{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 6786, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.000442265875502156, "grad_norm": 0.8781153575786518, "learning_rate": 0.0, "loss": 0.4272, "step": 1 }, { "epoch": 0.000884531751004312, "grad_norm": 0.6218387790131081, "learning_rate": 1.4727540500736378e-08, "loss": 0.4277, "step": 2 }, { "epoch": 0.0013267976265064682, "grad_norm": 0.8769587450624082, "learning_rate": 2.9455081001472756e-08, "loss": 0.4038, "step": 3 }, { "epoch": 0.001769063502008624, "grad_norm": 0.7494292654535274, "learning_rate": 4.418262150220913e-08, "loss": 0.4318, "step": 4 }, { "epoch": 0.00221132937751078, "grad_norm": 0.9227272561856895, "learning_rate": 5.891016200294551e-08, "loss": 0.4493, "step": 5 }, { "epoch": 0.0026535952530129365, "grad_norm": 0.7229946939156365, "learning_rate": 7.363770250368189e-08, "loss": 0.4334, "step": 6 }, { "epoch": 0.0030958611285150923, "grad_norm": 0.878110224094956, "learning_rate": 8.836524300441826e-08, "loss": 0.437, "step": 7 }, { "epoch": 0.003538127004017248, "grad_norm": 0.881104436555977, "learning_rate": 1.0309278350515465e-07, "loss": 0.46, "step": 8 }, { "epoch": 0.0039803928795194045, "grad_norm": 0.6415900131989126, "learning_rate": 1.1782032400589102e-07, "loss": 0.3987, "step": 9 }, { "epoch": 0.00442265875502156, "grad_norm": 0.6988098606655294, "learning_rate": 1.325478645066274e-07, "loss": 0.4246, "step": 10 }, { "epoch": 0.004864924630523716, "grad_norm": 1.097199254373804, "learning_rate": 1.4727540500736377e-07, "loss": 0.5093, "step": 11 }, { "epoch": 0.005307190506025873, "grad_norm": 0.7255478392281051, "learning_rate": 1.6200294550810016e-07, "loss": 0.4288, "step": 12 }, { "epoch": 0.005749456381528029, "grad_norm": 0.826910452370686, "learning_rate": 1.7673048600883652e-07, "loss": 0.4012, "step": 13 }, { "epoch": 0.006191722257030185, "grad_norm": 0.5958689554787003, "learning_rate": 1.9145802650957294e-07, "loss": 0.3574, "step": 14 }, { "epoch": 0.0066339881325323405, "grad_norm": 0.6974241027115355, "learning_rate": 2.061855670103093e-07, "loss": 0.4437, "step": 15 }, { "epoch": 0.007076254008034496, "grad_norm": 0.7136486353835451, "learning_rate": 2.209131075110457e-07, "loss": 0.47, "step": 16 }, { "epoch": 0.007518519883536653, "grad_norm": 0.7299487584486921, "learning_rate": 2.3564064801178205e-07, "loss": 0.4656, "step": 17 }, { "epoch": 0.007960785759038809, "grad_norm": 0.6395683338750987, "learning_rate": 2.5036818851251844e-07, "loss": 0.4292, "step": 18 }, { "epoch": 0.008403051634540965, "grad_norm": 0.6201590732592449, "learning_rate": 2.650957290132548e-07, "loss": 0.3943, "step": 19 }, { "epoch": 0.00884531751004312, "grad_norm": 0.7187273131018803, "learning_rate": 2.798232695139912e-07, "loss": 0.4277, "step": 20 }, { "epoch": 0.009287583385545277, "grad_norm": 0.6855828283103271, "learning_rate": 2.9455081001472755e-07, "loss": 0.4227, "step": 21 }, { "epoch": 0.009729849261047432, "grad_norm": 0.776632929635901, "learning_rate": 3.0927835051546394e-07, "loss": 0.4376, "step": 22 }, { "epoch": 0.010172115136549588, "grad_norm": 1.9296659665901026, "learning_rate": 3.240058910162003e-07, "loss": 0.605, "step": 23 }, { "epoch": 0.010614381012051746, "grad_norm": 0.58587245095869, "learning_rate": 3.3873343151693666e-07, "loss": 0.3805, "step": 24 }, { "epoch": 0.011056646887553902, "grad_norm": 0.6492002321286227, "learning_rate": 3.5346097201767305e-07, "loss": 0.4263, "step": 25 }, { "epoch": 0.011498912763056058, "grad_norm": 0.6434502545200099, "learning_rate": 3.681885125184095e-07, "loss": 0.3841, "step": 26 }, { "epoch": 0.011941178638558213, "grad_norm": 0.8067024224269329, "learning_rate": 3.829160530191459e-07, "loss": 0.4247, "step": 27 }, { "epoch": 0.01238344451406037, "grad_norm": 0.6877951228970819, "learning_rate": 3.976435935198822e-07, "loss": 0.4624, "step": 28 }, { "epoch": 0.012825710389562525, "grad_norm": 0.7172108484771123, "learning_rate": 4.123711340206186e-07, "loss": 0.4743, "step": 29 }, { "epoch": 0.013267976265064681, "grad_norm": 0.6683900304424291, "learning_rate": 4.27098674521355e-07, "loss": 0.4224, "step": 30 }, { "epoch": 0.013710242140566837, "grad_norm": 0.7056919084981061, "learning_rate": 4.418262150220914e-07, "loss": 0.384, "step": 31 }, { "epoch": 0.014152508016068993, "grad_norm": 0.6029019562491552, "learning_rate": 4.565537555228277e-07, "loss": 0.3975, "step": 32 }, { "epoch": 0.01459477389157115, "grad_norm": 0.6344373144035546, "learning_rate": 4.712812960235641e-07, "loss": 0.4515, "step": 33 }, { "epoch": 0.015037039767073306, "grad_norm": 0.47277012585096406, "learning_rate": 4.860088365243005e-07, "loss": 0.3849, "step": 34 }, { "epoch": 0.015479305642575462, "grad_norm": 0.5723931986008517, "learning_rate": 5.007363770250369e-07, "loss": 0.4122, "step": 35 }, { "epoch": 0.015921571518077618, "grad_norm": 0.4571928632354695, "learning_rate": 5.154639175257732e-07, "loss": 0.392, "step": 36 }, { "epoch": 0.016363837393579772, "grad_norm": 0.46937479067886817, "learning_rate": 5.301914580265096e-07, "loss": 0.4259, "step": 37 }, { "epoch": 0.01680610326908193, "grad_norm": 0.4764279209097324, "learning_rate": 5.44918998527246e-07, "loss": 0.3716, "step": 38 }, { "epoch": 0.017248369144584087, "grad_norm": 0.48255403675009434, "learning_rate": 5.596465390279824e-07, "loss": 0.4221, "step": 39 }, { "epoch": 0.01769063502008624, "grad_norm": 0.4945624969353443, "learning_rate": 5.743740795287188e-07, "loss": 0.4234, "step": 40 }, { "epoch": 0.0181329008955884, "grad_norm": 0.687872574934456, "learning_rate": 5.891016200294551e-07, "loss": 0.4581, "step": 41 }, { "epoch": 0.018575166771090553, "grad_norm": 0.5357082687419543, "learning_rate": 6.038291605301915e-07, "loss": 0.4597, "step": 42 }, { "epoch": 0.01901743264659271, "grad_norm": 0.5140001611741034, "learning_rate": 6.185567010309279e-07, "loss": 0.3908, "step": 43 }, { "epoch": 0.019459698522094865, "grad_norm": 0.5712432103182022, "learning_rate": 6.332842415316642e-07, "loss": 0.4223, "step": 44 }, { "epoch": 0.019901964397597022, "grad_norm": 0.42852942368391367, "learning_rate": 6.480117820324006e-07, "loss": 0.4025, "step": 45 }, { "epoch": 0.020344230273099177, "grad_norm": 0.3838406161567575, "learning_rate": 6.62739322533137e-07, "loss": 0.4042, "step": 46 }, { "epoch": 