0.9396475553512573, "learning_rate": 8.617300131061599e-06, "loss": 0.9783, "step": 2809 }, { "epoch": 3.6564736499674693, "grad_norm": 1.0220582485198975, "learning_rate": 8.584534731323722e-06, "loss": 0.81, "step": 2810 }, { "epoch": 3.657774886141835, "grad_norm": 1.3606343269348145, "learning_rate": 8.551769331585847e-06, "loss": 0.9687, "step": 2811 }, { "epoch": 3.6590761223162005, "grad_norm": 0.8804569840431213, "learning_rate": 8.519003931847968e-06, "loss": 0.5825, "step": 2812 }, { "epoch": 3.660377358490566, "grad_norm": 1.0917452573776245, "learning_rate": 8.486238532110093e-06, "loss": 0.7133, "step": 2813 }, { "epoch": 3.6616785946649317, "grad_norm": 1.4057462215423584, "learning_rate": 8.453473132372216e-06, "loss": 0.8726, "step": 2814 }, { "epoch": 3.6629798308392973, "grad_norm": 1.5779625177383423, "learning_rate": 8.420707732634339e-06, "loss": 0.3834, "step": 2815 }, { "epoch": 3.664281067013663, "grad_norm": 1.6935299634933472, "learning_rate": 8.387942332896462e-06, "loss": 0.3049, "step": 2816 }, { "epoch": 3.6655823031880286, "grad_norm": 1.7838581800460815, "learning_rate": 8.355176933158585e-06, "loss": 0.4017, "step": 2817 }, { "epoch": 3.666883539362394, "grad_norm": 0.8527935147285461, "learning_rate": 8.322411533420708e-06, "loss": 1.0602, "step": 2818 }, { "epoch": 3.66818477553676, "grad_norm": 1.2594292163848877, "learning_rate": 8.289646133682831e-06, "loss": 0.6772, "step": 2819 }, { "epoch": 3.6694860117111254, "grad_norm": 1.0284929275512695, "learning_rate": 8.256880733944954e-06, "loss": 0.9103, "step": 2820 }, { "epoch": 3.6707872478854915, "grad_norm": 1.0171302556991577, "learning_rate": 8.224115334207077e-06, "loss": 1.0869, "step": 2821 }, { "epoch": 3.6720884840598567, "grad_norm": 1.074184536933899, "learning_rate": 8.191349934469202e-06, "loss": 0.5015, "step": 2822 }, { "epoch": 3.6733897202342227, "grad_norm": 1.2699873447418213, "learning_rate": 8.158584534731325e-06, "loss": 0.2683, "step": 2823 }, { "epoch": 3.674690956408588, "grad_norm": 1.2072802782058716, "learning_rate": 8.125819134993446e-06, "loss": 0.7802, "step": 2824 }, { "epoch": 3.675992192582954, "grad_norm": 1.112255573272705, "learning_rate": 8.093053735255571e-06, "loss": 1.0251, "step": 2825 }, { "epoch": 3.6772934287573196, "grad_norm": 1.885940432548523, "learning_rate": 8.060288335517694e-06, "loss": 0.7253, "step": 2826 }, { "epoch": 3.678594664931685, "grad_norm": 1.391442060470581, "learning_rate": 8.027522935779817e-06, "loss": 0.428, "step": 2827 }, { "epoch": 3.679895901106051, "grad_norm": 1.1297540664672852, "learning_rate": 7.99475753604194e-06, "loss": 1.104, "step": 2828 }, { "epoch": 3.6811971372804164, "grad_norm": 1.2743631601333618, "learning_rate": 7.961992136304063e-06, "loss": 0.5221, "step": 2829 }, { "epoch": 3.682498373454782, "grad_norm": 1.3905729055404663, "learning_rate": 7.929226736566186e-06, "loss": 0.5952, "step": 2830 }, { "epoch": 3.6837996096291477, "grad_norm": 0.9651404619216919, "learning_rate": 7.89646133682831e-06, "loss": 0.2709, "step": 2831 }, { "epoch": 3.6851008458035133, "grad_norm": 1.2327051162719727, "learning_rate": 7.863695937090432e-06, "loss": 0.632, "step": 2832 }, { "epoch": 3.686402081977879, "grad_norm": 1.5038141012191772, "learning_rate": 