0.020786496148601334, "grad_norm": 0.4000780549680701, "learning_rate": 6.774668630338733e-07, "loss": 0.3606, "step": 47 }, { "epoch": 0.021228762024103492, "grad_norm": 0.4050216833367377, "learning_rate": 6.921944035346099e-07, "loss": 0.3991, "step": 48 }, { "epoch": 0.021671027899605646, "grad_norm": 0.4153450697489132, "learning_rate": 7.069219440353461e-07, "loss": 0.386, "step": 49 }, { "epoch": 0.022113293775107803, "grad_norm": 0.5019351706490736, "learning_rate": 7.216494845360824e-07, "loss": 0.4457, "step": 50 }, { "epoch": 0.022555559650609958, "grad_norm": 0.42598897927123414, "learning_rate": 7.36377025036819e-07, "loss": 0.4279, "step": 51 }, { "epoch": 0.022997825526112115, "grad_norm": 0.4958055929869116, "learning_rate": 7.511045655375553e-07, "loss": 0.48, "step": 52 }, { "epoch": 0.02344009140161427, "grad_norm": 0.46030430174893666, "learning_rate": 7.658321060382918e-07, "loss": 0.3903, "step": 53 }, { "epoch": 0.023882357277116427, "grad_norm": 0.5223045974517452, "learning_rate": 7.805596465390281e-07, "loss": 0.4384, "step": 54 }, { "epoch": 0.02432462315261858, "grad_norm": 1.3413427810954817, "learning_rate": 7.952871870397644e-07, "loss": 0.5965, "step": 55 }, { "epoch": 0.02476688902812074, "grad_norm": 0.3979685487112107, "learning_rate": 8.100147275405009e-07, "loss": 0.3954, "step": 56 }, { "epoch": 0.025209154903622896, "grad_norm": 0.3833002448223037, "learning_rate": 8.247422680412372e-07, "loss": 0.3581, "step": 57 }, { "epoch": 0.02565142077912505, "grad_norm": 0.3801253222506262, "learning_rate": 8.394698085419735e-07, "loss": 0.3719, "step": 58 }, { "epoch": 0.026093686654627208, "grad_norm": 0.4003745706044588, "learning_rate": 8.5419734904271e-07, "loss": 0.3694, "step": 59 }, { "epoch": 0.026535952530129362, "grad_norm": 0.44680953590157996, "learning_rate": 8.689248895434463e-07, "loss": 0.4328, "step": 60 }, { "epoch": 0.02697821840563152, "grad_norm": 0.3800722899648336, "learning_rate": 8.836524300441827e-07, "loss": 0.3691, "step": 61 }, { "epoch": 0.027420484281133674, "grad_norm": 0.4264407101598476, "learning_rate": 8.983799705449191e-07, "loss": 0.3854, "step": 62 }, { "epoch": 0.02786275015663583, "grad_norm": 0.37921447118918217, "learning_rate": 9.131075110456554e-07, "loss": 0.4232, "step": 63 }, { "epoch": 0.028305016032137986, "grad_norm": 0.5146829507172227, "learning_rate": 9.278350515463919e-07, "loss": 0.4098, "step": 64 }, { "epoch": 0.028747281907640143, "grad_norm": 0.46045353143096246, "learning_rate": 9.425625920471282e-07, "loss": 0.4137, "step": 65 }, { "epoch": 0.0291895477831423, "grad_norm": 0.3627792916920898, "learning_rate": 9.572901325478646e-07, "loss": 0.374, "step": 66 }, { "epoch": 0.029631813658644455, "grad_norm": 0.3961831381942691, "learning_rate": 9.72017673048601e-07, "loss": 0.426, "step": 67 }, { "epoch": 0.030074079534146612, "grad_norm": 0.39770631375179455, "learning_rate": 9.867452135493373e-07, "loss": 0.4166, "step": 68 }, { "epoch": 0.030516345409648767, "grad_norm": 0.329063732268352, "learning_rate": 1.0014727540500737e-06, "loss": 0.3741, "step": 69 }, { "epoch": 0.030958611285150924, "grad_norm": 0.37367245658532544, "learning_rate": 1.0162002945508102e-06, "loss": 0.4042, "step": 70 }, { "epoch": 0.03140087716065308, "grad_norm": 0.4166857290056908, "learning_rate": 1.0309278350515464e-06, "loss": 0.4232, "step": 71 }, { "epoch": 0.031843143036155236, "grad_norm": 0.37551088303988994, "learning_rate": 1.0456553755522829e-06, "loss": 0.3962, "step": 72 }, { "epoch": 0.03228540891165739, "grad_norm": 0.44251410793295, "learning_rate": 1.0603829160530193e-06, "loss": 0.3842, "step": 73 }, { "epoch": 0.032727674787159544, "grad_norm": 0.4599871080007012, "learning_rate": 1.0751104565537557e-06, "loss": 0.4311, "step": 74 }, { "epoch": 0.033169940662661705, "grad_norm": 0.4175178575183878, "learning_rate": 1.089837997054492e-06, "loss": 0.4299, "step": 75 }, { "epoch": 0.03361220653816386, "grad_norm": 0.3496784265320717, "learning_rate": 1.1045655375552284e-06, "loss": 0.3823, "step": 76 }, { "epoch": 0.03405447241366601, "grad_norm": 0.4710169321910289, "learning_rate": 1.1192930780559649e-06, "loss": 0.3915, "step": 77 }, { "epoch": 0.034496738289168175, "grad_norm": 0.4629751057242775, "learning_rate": 1.134020618556701e-06, "loss": 0.3878, "step": 78 }, { "epoch": 0.03493900416467033, "grad_norm": 0.5609159851710638, "learning_rate": 1.1487481590574375e-06, "loss": 0.4564, "step": 79 }, { "epoch": 0.03538127004017248, "grad_norm": 0.40818003727773317, "learning_rate": 1.163475699558174e-06, "loss": 0.3917, "step": 80 }, { "epoch": 0.03582353591567464, "grad_norm": 0.3238368056234642, "learning_rate": 1.1782032400589102e-06, "loss": 0.4017, "step": 81 }, { "epoch": 0.0362658017911768, "grad_norm": 0.37527863764670505, "learning_rate": 1.1929307805596466e-06, "loss": 0.4132, "step": 82 }, { "epoch": 0.03670806766667895, "grad_norm": 0.42752046901608387, "learning_rate": 1.207658321060383e-06, "loss": 0.4236, "step": 83 }, { "epoch": 0.037150333542181106, "grad_norm": 0.30933561572380097, "learning_rate": 1.2223858615611193e-06, "loss": 0.3817, "step": 84 }, { "epoch": 0.03759259941768327, "grad_norm": 0.360726874022633, "learning_rate": 1.2371134020618557e-06, "loss": 0.3793, "step": 85 }, { "epoch": 0.03803486529318542, "grad_norm": 0.348215037505524, "learning_rate": 1.2518409425625922e-06, "loss": 0.3673, "step": 86 }, { "epoch": 0.038477131168687576, "grad_norm": 0.31163796442697844, "learning_rate": 1.2665684830633284e-06, "loss": 0.391, "step": 87 }, { "epoch": 0.03891939704418973, "grad_norm": 0.7748340404764231, "learning_rate": 1.2812960235640649e-06, "loss": 0.4984, "step": 88 }, { "epoch": 0.03936166291969189, "grad_norm": 0.45595855836040716, "learning_rate": 1.2960235640648013e-06, "loss": 0.4142, "step": 89 }, { "epoch": 0.039803928795194045, "grad_norm": 0.39312776258825854, "learning_rate": 1.3107511045655377e-06, "loss": 0.4171, "step": 90 }, { "epoch": 0.0402461946706962, "grad_norm": 0.4627097488654707, "learning_rate": 1.325478645066274e-06, "loss": 0.435, "step": 91 }, { "epoch": 0.04068846054619835, "grad_norm": 0.38404790060883637, "learning_rate": 1.3402061855670104e-06, "loss": 0.3733, "step": 92 }, { "epoch": 0.041130726421700514, "grad_norm": 0.3866485563411229, "learning_rate": 1.3549337260677466e-06, "loss": 0.388, "step": 93 }, { "epoch": 0.04157299229720267, "grad_norm": 0.3022031276063246, "learning_rate": 1.3696612665684833e-06, "loss": 0.4046, "step": 94 }, { "epoch": 0.04201525817270482, "grad_norm": 0.4251208913576873, "learning_rate": 1.3843888070692197e-06, "loss": 0.4235, "step": 95 }, { "epoch": 0.042457524048206984, "grad_norm": 0.42652095671890927, "learning_rate": 1.399116347569956e-06, "loss": 0.4048, "step": 96 }, { "epoch": 0.04289978992370914, "grad_norm": 0.3457767272834785, "learning_rate": 1.4138438880706922e-06, "loss": 0.4444, "step": 97 }, { "epoch": 0.04334205579921129, "grad_norm": 0.3420664989551918, "learning_rate": 1.4285714285714286e-06, "loss": 0.4237, "step": 98 }, { "epoch": 0.043784321674713446, "grad_norm": 0.35160049382334463, "learning_rate": 1.4432989690721649e-06, "loss": 0.3891, "step": 99 }, { "epoch": 0.04422658755021561, "grad_norm": 0.3763603246563011, "learning_rate": 1.4580265095729015e-06, "loss": 0.3859, "step": 100 }, { "epoch": 0.04466885342571776, "grad_norm": 0.3651265256757238, "learning_rate": 1.472754050073638e-06, "loss": 0.3929, "step": 101 }, { "epoch": 0.045111119301219915, "grad_norm": 0.3337650557611495, "learning_rate": 1.4874815905743742e-06, "loss": 0.3281, "step": 102 }, { "epoch": 0.04555338517672207, "grad_norm": 0.4229459624753358, "learning_rate": 1.5022091310751106e-06, "loss": 0.4006, "step": 103 }, { "epoch": 0.04599565105222423, "grad_norm": 0.3707842099793697, "learning_rate": 1.5169366715758468e-06, "loss": 0.3809, "step": 104 }, { "epoch": 0.046437916927726385, "grad_norm": 0.4238070068420326, "learning_rate": 1.5316642120765835e-06, "loss": 0.4548, "step": 105 }, { "epoch": 0.04688018280322854, "grad_norm": 0.31118302312772483, "learning_rate": 1.5463917525773197e-06, "loss": 0.4073, "step": 106 }, { "epoch": 0.0473224486787307, "grad_norm": 0.2949787821467295, "learning_rate": 1.5611192930780562e-06, "loss": 0.3557, "step": 107 }, { "epoch": 0.047764714554232854, "grad_norm": 0.3082836860837132, "learning_rate": 1.5758468335787924e-06, "loss": 0.404, "step": 108 }, { "epoch": 0.04820698042973501, "grad_norm": 0.3932477480898268, "learning_rate": 1.5905743740795288e-06, "loss": 0.3696, "step": 109 }, { "epoch": 0.04864924630523716, "grad_norm": 0.3848331155061798, "learning_rate": 1.6053019145802653e-06, "loss": 0.4089, "step": 110 }, { "epoch": 0.04909151218073932, "grad_norm": 0.41545706462478676, "learning_rate": 1.6200294550810017e-06, "loss": 0.4023, "step": 111 }, { "epoch": 0.04953377805624148, "grad_norm": 0.39778337309183687, "learning_rate": 1.634756995581738e-06, "loss": 0.4712, "step": 112 }, { "epoch": 0.04997604393174363, "grad_norm": 0.3010913388433542, "learning_rate": 1.6494845360824744e-06, "loss": 0.3944, "step": 113 }, { "epoch": 0.05041830980724579, "grad_norm": 0.5029100791524123, "learning_rate": 1.6642120765832106e-06, "loss": 0.4271, "step": 114 }, { "epoch": 0.05086057568274795, "grad_norm": 0.36215994530043855, "learning_rate": 1.678939617083947e-06, "loss": 0.3659, "step": 115 }, { "epoch": 0.0513028415582501, "grad_norm": 0.27003392716649127, "learning_rate": 1.6936671575846835e-06, "loss": 0.376, "step": 116 }, { "epoch": 0.051745107433752255, "grad_norm": 0.3830543203262692, "learning_rate": 1.70839469808542e-06, "loss": 0.3785, "step": 117 }, { "epoch": 0.052187373309254416, "grad_norm": 0.3756654959878932, "learning_rate": 1.7231222385861562e-06, "loss": 0.4014, "step": 118 }, { "epoch": 0.05262963918475657, "grad_norm": 0.3898802910657805, "learning_rate": 1.7378497790868926e-06, "loss": 0.4306, "step": 119 }, { "epoch": 0.053071905060258724, "grad_norm": 0.30722226059133356, "learning_rate": 1.7525773195876288e-06, "loss": 0.3898, "step": 120 }, { "epoch": 0.05351417093576088, "grad_norm": 0.30056857362921896, "learning_rate": 1.7673048600883655e-06, "loss": 0.3785, "step": 121 }, { "epoch": 0.05395643681126304, "grad_norm": 0.34915967143589566, "learning_rate": 1.7820324005891017e-06, "loss": 0.3954, "step": 122 }, { "epoch": 0.054398702686765193, "grad_norm": 0.30733143614088954, "learning_rate": 1.7967599410898382e-06, "loss": 0.3848, "step": 123 }, { "epoch": 0.05484096856226735, "grad_norm": 0.2967344611676499, "learning_rate": 1.8114874815905744e-06, "loss": 0.3735, "step": 124 }, { "epoch": 0.05528323443776951, "grad_norm": 0.38791918308844153, "learning_rate": 1.8262150220913108e-06, "loss": 0.3823, "step": 125 }, { "epoch": 0.05572550031327166, "grad_norm": 0.35562127629510765, "learning_rate": 1.8409425625920475e-06, "loss": 0.3833, "step": 126 }, { "epoch": 0.05616776618877382, "grad_norm": 0.33647462175871834, "learning_rate": 1.8556701030927837e-06, "loss": 0.3899, "step": 127 }, { "epoch": 0.05661003206427597, "grad_norm": 0.39465087079987904, "learning_rate": 1.87039764359352e-06, "loss": 0.3781, "step": 128 }, { "epoch": 0.05705229793977813, "grad_norm": 0.36928779581793963, "learning_rate": 1.8851251840942564e-06, "loss": 0.4032, "step": 129 }, { "epoch": 0.057494563815280286, "grad_norm": 0.35790807310427497, "learning_rate": 1.8998527245949926e-06, "loss": 0.4093, "step": 130 }, { "epoch": 0.05793682969078244, "grad_norm": 0.3248430351039431, "learning_rate": 1.9145802650957293e-06, "loss": 0.3734, "step": 131 }, { "epoch": 0.0583790955662846, "grad_norm": 0.3902418550153629, "learning_rate": 1.9293078055964655e-06, "loss": 0.4292, "step": 132 }, { "epoch": 0.058821361441786756, "grad_norm": 0.34739110814314483, "learning_rate": 1.944035346097202e-06, "loss": 0.4192, "step": 133 }, { "epoch": 0.05926362731728891, "grad_norm": 0.3294497825797575, "learning_rate": 1.9587628865979384e-06, "loss": 0.3926, "step": 134 }, { "epoch": 0.059705893192791064, "grad_norm": 0.3603546079866457, "learning_rate": 1.9734904270986746e-06, "loss": 0.39, "step": 135 }, { "epoch": 0.060148159068293225, "grad_norm": 0.4217075060807565, "learning_rate": 1.988217967599411e-06, "loss": 0.4238, "step": 136 }, { "epoch": 0.06059042494379538, "grad_norm": 0.32556108229601094, "learning_rate": 2.0029455081001475e-06, "loss": 0.3482, "step": 137 }, { "epoch": 0.06103269081929753, "grad_norm": 0.35242756101017586, "learning_rate": 2.0176730486008837e-06, "loss": 0.4022, "step": 138 }, { "epoch": 0.06147495669479969, "grad_norm": 0.36417922256041335, "learning_rate": 2.0324005891016204e-06, "loss": 0.4008, "step": 139 }, { "epoch": 