7.830930537352555e-06, "loss": 0.4902, "step": 2833 }, { "epoch": 3.6877033181522445, "grad_norm": 1.2557305097579956, "learning_rate": 7.79816513761468e-06, "loss": 0.2719, "step": 2834 }, { "epoch": 3.68900455432661, "grad_norm": 0.9384816884994507, "learning_rate": 7.765399737876802e-06, "loss": 0.8376, "step": 2835 }, { "epoch": 3.690305790500976, "grad_norm": 1.2277883291244507, "learning_rate": 7.732634338138926e-06, "loss": 1.0737, "step": 2836 }, { "epoch": 3.6916070266753414, "grad_norm": 1.5773671865463257, "learning_rate": 7.69986893840105e-06, "loss": 0.8036, "step": 2837 }, { "epoch": 3.6929082628497074, "grad_norm": 0.9968787431716919, "learning_rate": 7.667103538663172e-06, "loss": 0.7393, "step": 2838 }, { "epoch": 3.694209499024073, "grad_norm": 1.0832473039627075, "learning_rate": 7.634338138925295e-06, "loss": 1.5288, "step": 2839 }, { "epoch": 3.6955107351984386, "grad_norm": 1.5942686796188354, "learning_rate": 7.601572739187418e-06, "loss": 0.4141, "step": 2840 }, { "epoch": 3.6968119713728043, "grad_norm": 1.3327304124832153, "learning_rate": 7.568807339449542e-06, "loss": 0.3756, "step": 2841 }, { "epoch": 3.69811320754717, "grad_norm": 1.300759196281433, "learning_rate": 7.5360419397116645e-06, "loss": 0.8255, "step": 2842 }, { "epoch": 3.6994144437215355, "grad_norm": 0.962526798248291, "learning_rate": 7.503276539973788e-06, "loss": 0.8096, "step": 2843 }, { "epoch": 3.700715679895901, "grad_norm": 1.5482956171035767, "learning_rate": 7.4705111402359114e-06, "loss": 0.808, "step": 2844 }, { "epoch": 3.7020169160702667, "grad_norm": 1.183812141418457, "learning_rate": 7.437745740498034e-06, "loss": 0.5625, "step": 2845 }, { "epoch": 3.7033181522446323, "grad_norm": 1.1432191133499146, "learning_rate": 7.4049803407601575e-06, "loss": 0.6762, "step": 2846 }, { "epoch": 3.704619388418998, "grad_norm": 1.0869256258010864, "learning_rate": 7.3722149410222806e-06, "loss": 0.7722, "step": 2847 }, { "epoch": 3.7059206245933636, "grad_norm": 1.2133201360702515, "learning_rate": 7.3394495412844045e-06, "loss": 0.6498, "step": 2848 }, { "epoch": 3.707221860767729, "grad_norm": 0.9540015459060669, "learning_rate": 7.306684141546527e-06, "loss": 0.514, "step": 2849 }, { "epoch": 3.708523096942095, "grad_norm": 1.1689519882202148, "learning_rate": 7.273918741808651e-06, "loss": 0.8032, "step": 2850 }, { "epoch": 3.709824333116461, "grad_norm": 1.3757344484329224, "learning_rate": 7.241153342070774e-06, "loss": 0.6011, "step": 2851 }, { "epoch": 3.711125569290826, "grad_norm": 1.1116350889205933, "learning_rate": 7.208387942332897e-06, "loss": 0.5744, "step": 2852 }, { "epoch": 3.712426805465192, "grad_norm": 1.4113695621490479, "learning_rate": 7.1756225425950205e-06, "loss": 0.9997, "step": 2853 }, { "epoch": 3.7137280416395577, "grad_norm": 1.556117057800293, "learning_rate": 7.142857142857143e-06, "loss": 0.483, "step": 2854 }, { "epoch": 3.7150292778139233, "grad_norm": 1.0709412097930908, "learning_rate": 7.110091743119267e-06, "loss": 0.5171, "step": 2855 }, { "epoch": 3.716330513988289, "grad_norm": 1.3775649070739746, "learning_rate": 7.07732634338139e-06, "loss": 0.4673, "step": 2856 }, { "epoch": 3.7176317501626546, "grad_norm": 1.5115619897842407, "learning_rate": 