0.06191722257030185, "grad_norm": 0.4159021084695606, "learning_rate": 2.0471281296023566e-06, "loss": 0.4038, "step": 140 }, { "epoch": 0.062359488445804, "grad_norm": 0.3606644581518755, "learning_rate": 2.061855670103093e-06, "loss": 0.3999, "step": 141 }, { "epoch": 0.06280175432130616, "grad_norm": 0.30057722354978267, "learning_rate": 2.0765832106038295e-06, "loss": 0.3848, "step": 142 }, { "epoch": 0.06324402019680832, "grad_norm": 0.30106382894828265, "learning_rate": 2.0913107511045657e-06, "loss": 0.3658, "step": 143 }, { "epoch": 0.06368628607231047, "grad_norm": 0.3219724594269278, "learning_rate": 2.106038291605302e-06, "loss": 0.395, "step": 144 }, { "epoch": 0.06412855194781263, "grad_norm": 0.3793713244225711, "learning_rate": 2.1207658321060386e-06, "loss": 0.3742, "step": 145 }, { "epoch": 0.06457081782331478, "grad_norm": 0.3938293777317276, "learning_rate": 2.135493372606775e-06, "loss": 0.4004, "step": 146 }, { "epoch": 0.06501308369881693, "grad_norm": 0.5323261289094346, "learning_rate": 2.1502209131075115e-06, "loss": 0.4591, "step": 147 }, { "epoch": 0.06545534957431909, "grad_norm": 0.2874318173492164, "learning_rate": 2.1649484536082477e-06, "loss": 0.3597, "step": 148 }, { "epoch": 0.06589761544982126, "grad_norm": 0.3098681641780344, "learning_rate": 2.179675994108984e-06, "loss": 0.384, "step": 149 }, { "epoch": 0.06633988132532341, "grad_norm": 0.46341394292697824, "learning_rate": 2.19440353460972e-06, "loss": 0.4561, "step": 150 }, { "epoch": 0.06678214720082556, "grad_norm": 0.4815092624288841, "learning_rate": 2.209131075110457e-06, "loss": 0.4211, "step": 151 }, { "epoch": 0.06722441307632772, "grad_norm": 0.3746318419536617, "learning_rate": 2.223858615611193e-06, "loss": 0.4174, "step": 152 }, { "epoch": 0.06766667895182987, "grad_norm": 0.31762678763720714, "learning_rate": 2.2385861561119297e-06, "loss": 0.4127, "step": 153 }, { "epoch": 0.06810894482733203, "grad_norm": 0.5063852545542253, "learning_rate": 2.253313696612666e-06, "loss": 0.4013, "step": 154 }, { "epoch": 0.06855121070283418, "grad_norm": 0.3569282463598872, "learning_rate": 2.268041237113402e-06, "loss": 0.389, "step": 155 }, { "epoch": 0.06899347657833635, "grad_norm": 0.31431055215929654, "learning_rate": 2.2827687776141384e-06, "loss": 0.4047, "step": 156 }, { "epoch": 0.0694357424538385, "grad_norm": 0.33141752857587353, "learning_rate": 2.297496318114875e-06, "loss": 0.4053, "step": 157 }, { "epoch": 0.06987800832934066, "grad_norm": 0.33564756069739576, "learning_rate": 2.3122238586156117e-06, "loss": 0.3853, "step": 158 }, { "epoch": 0.07032027420484281, "grad_norm": 0.312782330125141, "learning_rate": 2.326951399116348e-06, "loss": 0.3947, "step": 159 }, { "epoch": 0.07076254008034497, "grad_norm": 0.3109459457766325, "learning_rate": 2.341678939617084e-06, "loss": 0.3671, "step": 160 }, { "epoch": 0.07120480595584712, "grad_norm": 0.36052475237698367, "learning_rate": 2.3564064801178204e-06, "loss": 0.3759, "step": 161 }, { "epoch": 0.07164707183134927, "grad_norm": 0.32590219951660593, "learning_rate": 2.3711340206185566e-06, "loss": 0.3591, "step": 162 }, { "epoch": 0.07208933770685144, "grad_norm": 0.3361350763498865, "learning_rate": 2.3858615611192933e-06, "loss": 0.3632, "step": 163 }, { "epoch": 0.0725316035823536, "grad_norm": 0.39648053648575343, "learning_rate": 2.40058910162003e-06, "loss": 0.4122, "step": 164 }, { "epoch": 0.07297386945785575, "grad_norm": 0.373638951745047, "learning_rate": 2.415316642120766e-06, "loss": 0.3736, "step": 165 }, { "epoch": 0.0734161353333579, "grad_norm": 0.3630223050832077, "learning_rate": 2.4300441826215024e-06, "loss": 0.403, "step": 166 }, { "epoch": 0.07385840120886006, "grad_norm": 0.34959154417767807, "learning_rate": 2.4447717231222386e-06, "loss": 0.4017, "step": 167 }, { "epoch": 0.07430066708436221, "grad_norm": 0.35972826022208454, "learning_rate": 2.459499263622975e-06, "loss": 0.3817, "step": 168 }, { "epoch": 0.07474293295986437, "grad_norm": 0.4396491179333677, "learning_rate": 2.4742268041237115e-06, "loss": 0.4224, "step": 169 }, { "epoch": 0.07518519883536653, "grad_norm": 0.31412047908146606, "learning_rate": 2.488954344624448e-06, "loss": 0.4357, "step": 170 }, { "epoch": 0.07562746471086869, "grad_norm": 0.36938243585475644, "learning_rate": 2.5036818851251844e-06, "loss": 0.3555, "step": 171 }, { "epoch": 0.07606973058637084, "grad_norm": 0.3639639844334975, "learning_rate": 2.5184094256259206e-06, "loss": 0.4094, "step": 172 }, { "epoch": 0.076511996461873, "grad_norm": 0.3068498333258657, "learning_rate": 2.533136966126657e-06, "loss": 0.3486, "step": 173 }, { "epoch": 0.07695426233737515, "grad_norm": 0.4502169592953238, "learning_rate": 2.547864506627393e-06, "loss": 0.4101, "step": 174 }, { "epoch": 0.0773965282128773, "grad_norm": 0.46749005348141365, "learning_rate": 2.5625920471281297e-06, "loss": 0.3988, "step": 175 }, { "epoch": 0.07783879408837946, "grad_norm": 0.3205072360171259, "learning_rate": 2.577319587628866e-06, "loss": 0.3911, "step": 176 }, { "epoch": 0.07828105996388161, "grad_norm": 0.33913417242963023, "learning_rate": 2.5920471281296026e-06, "loss": 0.3779, "step": 177 }, { "epoch": 0.07872332583938378, "grad_norm": 0.394222675203982, "learning_rate": 2.6067746686303392e-06, "loss": 0.4404, "step": 178 }, { "epoch": 0.07916559171488594, "grad_norm": 0.3496314618702274, "learning_rate": 2.6215022091310755e-06, "loss": 0.403, "step": 179 }, { "epoch": 0.07960785759038809, "grad_norm": 0.3231478101752516, "learning_rate": 2.6362297496318117e-06, "loss": 0.3921, "step": 180 }, { "epoch": 0.08005012346589024, "grad_norm": 0.3369841539457067, "learning_rate": 2.650957290132548e-06, "loss": 0.394, "step": 181 }, { "epoch": 0.0804923893413924, "grad_norm": 0.35525575283201355, "learning_rate": 2.6656848306332846e-06, "loss": 0.3974, "step": 182 }, { "epoch": 0.08093465521689455, "grad_norm": 0.3402395007695635, "learning_rate": 2.680412371134021e-06, "loss": 0.3643, "step": 183 }, { "epoch": 0.0813769210923967, "grad_norm": 0.31008769719743207, "learning_rate": 2.695139911634757e-06, "loss": 0.3816, "step": 184 }, { "epoch": 0.08181918696789887, "grad_norm": 0.33588171402597344, "learning_rate": 2.7098674521354933e-06, "loss": 0.3854, "step": 185 }, { "epoch": 0.08226145284340103, "grad_norm": 0.35192947178435263, "learning_rate": 2.72459499263623e-06, "loss": 0.3532, "step": 186 }, { "epoch": 0.08270371871890318, "grad_norm": 0.3842340362523307, "learning_rate": 2.7393225331369666e-06, "loss": 0.4046, "step": 187 }, { "epoch": 0.08314598459440534, "grad_norm": 0.3671475395536957, "learning_rate": 2.754050073637703e-06, "loss": 0.4276, "step": 188 }, { "epoch": 0.08358825046990749, "grad_norm": 0.3541344394502841, "learning_rate": 2.7687776141384395e-06, "loss": 0.3863, "step": 189 }, { "epoch": 0.08403051634540964, "grad_norm": 0.27326224962797524, "learning_rate": 2.7835051546391757e-06, "loss": 0.3683, "step": 190 }, { "epoch": 0.0844727822209118, "grad_norm": 0.34054996111752145, "learning_rate": 2.798232695139912e-06, "loss": 0.3724, "step": 191 }, { "epoch": 0.08491504809641397, "grad_norm": 0.36201022109552355, "learning_rate": 2.812960235640648e-06, "loss": 0.3826, "step": 192 }, { "epoch": 0.08535731397191612, "grad_norm": 0.32856868261806366, "learning_rate": 2.8276877761413844e-06, "loss": 0.3749, "step": 193 }, { "epoch": 0.08579957984741828, "grad_norm": 0.323117008264352, "learning_rate": 2.842415316642121e-06, "loss": 0.3887, "step": 194 }, { "epoch": 0.08624184572292043, "grad_norm": 0.2844087954706406, "learning_rate": 2.8571428571428573e-06, "loss": 0.36, "step": 195 }, { "epoch": 0.08668411159842258, "grad_norm": 0.3763929150165721, "learning_rate": 2.8718703976435935e-06, "loss": 0.3997, "step": 196 }, { "epoch": 0.08712637747392474, "grad_norm": 0.343693296477661, "learning_rate": 2.8865979381443297e-06, "loss": 0.4029, "step": 197 }, { "epoch": 0.08756864334942689, "grad_norm": 0.31686233339644826, "learning_rate": 2.901325478645067e-06, "loss": 0.3988, "step": 198 }, { "epoch": 0.08801090922492906, "grad_norm": 0.3764663367928946, "learning_rate": 2.916053019145803e-06, "loss": 0.405, "step": 199 }, { "epoch": 0.08845317510043121, "grad_norm": 0.4220537354895963, "learning_rate": 2.9307805596465392e-06, "loss": 0.3908, "step": 200 }, { "epoch": 0.08889544097593337, "grad_norm": 0.324620697441631, "learning_rate": 2.945508100147276e-06, "loss": 0.3963, "step": 201 }, { "epoch": 0.08933770685143552, "grad_norm": 0.3043162384666142, "learning_rate": 2.960235640648012e-06, "loss": 0.372, "step": 202 }, { "epoch": 0.08977997272693768, "grad_norm": 0.3326623110471265, "learning_rate": 2.9749631811487484e-06, "loss": 0.3962, "step": 203 }, { "epoch": 0.09022223860243983, "grad_norm": 0.33178011696941734, "learning_rate": 2.9896907216494846e-06, "loss": 0.3843, "step": 204 }, { "epoch": 0.09066450447794198, "grad_norm": 0.36873922865112, "learning_rate": 3.0044182621502212e-06, "loss": 0.4061, "step": 205 }, { "epoch": 0.09110677035344414, "grad_norm": 0.30552218987959084, "learning_rate": 3.0191458026509575e-06, "loss": 0.3899, "step": 206 }, { "epoch": 0.0915490362289463, "grad_norm": 0.31472984464881365, "learning_rate": 3.0338733431516937e-06, "loss": 0.4231, "step": 207 }, { "epoch": 0.09199130210444846, "grad_norm": 0.3212229875346074, "learning_rate": 3.0486008836524308e-06, "loss": 0.4203, "step": 208 }, { "epoch": 0.09243356797995061, "grad_norm": 0.34062379262705705, "learning_rate": 3.063328424153167e-06, "loss": 0.4188, "step": 209 }, { "epoch": 0.09287583385545277, "grad_norm": 0.35457864452692844, "learning_rate": 3.0780559646539032e-06, "loss": 0.3826, "step": 210 }, { "epoch": 0.09331809973095492, "grad_norm": 0.33108673301308916, "learning_rate": 3.0927835051546395e-06, "loss": 0.4095, "step": 211 }, { "epoch": 0.09376036560645708, "grad_norm": 0.35847261611611253, "learning_rate": 3.1075110456553757e-06, "loss": 0.411, "step": 212 }, { "epoch": 0.09420263148195923, "grad_norm": 0.33794948687450477, "learning_rate": 3.1222385861561123e-06, "loss": 0.3844, "step": 213 }, { "epoch": 0.0946448973574614, "grad_norm": 0.2970217549491036, "learning_rate": 3.1369661266568486e-06, "loss": 0.3721, "step": 214 }, { "epoch": 0.09508716323296355, "grad_norm": 0.3353340679229872, "learning_rate": 3.151693667157585e-06, "loss": 0.3844, "step": 215 }, { "epoch": 0.09552942910846571, "grad_norm": 0.37756946346361414, "learning_rate": 3.166421207658321e-06, "loss": 0.3845, "step": 216 }, { "epoch": 0.09597169498396786, "grad_norm": 0.37378232722589083, "learning_rate": 3.1811487481590577e-06, "loss": 0.3902, "step": 217 }, { "epoch": 0.09641396085947002, "grad_norm": 0.34410080346817207, "learning_rate": 3.195876288659794e-06, "loss": 0.3872, "step": 218 }, { "epoch": 0.09685622673497217, "grad_norm": 0.5271760360924385, "learning_rate": 3.2106038291605306e-06, "loss": 0.419, "step": 219 }, { "epoch": 0.09729849261047432, "grad_norm": 0.3134250719077149, "learning_rate": 3.2253313696612672e-06, "loss": 0.3641, "step": 220 }, { "epoch": 0.09774075848597649, "grad_norm": 0.35937479590077065, "learning_rate": 3.2400589101620034e-06, "loss": 0.43, "step": 221 }, { "epoch": 0.09818302436147865, "grad_norm": 0.31387046273563507, "learning_rate": 3.2547864506627397e-06, "loss": 0.3878, "step": 222 }, { "epoch": 0.0986252902369808, "grad_norm": 0.4024634045057449, "learning_rate": 3.269513991163476e-06, "loss": 0.3816, "step": 223 }, { "epoch": 0.09906755611248295, "grad_norm": 0.304708448812162, "learning_rate": 3.284241531664212e-06, "loss": 0.3869, "step": 224 }, { "epoch": 0.09950982198798511, "grad_norm": 0.49444886271403454, "learning_rate": 3.298969072164949e-06, "loss": 0.4471, "step": 225 }, { "epoch": 0.09995208786348726, "grad_norm": 0.3176873857820558, "learning_rate": 3.313696612665685e-06, "loss": 0.3714, "step": 226 }, { "epoch": 0.10039435373898942, "grad_norm": 0.343008775520098, "learning_rate": 3.3284241531664212e-06, "loss": 0.4443, "step": 227 }, { "epoch": 0.10083661961449158, "grad_norm": 0.32452046963142206, "learning_rate": 3.3431516936671575e-06, "loss": 0.3858, "step": 228 }, { "epoch": 0.10127888548999374, "grad_norm": 0.42460325769058543, "learning_rate": 3.357879234167894e-06, "loss": 0.3726, "step": 229 }, { "epoch": 0.1017211513654959, "grad_norm": 0.376509591495129, "learning_rate": 3.3726067746686308e-06, "loss": 0.4261, "step": 230 }, { "epoch": 0.10216341724099805, "grad_norm": 0.3305185062486174, "learning_rate": 3.387334315169367e-06, "loss": 0.3695, "step": 231 }, { "epoch": 0.1026056831165002, "grad_norm": 0.30208394602243477, "learning_rate": 3.4020618556701037e-06, "loss": 0.402, "step": 232 }, { "epoch": 