7.0445609436435135e-06, "loss": 0.8605, "step": 2857 }, { "epoch": 3.71893298633702, "grad_norm": 1.3077727556228638, "learning_rate": 7.011795543905636e-06, "loss": 0.5291, "step": 2858 }, { "epoch": 3.720234222511386, "grad_norm": 1.2592308521270752, "learning_rate": 6.979030144167759e-06, "loss": 0.3833, "step": 2859 }, { "epoch": 3.7215354586857514, "grad_norm": 1.2495887279510498, "learning_rate": 6.946264744429883e-06, "loss": 0.8078, "step": 2860 }, { "epoch": 3.722836694860117, "grad_norm": 1.1660330295562744, "learning_rate": 6.913499344692005e-06, "loss": 0.552, "step": 2861 }, { "epoch": 3.7241379310344827, "grad_norm": 1.0339685678482056, "learning_rate": 6.880733944954129e-06, "loss": 0.1485, "step": 2862 }, { "epoch": 3.7254391672088483, "grad_norm": 1.3942471742630005, "learning_rate": 6.847968545216252e-06, "loss": 0.6679, "step": 2863 }, { "epoch": 3.7267404033832143, "grad_norm": 1.1803205013275146, "learning_rate": 6.815203145478376e-06, "loss": 0.7743, "step": 2864 }, { "epoch": 3.7280416395575795, "grad_norm": 0.9076970219612122, "learning_rate": 6.782437745740499e-06, "loss": 0.9282, "step": 2865 }, { "epoch": 3.7293428757319456, "grad_norm": 1.4049845933914185, "learning_rate": 6.749672346002621e-06, "loss": 0.3824, "step": 2866 }, { "epoch": 3.7306441119063107, "grad_norm": 1.2189512252807617, "learning_rate": 6.716906946264745e-06, "loss": 0.4967, "step": 2867 }, { "epoch": 3.731945348080677, "grad_norm": 1.1928021907806396, "learning_rate": 6.684141546526868e-06, "loss": 0.6521, "step": 2868 }, { "epoch": 3.7332465842550424, "grad_norm": 1.107861876487732, "learning_rate": 6.651376146788992e-06, "loss": 0.4149, "step": 2869 }, { "epoch": 3.734547820429408, "grad_norm": 1.1047502756118774, "learning_rate": 6.618610747051114e-06, "loss": 0.516, "step": 2870 }, { "epoch": 3.7358490566037736, "grad_norm": 0.7596063017845154, "learning_rate": 6.585845347313238e-06, "loss": 0.7289, "step": 2871 }, { "epoch": 3.7371502927781393, "grad_norm": 1.5554673671722412, "learning_rate": 6.553079947575361e-06, "loss": 0.4425, "step": 2872 }, { "epoch": 3.738451528952505, "grad_norm": 1.2365245819091797, "learning_rate": 6.520314547837483e-06, "loss": 0.328, "step": 2873 }, { "epoch": 3.7397527651268705, "grad_norm": 1.2272427082061768, "learning_rate": 6.487549148099607e-06, "loss": 0.4102, "step": 2874 }, { "epoch": 3.741054001301236, "grad_norm": 1.1512951850891113, "learning_rate": 6.45478374836173e-06, "loss": 0.7288, "step": 2875 }, { "epoch": 3.7423552374756017, "grad_norm": 1.3824795484542847, "learning_rate": 6.422018348623854e-06, "loss": 1.182, "step": 2876 }, { "epoch": 3.7436564736499673, "grad_norm": 1.6226873397827148, "learning_rate": 6.389252948885976e-06, "loss": 0.3519, "step": 2877 }, { "epoch": 3.744957709824333, "grad_norm": 1.131341814994812, "learning_rate": 6.356487549148101e-06, "loss": 0.657, "step": 2878 }, { "epoch": 3.746258945998699, "grad_norm": 1.4460186958312988, "learning_rate": 6.323722149410223e-06, "loss": 0.9889, "step": 2879 }, { "epoch": 3.747560182173064, "grad_norm": 1.520754337310791, "learning_rate": 6.290956749672346e-06, "loss": 0.317, "step": 2880 }, { "epoch": 3.7488614183474303, "grad_norm": 1.4077394008636475, "learning_rate": 6.25819134993447e-06, "loss": 