0.10304794899200236, "grad_norm": 0.43157920111942905, "learning_rate": 3.41678939617084e-06, "loss": 0.3855, "step": 233 }, { "epoch": 0.10349021486750451, "grad_norm": 0.40830970991876736, "learning_rate": 3.431516936671576e-06, "loss": 0.3985, "step": 234 }, { "epoch": 0.10393248074300668, "grad_norm": 0.3709627419850503, "learning_rate": 3.4462444771723124e-06, "loss": 0.3979, "step": 235 }, { "epoch": 0.10437474661850883, "grad_norm": 0.31798050848609016, "learning_rate": 3.460972017673049e-06, "loss": 0.3972, "step": 236 }, { "epoch": 0.10481701249401099, "grad_norm": 0.29623761522565784, "learning_rate": 3.4756995581737852e-06, "loss": 0.3644, "step": 237 }, { "epoch": 0.10525927836951314, "grad_norm": 0.29335220484247104, "learning_rate": 3.4904270986745215e-06, "loss": 0.3985, "step": 238 }, { "epoch": 0.1057015442450153, "grad_norm": 0.4191380586961168, "learning_rate": 3.5051546391752577e-06, "loss": 0.4046, "step": 239 }, { "epoch": 0.10614381012051745, "grad_norm": 0.33088992143710816, "learning_rate": 3.5198821796759948e-06, "loss": 0.3969, "step": 240 }, { "epoch": 0.1065860759960196, "grad_norm": 0.3074267779194083, "learning_rate": 3.534609720176731e-06, "loss": 0.3244, "step": 241 }, { "epoch": 0.10702834187152176, "grad_norm": 0.3405615502775006, "learning_rate": 3.5493372606774672e-06, "loss": 0.3964, "step": 242 }, { "epoch": 0.10747060774702392, "grad_norm": 0.347947069279981, "learning_rate": 3.5640648011782035e-06, "loss": 0.3985, "step": 243 }, { "epoch": 0.10791287362252608, "grad_norm": 0.3219749714076349, "learning_rate": 3.57879234167894e-06, "loss": 0.3608, "step": 244 }, { "epoch": 0.10835513949802823, "grad_norm": 0.3593933663428767, "learning_rate": 3.5935198821796763e-06, "loss": 0.3828, "step": 245 }, { "epoch": 0.10879740537353039, "grad_norm": 0.39086353794041623, "learning_rate": 3.6082474226804126e-06, "loss": 0.4143, "step": 246 }, { "epoch": 0.10923967124903254, "grad_norm": 0.3020465688477811, "learning_rate": 3.622974963181149e-06, "loss": 0.3728, "step": 247 }, { "epoch": 0.1096819371245347, "grad_norm": 0.35622676444190327, "learning_rate": 3.6377025036818854e-06, "loss": 0.4057, "step": 248 }, { "epoch": 0.11012420300003685, "grad_norm": 0.45149998487674425, "learning_rate": 3.6524300441826217e-06, "loss": 0.4209, "step": 249 }, { "epoch": 0.11056646887553902, "grad_norm": 0.3531184330500681, "learning_rate": 3.667157584683358e-06, "loss": 0.4024, "step": 250 }, { "epoch": 0.11100873475104117, "grad_norm": 0.33341858894278575, "learning_rate": 3.681885125184095e-06, "loss": 0.3606, "step": 251 }, { "epoch": 0.11145100062654333, "grad_norm": 0.3937802658245101, "learning_rate": 3.696612665684831e-06, "loss": 0.3743, "step": 252 }, { "epoch": 0.11189326650204548, "grad_norm": 0.32081735820039897, "learning_rate": 3.7113402061855674e-06, "loss": 0.3523, "step": 253 }, { "epoch": 0.11233553237754763, "grad_norm": 0.3297143113205817, "learning_rate": 3.7260677466863037e-06, "loss": 0.3575, "step": 254 }, { "epoch": 0.11277779825304979, "grad_norm": 0.3022027145139146, "learning_rate": 3.74079528718704e-06, "loss": 0.4452, "step": 255 }, { "epoch": 0.11322006412855194, "grad_norm": 0.3168411920875108, "learning_rate": 3.7555228276877766e-06, "loss": 0.4153, "step": 256 }, { "epoch": 0.11366233000405411, "grad_norm": 0.48309425506022285, "learning_rate": 3.7702503681885128e-06, "loss": 0.4567, "step": 257 }, { "epoch": 0.11410459587955626, "grad_norm": 0.3095805820480119, "learning_rate": 3.784977908689249e-06, "loss": 0.4203, "step": 258 }, { "epoch": 0.11454686175505842, "grad_norm": 0.5009536282751612, "learning_rate": 3.7997054491899852e-06, "loss": 0.4256, "step": 259 }, { "epoch": 0.11498912763056057, "grad_norm": 0.39230259006481016, "learning_rate": 3.814432989690722e-06, "loss": 0.3969, "step": 260 }, { "epoch": 0.11543139350606273, "grad_norm": 0.3458816848088728, "learning_rate": 3.8291605301914585e-06, "loss": 0.3968, "step": 261 }, { "epoch": 0.11587365938156488, "grad_norm": 0.3200788418476786, "learning_rate": 3.843888070692195e-06, "loss": 0.3851, "step": 262 }, { "epoch": 0.11631592525706703, "grad_norm": 0.3380258497302857, "learning_rate": 3.858615611192931e-06, "loss": 0.4385, "step": 263 }, { "epoch": 0.1167581911325692, "grad_norm": 0.35202949974003367, "learning_rate": 3.873343151693668e-06, "loss": 0.4218, "step": 264 }, { "epoch": 0.11720045700807136, "grad_norm": 0.3159289979034962, "learning_rate": 3.888070692194404e-06, "loss": 0.3873, "step": 265 }, { "epoch": 0.11764272288357351, "grad_norm": 0.3715387237831565, "learning_rate": 3.90279823269514e-06, "loss": 0.4049, "step": 266 }, { "epoch": 0.11808498875907567, "grad_norm": 0.30223298927339876, "learning_rate": 3.917525773195877e-06, "loss": 0.3975, "step": 267 }, { "epoch": 0.11852725463457782, "grad_norm": 0.3432347954115166, "learning_rate": 3.9322533136966126e-06, "loss": 0.3698, "step": 268 }, { "epoch": 0.11896952051007997, "grad_norm": 0.3532281014135538, "learning_rate": 3.946980854197349e-06, "loss": 0.3805, "step": 269 }, { "epoch": 0.11941178638558213, "grad_norm": 0.4458597617524757, "learning_rate": 3.961708394698086e-06, "loss": 0.3855, "step": 270 }, { "epoch": 0.11985405226108428, "grad_norm": 0.3076339219294476, "learning_rate": 3.976435935198822e-06, "loss": 0.3741, "step": 271 }, { "epoch": 0.12029631813658645, "grad_norm": 0.4109317619872949, "learning_rate": 3.991163475699558e-06, "loss": 0.4028, "step": 272 }, { "epoch": 0.1207385840120886, "grad_norm": 0.3285313235112913, "learning_rate": 4.005891016200295e-06, "loss": 0.3594, "step": 273 }, { "epoch": 0.12118084988759076, "grad_norm": 0.3564992702306064, "learning_rate": 4.020618556701032e-06, "loss": 0.3944, "step": 274 }, { "epoch": 0.12162311576309291, "grad_norm": 0.4147439837606847, "learning_rate": 4.0353460972017674e-06, "loss": 0.4087, "step": 275 }, { "epoch": 0.12206538163859507, "grad_norm": 0.3215862708245971, "learning_rate": 4.050073637702504e-06, "loss": 0.3791, "step": 276 }, { "epoch": 0.12250764751409722, "grad_norm": 0.3634159915402298, "learning_rate": 4.064801178203241e-06, "loss": 0.3915, "step": 277 }, { "epoch": 0.12294991338959937, "grad_norm": 0.36832758888190525, "learning_rate": 4.0795287187039766e-06, "loss": 0.4003, "step": 278 }, { "epoch": 0.12339217926510154, "grad_norm": 0.3002254148585283, "learning_rate": 