0.4812, "step": 2881 }, { "epoch": 3.750162654521796, "grad_norm": 1.1954708099365234, "learning_rate": 6.225425950196593e-06, "loss": 0.7953, "step": 2882 }, { "epoch": 3.7514638906961615, "grad_norm": 1.286417841911316, "learning_rate": 6.192660550458716e-06, "loss": 0.8902, "step": 2883 }, { "epoch": 3.752765126870527, "grad_norm": 1.9368194341659546, "learning_rate": 6.159895150720839e-06, "loss": 0.7977, "step": 2884 }, { "epoch": 3.7540663630448927, "grad_norm": 1.3488881587982178, "learning_rate": 6.127129750982962e-06, "loss": 0.6271, "step": 2885 }, { "epoch": 3.7553675992192583, "grad_norm": 1.0769685506820679, "learning_rate": 6.094364351245085e-06, "loss": 0.6709, "step": 2886 }, { "epoch": 3.756668835393624, "grad_norm": 1.3828927278518677, "learning_rate": 6.061598951507209e-06, "loss": 0.8335, "step": 2887 }, { "epoch": 3.7579700715679896, "grad_norm": 0.9474514722824097, "learning_rate": 6.028833551769332e-06, "loss": 0.9507, "step": 2888 }, { "epoch": 3.759271307742355, "grad_norm": 1.1869299411773682, "learning_rate": 5.996068152031455e-06, "loss": 0.4871, "step": 2889 }, { "epoch": 3.760572543916721, "grad_norm": 0.924910306930542, "learning_rate": 5.963302752293578e-06, "loss": 0.588, "step": 2890 }, { "epoch": 3.7618737800910864, "grad_norm": 0.9285858869552612, "learning_rate": 5.930537352555701e-06, "loss": 1.1408, "step": 2891 }, { "epoch": 3.763175016265452, "grad_norm": 1.3063764572143555, "learning_rate": 5.897771952817824e-06, "loss": 0.4535, "step": 2892 }, { "epoch": 3.7644762524398176, "grad_norm": 1.0144656896591187, "learning_rate": 5.865006553079948e-06, "loss": 0.9401, "step": 2893 }, { "epoch": 3.7657774886141837, "grad_norm": 1.3274751901626587, "learning_rate": 5.832241153342071e-06, "loss": 0.7161, "step": 2894 }, { "epoch": 3.767078724788549, "grad_norm": 1.3066692352294922, "learning_rate": 5.799475753604194e-06, "loss": 0.7209, "step": 2895 }, { "epoch": 3.768379960962915, "grad_norm": 1.1454139947891235, "learning_rate": 5.766710353866317e-06, "loss": 0.5174, "step": 2896 }, { "epoch": 3.7696811971372806, "grad_norm": 1.520671010017395, "learning_rate": 5.733944954128441e-06, "loss": 0.8653, "step": 2897 }, { "epoch": 3.770982433311646, "grad_norm": 1.5656856298446655, "learning_rate": 5.701179554390563e-06, "loss": 0.382, "step": 2898 }, { "epoch": 3.772283669486012, "grad_norm": 1.017261266708374, "learning_rate": 5.6684141546526864e-06, "loss": 0.6725, "step": 2899 }, { "epoch": 3.7735849056603774, "grad_norm": 0.8045517802238464, "learning_rate": 5.63564875491481e-06, "loss": 0.9332, "step": 2900 }, { "epoch": 3.774886141834743, "grad_norm": 1.147047996520996, "learning_rate": 5.602883355176933e-06, "loss": 0.4266, "step": 2901 }, { "epoch": 3.7761873780091086, "grad_norm": 1.2275270223617554, "learning_rate": 5.570117955439056e-06, "loss": 1.0922, "step": 2902 }, { "epoch": 3.7774886141834743, "grad_norm": 0.7681794762611389, "learning_rate": 5.53735255570118e-06, "loss": 0.3363, "step": 2903 }, { "epoch": 3.77878985035784, "grad_norm": 0.7455167770385742, "learning_rate": 5.504587155963303e-06, "loss": 1.0908, "step": 2904 }, { "epoch": 3.7800910865322055, "grad_norm": 1.044960618019104, "learning_rate": 5.4718217562254255e-06, "loss": 0.3081, "step": 2905 }, { "epoch": 3.781392322706571, "grad_norm": 1.0693838596343994, "learning_rate": 5.4390563564875494e-06, "loss": 0.6591, "step": 2906 }, { "epoch": 3.7826935588809367, "grad_norm": 1.1125121116638184, "learning_rate": 5.4062909567496725e-06, "loss": 0.8719, "step": 2907 }, { "epoch": 3.7839947950553023, "grad_norm": 0.9367749094963074, "learning_rate": 5.3735255570117955e-06, "loss": 0.3275, "step": 2908 }, { "epoch": 3.7852960312296684, "grad_norm": 1.1559170484542847, "learning_rate": 5.340760157273919e-06, "loss": 0.5108, "step": 2909 }, { "epoch": 3.7865972674040336, "grad_norm": 1.0347143411636353, "learning_rate": 5.3079947575360424e-06, "loss": 1.5218, "step": 2910 }, { "epoch": 3.7878985035783996, "grad_norm": 1.4686614274978638, "learning_rate": 5.2752293577981655e-06, "loss": 0.6388, "step": 2911 }, { "epoch": 3.7891997397527653, "grad_norm": 1.2606450319290161, "learning_rate": 5.2424639580602885e-06, "loss": 1.0854, "step": 2912 }, { "epoch": 3.790500975927131, "grad_norm": 0.789685070514679, "learning_rate": 5.209698558322412e-06, "loss": 0.6026, "step": 2913 }, { "epoch": 3.7918022121014965, "grad_norm": 0.9540252089500427, "learning_rate": 5.176933158584535e-06, "loss": 0.2778, "step": 2914 }, { "epoch": 3.793103448275862, "grad_norm": 1.251987338066101, "learning_rate": 5.1441677588466585e-06, "loss": 1.0102, "step": 2915 }, { "epoch": 3.7944046844502277, "grad_norm": 1.4802610874176025, "learning_rate": 5.1114023591087816e-06, "loss": 0.325, "step": 2916 }, { "epoch": 3.7957059206245933, "grad_norm": 1.344650387763977, "learning_rate": 5.078636959370905e-06, "loss": 0.3686, "step": 2917 }, { "epoch": 3.797007156798959, "grad_norm": 1.2692309617996216, "learning_rate": 5.045871559633028e-06, "loss": 0.2511, "step": 2918 }, { "epoch": 3.7983083929733246, "grad_norm": 1.1660329103469849, "learning_rate": 5.013106159895151e-06, "loss": 0.6327, "step": 2919 }, { "epoch": 3.79960962914769, "grad_norm": 1.4321316480636597, "learning_rate": 4.980340760157274e-06, "loss": 0.612, "step": 2920 }, { "epoch": 3.800910865322056, "grad_norm": 1.1449389457702637, "learning_rate": 4.947575360419398e-06, "loss": 0.3581, "step": 2921 }, { "epoch": 3.802212101496422, "grad_norm": 1.3535315990447998, "learning_rate": 4.914809960681521e-06, "loss": 0.4995, "step": 2922 }, { "epoch": 3.803513337670787, "grad_norm": 1.1499452590942383, "learning_rate": 4.882044560943644e-06, "loss": 0.6917, "step": 2923 }, { "epoch": 3.804814573845153, "grad_norm": 1.4985668659210205, "learning_rate": 4.849279161205767e-06, "loss": 0.4653, "step": 2924 }, { "epoch": 3.8061158100195187, "grad_norm": 0.9997090697288513, "learning_rate": 4.816513761467891e-06, "loss": 0.8017, "step": 2925 }, { "epoch": 3.8074170461938843, "grad_norm": 1.2876784801483154, "learning_rate": 4.783748361730013e-06, "loss": 0.2346, "step": 2926 }, { "epoch": 3.80871828236825, "grad_norm": 1.3681167364120483, "learning_rate": 4.750982961992137e-06, "loss": 0.7614, "step": 2927 }, { "epoch": 3.8100195185426156, "grad_norm": 1.437312364578247, "learning_rate": 4.71821756225426e-06, "loss": 1.1087, "step": 2928 }, { "epoch": 3.811320754716981, "grad_norm": 1.000667691230774, "learning_rate": 4.685452162516383e-06, "loss": 0.6006, "step": 2929 }, { "epoch": 