4.094256259204713e-06, "loss": 0.3738, "step": 279 }, { "epoch": 0.1238344451406037, "grad_norm": 0.2884843204972599, "learning_rate": 4.108983799705449e-06, "loss": 0.377, "step": 280 }, { "epoch": 0.12427671101610585, "grad_norm": 0.3662700714635297, "learning_rate": 4.123711340206186e-06, "loss": 0.402, "step": 281 }, { "epoch": 0.124718976891608, "grad_norm": 0.4749664423545888, "learning_rate": 4.138438880706922e-06, "loss": 0.4348, "step": 282 }, { "epoch": 0.12516124276711016, "grad_norm": 0.3105463805498842, "learning_rate": 4.153166421207659e-06, "loss": 0.3676, "step": 283 }, { "epoch": 0.12560350864261233, "grad_norm": 0.3392337972531778, "learning_rate": 4.167893961708396e-06, "loss": 0.3639, "step": 284 }, { "epoch": 0.12604577451811447, "grad_norm": 0.36065990854879626, "learning_rate": 4.1826215022091314e-06, "loss": 0.3764, "step": 285 }, { "epoch": 0.12648804039361664, "grad_norm": 0.4195562660873782, "learning_rate": 4.197349042709868e-06, "loss": 0.386, "step": 286 }, { "epoch": 0.12693030626911878, "grad_norm": 0.409998422804919, "learning_rate": 4.212076583210604e-06, "loss": 0.4152, "step": 287 }, { "epoch": 0.12737257214462094, "grad_norm": 0.3206801283701468, "learning_rate": 4.2268041237113405e-06, "loss": 0.4098, "step": 288 }, { "epoch": 0.12781483802012308, "grad_norm": 0.3391144905756185, "learning_rate": 4.241531664212077e-06, "loss": 0.3886, "step": 289 }, { "epoch": 0.12825710389562525, "grad_norm": 0.3322735234619148, "learning_rate": 4.256259204712813e-06, "loss": 0.3638, "step": 290 }, { "epoch": 0.12869936977112742, "grad_norm": 0.3607617671770448, "learning_rate": 4.27098674521355e-06, "loss": 0.4065, "step": 291 }, { "epoch": 0.12914163564662956, "grad_norm": 0.35104436018592766, "learning_rate": 4.2857142857142855e-06, "loss": 0.382, "step": 292 }, { "epoch": 0.12958390152213173, "grad_norm": 0.31605513897206333, "learning_rate": 4.300441826215023e-06, "loss": 0.3939, "step": 293 }, { "epoch": 0.13002616739763387, "grad_norm": 0.2760072510774275, "learning_rate": 4.315169366715759e-06, "loss": 0.3472, "step": 294 }, { "epoch": 0.13046843327313604, "grad_norm": 0.3706673879688958, "learning_rate": 4.329896907216495e-06, "loss": 0.3937, "step": 295 }, { "epoch": 0.13091069914863818, "grad_norm": 0.32764224942528397, "learning_rate": 4.344624447717232e-06, "loss": 0.3688, "step": 296 }, { "epoch": 0.13135296502414034, "grad_norm": 0.32713312873665173, "learning_rate": 4.359351988217968e-06, "loss": 0.3832, "step": 297 }, { "epoch": 0.1317952308996425, "grad_norm": 0.3584738828719295, "learning_rate": 4.3740795287187045e-06, "loss": 0.3829, "step": 298 }, { "epoch": 0.13223749677514465, "grad_norm": 0.3173716552079419, "learning_rate": 4.38880706921944e-06, "loss": 0.3811, "step": 299 }, { "epoch": 0.13267976265064682, "grad_norm": 0.2925360175067458, "learning_rate": 4.403534609720177e-06, "loss": 0.3617, "step": 300 }, { "epoch": 0.13312202852614896, "grad_norm": 0.37572241156777303, "learning_rate": 4.418262150220914e-06, "loss": 0.4084, "step": 301 }, { "epoch": 0.13356429440165113, "grad_norm": 0.3714255329766156, "learning_rate": 4.4329896907216494e-06, "loss": 0.3951, "step": 302 }, { "epoch": 0.13400656027715327, "grad_norm": 0.35048268722417847, "learning_rate": 4.447717231222386e-06, "loss": 0.383, "step": 303 }, { "epoch": 0.13444882615265544, "grad_norm": 0.29256748801281435, "learning_rate": 4.462444771723123e-06, "loss": 0.3548, "step": 304 }, { "epoch": 0.1348910920281576, "grad_norm": 0.422451471784086, "learning_rate": 4.477172312223859e-06, "loss": 0.3725, "step": 305 }, { "epoch": 0.13533335790365975, "grad_norm": 0.3378490344248794, "learning_rate": 4.491899852724595e-06, "loss": 0.3836, "step": 306 }, { "epoch": 0.1357756237791619, "grad_norm": 0.48722496061734893, "learning_rate": 4.506627393225332e-06, "loss": 0.444, "step": 307 }, { "epoch": 0.13621788965466405, "grad_norm": 0.4671737182356339, "learning_rate": 4.5213549337260685e-06, "loss": 0.3902, "step": 308 }, { "epoch": 0.13666015553016622, "grad_norm": 0.5842002287364836, "learning_rate": 4.536082474226804e-06, "loss": 0.3828, "step": 309 }, { "epoch": 0.13710242140566836, "grad_norm": 0.337243050379789, "learning_rate": 4.550810014727541e-06, "loss": 0.3906, "step": 310 }, { "epoch": 0.13754468728117053, "grad_norm": 0.39162680206654943, "learning_rate": 4.565537555228277e-06, "loss": 0.3929, "step": 311 }, { "epoch": 0.1379869531566727, "grad_norm": 0.3396750211164344, "learning_rate": 4.5802650957290134e-06, "loss": 0.412, "step": 312 }, { "epoch": 0.13842921903217484, "grad_norm": 0.4423974492133541, "learning_rate": 4.59499263622975e-06, "loss": 0.3818, "step": 313 }, { "epoch": 0.138871484907677, "grad_norm": 0.39044280778726614, "learning_rate": 4.609720176730487e-06, "loss": 0.4429, "step": 314 }, { "epoch": 0.13931375078317915, "grad_norm": 0.6937247135065527, "learning_rate": 4.624447717231223e-06, "loss": 0.5188, "step": 315 }, { "epoch": 0.13975601665868131, "grad_norm": 0.2956503984968021, "learning_rate": 4.639175257731959e-06, "loss": 0.3848, "step": 316 }, { "epoch": 0.14019828253418345, "grad_norm": 0.29957583747447264, "learning_rate": 4.653902798232696e-06, "loss": 0.346, "step": 317 }, { "epoch": 0.14064054840968562, "grad_norm": 0.3458980765130136, "learning_rate": 4.668630338733432e-06, "loss": 0.3823, "step": 318 }, { "epoch": 0.1410828142851878, "grad_norm": 0.3796825485131053, "learning_rate": 4.683357879234168e-06, "loss": 0.3763, "step": 319 }, { "epoch": 0.14152508016068993, "grad_norm": 0.3410103962657027, "learning_rate": 4.698085419734905e-06, "loss": 0.4204, "step": 320 }, { "epoch": 0.1419673460361921, "grad_norm": 0.3139589667747337, "learning_rate": 4.712812960235641e-06, "loss": 0.395, "step": 321 }, { "epoch": 0.14240961191169424, "grad_norm": 0.33338891775086266, "learning_rate": 4.727540500736377e-06, "loss": 0.4248, "step": 322 }, { "epoch": 0.1428518777871964, "grad_norm": 0.373404041539284, "learning_rate": 4.742268041237113e-06, "loss": 0.3924, "step": 323 }, { "epoch": 0.14329414366269855, "grad_norm": 0.314632050462846, "learning_rate": 4.75699558173785e-06, "loss": 0.3753, "step": 324 }, { "epoch": 0.14373640953820072, "grad_norm": 0.3136241503985317, "learning_rate": 4.7717231222385865e-06, "loss": 0.3737, "step": 325 }, { "epoch": 0.14417867541370288, "grad_norm": 