3.812621990891347, "grad_norm": 0.9426958560943604, "learning_rate": 4.652686762778506e-06, "loss": 0.4839, "step": 2930 }, { "epoch": 3.8139232270657124, "grad_norm": 1.2187105417251587, "learning_rate": 4.61992136304063e-06, "loss": 0.8629, "step": 2931 }, { "epoch": 3.815224463240078, "grad_norm": 1.0024930238723755, "learning_rate": 4.587155963302753e-06, "loss": 1.0116, "step": 2932 }, { "epoch": 3.8165256994144436, "grad_norm": 1.3322170972824097, "learning_rate": 4.554390563564875e-06, "loss": 0.37, "step": 2933 }, { "epoch": 3.8178269355888093, "grad_norm": 1.1999846696853638, "learning_rate": 4.521625163826999e-06, "loss": 0.5487, "step": 2934 }, { "epoch": 3.819128171763175, "grad_norm": 1.0977870225906372, "learning_rate": 4.488859764089122e-06, "loss": 0.4719, "step": 2935 }, { "epoch": 3.8204294079375405, "grad_norm": 1.1847038269042969, "learning_rate": 4.456094364351245e-06, "loss": 1.0249, "step": 2936 }, { "epoch": 3.8217306441119065, "grad_norm": 1.0872751474380493, "learning_rate": 4.423328964613369e-06, "loss": 0.9547, "step": 2937 }, { "epoch": 3.8230318802862717, "grad_norm": 1.391088604927063, "learning_rate": 4.390563564875492e-06, "loss": 0.4683, "step": 2938 }, { "epoch": 3.824333116460638, "grad_norm": 1.3338350057601929, "learning_rate": 4.357798165137615e-06, "loss": 0.9974, "step": 2939 }, { "epoch": 3.8256343526350034, "grad_norm": 1.3243476152420044, "learning_rate": 4.325032765399739e-06, "loss": 0.8473, "step": 2940 }, { "epoch": 3.826935588809369, "grad_norm": 1.1241563558578491, "learning_rate": 4.292267365661861e-06, "loss": 0.2913, "step": 2941 }, { "epoch": 3.8282368249837346, "grad_norm": 1.0621124505996704, "learning_rate": 4.259501965923984e-06, "loss": 1.1436, "step": 2942 }, { "epoch": 3.8295380611581002, "grad_norm": 1.2437814474105835, "learning_rate": 4.226736566186108e-06, "loss": 0.4168, "step": 2943 }, { "epoch": 3.830839297332466, "grad_norm": 1.5372247695922852, "learning_rate": 4.193971166448231e-06, "loss": 0.3165, "step": 2944 }, { "epoch": 3.8321405335068315, "grad_norm": 1.15373694896698, "learning_rate": 4.161205766710354e-06, "loss": 0.7818, "step": 2945 }, { "epoch": 3.833441769681197, "grad_norm": 1.1348756551742554, "learning_rate": 4.128440366972477e-06, "loss": 0.9367, "step": 2946 }, { "epoch": 3.8347430058555627, "grad_norm": 1.2052911520004272, "learning_rate": 4.095674967234601e-06, "loss": 0.411, "step": 2947 }, { "epoch": 3.8360442420299283, "grad_norm": 1.2902486324310303, "learning_rate": 4.062909567496723e-06, "loss": 0.7623, "step": 2948 }, { "epoch": 3.837345478204294, "grad_norm": 1.605956792831421, "learning_rate": 4.030144167758847e-06, "loss": 0.6304, "step": 2949 }, { "epoch": 3.8386467143786596, "grad_norm": 1.2989495992660522, "learning_rate": 3.99737876802097e-06, "loss": 0.3751, "step": 2950 }, { "epoch": 3.839947950553025, "grad_norm": 1.0615744590759277, "learning_rate": 3.964613368283093e-06, "loss": 1.1675, "step": 2951 }, { "epoch": 3.8412491867273912, "grad_norm": 1.3745225667953491, "learning_rate": 3.931847968545216e-06, "loss": 0.9109, "step": 2952 }, { "epoch": 3.8425504229017564, "grad_norm": 1.3383519649505615, "learning_rate": 3.89908256880734e-06, "loss": 0.2246, "step": 2953 }, { "epoch": 3.8438516590761225, "grad_norm": 1.148453712463379, "learning_rate": 3.866317169069463e-06, "loss": 0.5494, "step": 2954 }, { "epoch": 3.845152895250488, "grad_norm": 1.4404183626174927, "learning_rate": 3.833551769331586e-06, "loss": 0.7037, "step": 2955 }, { "epoch": 3.8464541314248537, "grad_norm": 1.299943208694458, "learning_rate": 3.800786369593709e-06, "loss": 0.4765, "step": 2956 }, { "epoch": 3.8477553675992193, "grad_norm": 1.482748031616211, "learning_rate": 3.7680209698558323e-06, "loss": 0.7831, "step": 2957 }, { "epoch": 3.849056603773585, "grad_norm": 1.0106803178787231, "learning_rate": 3.7352555701179557e-06, "loss": 0.3469, "step": 2958 }, { "epoch": 3.8503578399479506, "grad_norm": 0.9227120280265808, "learning_rate": 3.7024901703800788e-06, "loss": 0.7297, "step": 2959 }, { "epoch": 3.851659076122316, "grad_norm": 1.540203332901001, "learning_rate": 3.6697247706422022e-06, "loss": 0.4645, "step": 2960 }, { "epoch": 3.852960312296682, "grad_norm": 1.1309471130371094, "learning_rate": 3.6369593709043257e-06, "loss": 0.7292, "step": 2961 }, { "epoch": 3.8542615484710474, "grad_norm": 0.9003843069076538, "learning_rate": 3.6041939711664483e-06, "loss": 0.5694, "step": 2962 }, { "epoch": 3.855562784645413, "grad_norm": 1.396562933921814, "learning_rate": 3.5714285714285714e-06, "loss": 0.3152, "step": 2963 }, { "epoch": 3.8568640208197786, "grad_norm": 1.3350234031677246, "learning_rate": 3.538663171690695e-06, "loss": 0.7216, "step": 2964 }, { "epoch": 3.8581652569941447, "grad_norm": 1.7421214580535889, "learning_rate": 3.505897771952818e-06, "loss": 0.2889, "step": 2965 }, { "epoch": 3.85946649316851, "grad_norm": 1.5232897996902466, "learning_rate": 3.4731323722149413e-06, "loss": 1.1151, "step": 2966 }, { "epoch": 3.860767729342876, "grad_norm": 1.1938464641571045, "learning_rate": 3.4403669724770644e-06, "loss": 0.5741, "step": 2967 }, { "epoch": 3.862068965517241, "grad_norm": 1.0993099212646484, "learning_rate": 3.407601572739188e-06, "loss": 0.2489, "step": 2968 }, { "epoch": 3.863370201691607, "grad_norm": 1.212775468826294, "learning_rate": 3.3748361730013105e-06, "loss": 0.4051, "step": 2969 }, { "epoch": 3.864671437865973, "grad_norm": 1.029958724975586, "learning_rate": 3.342070773263434e-06, "loss": 0.0938, "step": 2970 }, { "epoch": 3.8659726740403384, "grad_norm": 0.9831691980361938, "learning_rate": 3.309305373525557e-06, "loss": 0.7103, "step": 2971 }, { "epoch": 3.867273910214704, "grad_norm": 0.8898549675941467, "learning_rate": 3.2765399737876804e-06, "loss": 0.5182, "step": 2972 }, { "epoch": 3.8685751463890696, "grad_norm": 1.460985779762268, "learning_rate": 3.2437745740498035e-06, "loss": 0.3284, "step": 2973 }, { "epoch": 3.8698763825634352, "grad_norm": 1.3370801210403442, "learning_rate": 3.211009174311927e-06, "loss": 0.4216, "step": 2974 }, { "epoch": 3.871177618737801, "grad_norm": 1.1864559650421143, "learning_rate": 3.1782437745740504e-06, "loss": 0.4454, "step": 2975 }, { "epoch": 3.8724788549121665, "grad_norm": 1.54583740234375, "learning_rate": 3.145478374836173e-06, "loss": 0.4949, "step": 2976 }, { "epoch": 3.873780091086532, "grad_norm": 1.1163302659988403, "learning_rate": 3.1127129750982965e-06, "loss": 0.6234, "step": 2977 }, { "epoch": 3.8750813272608977, "grad_norm": 0.955952525138855, "learning_rate": 3.0799475753604195e-06, "loss": 1.0813, "step": 2978 }, { "epoch": 3.8763825634352633, "grad_norm": 1.2365391254425049, "learning_rate": 3.0471821756225426e-06, "loss": 0.9685, "step": 2979 }, { "epoch": 3.8776837996096294, "grad_norm": 1.0728838443756104, "learning_rate": 3.014416775884666e-06, "loss": 0.3349, "step": 2980 }, { "epoch": 3.8789850357839946, "grad_norm": 1.2007548809051514, "learning_rate": 2.981651376146789e-06, "loss": 1.1158, "step": 2981 }, { "epoch": 3.8802862719583606, "grad_norm": 1.1061577796936035, "learning_rate": 2.948885976408912e-06, "loss": 0.4368, "step": 2982 }, { "epoch": 3.8815875081327262, "grad_norm": 0.788131833076477, "learning_rate": 2.9161205766710356e-06, "loss": 0.7055, "step": 2983 }, { "epoch": 3.882888744307092, "grad_norm": 0.9695775508880615, "learning_rate": 2.8833551769331587e-06, "loss": 0.7296, "step": 2984 }, { "epoch": 3.8841899804814575, "grad_norm": 1.350740909576416, "learning_rate": 2.8505897771952817e-06, "loss": 0.4433, "step": 2985 }, { "epoch": 3.885491216655823, "grad_norm": 1.0974007844924927, "learning_rate": 2.817824377457405e-06, "loss": 1.0219, "step": 2986 }, { "epoch": 3.8867924528301887, "grad_norm": 1.1883139610290527, "learning_rate": 2.785058977719528e-06, "loss": 0.7584, "step": 2987 }, { "epoch": 3.8880936890045543, "grad_norm": 1.043285846710205, "learning_rate": 2.7522935779816517e-06, "loss": 0.8446, "step": 2988 }, { "epoch": 3.88939492517892, "grad_norm": 1.4048840999603271, "learning_rate": 2.7195281782437747e-06, "loss": 0.5741, "step": 2989 }, { "epoch": 3.8906961613532856, "grad_norm": 0.9970018267631531, "learning_rate": 2.6867627785058978e-06, "loss": 0.2613, "step": 2990 }, { "epoch": 3.891997397527651, "grad_norm": 1.1650549173355103, "learning_rate": 2.6539973787680212e-06, "loss": 0.7268, "step": 2991 }, { "epoch": 3.893298633702017, "grad_norm": 1.103330135345459, "learning_rate": 2.6212319790301443e-06, "loss": 0.508, "step": 2992 }, { "epoch": 3.8945998698763824, "grad_norm": 0.9162395596504211, "learning_rate": 2.5884665792922673e-06, "loss": 0.6445, "step": 2993 }, { "epoch": 3.895901106050748, "grad_norm": 1.3857535123825073, "learning_rate": 2.5557011795543908e-06, "loss": 0.4204, "step": 2994 }, { "epoch": 3.897202342225114, "grad_norm": 0.9722936749458313, "learning_rate": 2.522935779816514e-06, "loss": 1.2443, "step": 2995 }, { "epoch": 3.8985035783994793, "grad_norm": 1.382798194885254, "learning_rate": 2.490170380078637e-06, "loss": 0.5781, "step": 2996 }, { "epoch": 3.8998048145738453, "grad_norm": 1.0721720457077026, "learning_rate": 2.4574049803407603e-06, "loss": 0.2842, "step": 2997 }, { "epoch": 3.901106050748211, "grad_norm": 1.1454334259033203, "learning_rate": 2.4246395806028834e-06, "loss": 0.7311, "step": 2998 }, { "epoch": 3.9024072869225765, "grad_norm": 1.3655797243118286, "learning_rate": 2.3918741808650064e-06, "loss": 0.5154, "step": 2999 }, { "epoch": 3.903708523096942, "grad_norm": 1.4753929376602173, "learning_rate": 2.35910878112713e-06, "loss": 0.504, "step": 3000 } ], "logging_steps": 1, "max_steps": 3072, "num_input_tokens_seen": 0, "num_train_epochs": 4, "save_steps": 500, "total_flos": 2.668216108169134e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null }