0.3720839939766779, "learning_rate": 4.786450662739323e-06, "loss": 0.3986, "step": 326 }, { "epoch": 0.14462094128920502, "grad_norm": 0.35692558764212273, "learning_rate": 4.80117820324006e-06, "loss": 0.3756, "step": 327 }, { "epoch": 0.1450632071647072, "grad_norm": 0.3714991801997026, "learning_rate": 4.815905743740796e-06, "loss": 0.389, "step": 328 }, { "epoch": 0.14550547304020933, "grad_norm": 0.3379503633997025, "learning_rate": 4.830633284241532e-06, "loss": 0.4075, "step": 329 }, { "epoch": 0.1459477389157115, "grad_norm": 0.33691008980273907, "learning_rate": 4.845360824742268e-06, "loss": 0.3818, "step": 330 }, { "epoch": 0.14639000479121364, "grad_norm": 0.3059234825011972, "learning_rate": 4.860088365243005e-06, "loss": 0.3738, "step": 331 }, { "epoch": 0.1468322706667158, "grad_norm": 0.38851572123545525, "learning_rate": 4.874815905743741e-06, "loss": 0.4046, "step": 332 }, { "epoch": 0.14727453654221798, "grad_norm": 0.36423436851017915, "learning_rate": 4.889543446244477e-06, "loss": 0.3903, "step": 333 }, { "epoch": 0.14771680241772012, "grad_norm": 0.4073456249876006, "learning_rate": 4.904270986745214e-06, "loss": 0.3995, "step": 334 }, { "epoch": 0.14815906829322228, "grad_norm": 0.35019879358046363, "learning_rate": 4.91899852724595e-06, "loss": 0.3622, "step": 335 }, { "epoch": 0.14860133416872442, "grad_norm": 0.2852296622952607, "learning_rate": 4.933726067746687e-06, "loss": 0.3847, "step": 336 }, { "epoch": 0.1490436000442266, "grad_norm": 0.42001170741725236, "learning_rate": 4.948453608247423e-06, "loss": 0.4708, "step": 337 }, { "epoch": 0.14948586591972873, "grad_norm": 0.3589168205378101, "learning_rate": 4.96318114874816e-06, "loss": 0.3626, "step": 338 }, { "epoch": 0.1499281317952309, "grad_norm": 0.3124928419826664, "learning_rate": 4.977908689248896e-06, "loss": 0.3943, "step": 339 }, { "epoch": 0.15037039767073307, "grad_norm": 0.34126553774695134, "learning_rate": 4.992636229749632e-06, "loss": 0.3815, "step": 340 }, { "epoch": 0.1508126635462352, "grad_norm": 0.31253693597060633, "learning_rate": 5.007363770250369e-06, "loss": 0.3805, "step": 341 }, { "epoch": 0.15125492942173738, "grad_norm": 0.34004357856385103, "learning_rate": 5.0220913107511045e-06, "loss": 0.3922, "step": 342 }, { "epoch": 0.15169719529723952, "grad_norm": 0.3077334730507332, "learning_rate": 5.036818851251841e-06, "loss": 0.3463, "step": 343 }, { "epoch": 0.15213946117274169, "grad_norm": 0.3832513882855968, "learning_rate": 5.051546391752578e-06, "loss": 0.3732, "step": 344 }, { "epoch": 0.15258172704824383, "grad_norm": 0.30560371575540207, "learning_rate": 5.066273932253314e-06, "loss": 0.3624, "step": 345 }, { "epoch": 0.153023992923746, "grad_norm": 0.3149387158251376, "learning_rate": 5.08100147275405e-06, "loss": 0.3678, "step": 346 }, { "epoch": 0.15346625879924813, "grad_norm": 0.3420601463207998, "learning_rate": 5.095729013254786e-06, "loss": 0.4129, "step": 347 }, { "epoch": 0.1539085246747503, "grad_norm": 0.33769195284251485, "learning_rate": 5.110456553755523e-06, "loss": 0.3932, "step": 348 }, { "epoch": 0.15435079055025247, "grad_norm": 0.3104062337106549, "learning_rate": 5.125184094256259e-06, "loss": 0.3602, "step": 349 }, { "epoch": 0.1547930564257546, "grad_norm": 0.35056993885799737, "learning_rate": 5.139911634756995e-06, "loss": 0.4361, "step": 350 }, { "epoch": 0.15523532230125678, "grad_norm": 0.30429181119618626, "learning_rate": 5.154639175257732e-06, "loss": 0.3913, "step": 351 }, { "epoch": 0.15567758817675892, "grad_norm": 0.32804226270949527, "learning_rate": 5.169366715758469e-06, "loss": 0.381, "step": 352 }, { "epoch": 0.1561198540522611, "grad_norm": 0.3101301833895385, "learning_rate": 5.184094256259205e-06, "loss": 0.4064, "step": 353 }, { "epoch": 0.15656211992776323, "grad_norm": 0.37585147077433734, "learning_rate": 5.198821796759942e-06, "loss": 0.3813, "step": 354 }, { "epoch": 0.1570043858032654, "grad_norm": 0.36863372402955863, "learning_rate": 5.2135493372606785e-06, "loss": 0.3929, "step": 355 }, { "epoch": 0.15744665167876756, "grad_norm": 0.40716149468922846, "learning_rate": 5.228276877761414e-06, "loss": 0.3791, "step": 356 }, { "epoch": 0.1578889175542697, "grad_norm": 0.3270477495009545, "learning_rate": 5.243004418262151e-06, "loss": 0.438, "step": 357 }, { "epoch": 0.15833118342977187, "grad_norm": 0.3214393138217844, "learning_rate": 5.257731958762888e-06, "loss": 0.3633, "step": 358 }, { "epoch": 0.158773449305274, "grad_norm": 0.43412386081352616, "learning_rate": 5.272459499263623e-06, "loss": 0.3773, "step": 359 }, { "epoch": 0.15921571518077618, "grad_norm": 0.3322175390153572, "learning_rate": 5.28718703976436e-06, "loss": 0.3821, "step": 360 }, { "epoch": 0.15965798105627832, "grad_norm": 0.3548796460963499, "learning_rate": 5.301914580265096e-06, "loss": 0.4029, "step": 361 }, { "epoch": 0.1601002469317805, "grad_norm": 0.44872493144407644, "learning_rate": 5.3166421207658325e-06, "loss": 0.4161, "step": 362 }, { "epoch": 0.16054251280728266, "grad_norm": 0.3263563927372097, "learning_rate": 5.331369661266569e-06, "loss": 0.3782, "step": 363 }, { "epoch": 0.1609847786827848, "grad_norm": 0.398597009288506, "learning_rate": 5.346097201767305e-06, "loss": 0.3711, "step": 364 }, { "epoch": 0.16142704455828696, "grad_norm": 0.30884993738844085, "learning_rate": 5.360824742268042e-06, "loss": 0.404, "step": 365 }, { "epoch": 0.1618693104337891, "grad_norm": 0.3040869491073254, "learning_rate": 5.3755522827687774e-06, "loss": 0.37, "step": 366 }, { "epoch": 0.16231157630929127, "grad_norm": 0.3109093650826845, "learning_rate": 5.390279823269514e-06, "loss": 0.4106, "step": 367 }, { "epoch": 0.1627538421847934, "grad_norm": 0.3973079548334136, "learning_rate": 5.405007363770251e-06, "loss": 0.4312, "step": 368 }, { "epoch": 0.16319610806029558, "grad_norm": 0.3432816817299342, "learning_rate": 5.4197349042709865e-06, "loss": 0.4008, "step": 369 }, { "epoch": 0.16363837393579775, "grad_norm": 0.3708435252265001, "learning_rate": 5.434462444771723e-06, "loss": 0.3628, "step": 370 }, { "epoch": 0.1640806398112999, "grad_norm": 0.31765255588734925, "learning_rate": 5.44918998527246e-06, "loss": 0.3844, "step": 371 }, { "epoch": 0.16452290568680206, "grad_norm": 0.31250439169273275, "learning_rate": 5.463917525773196e-06, "loss": 0.3575, "step": 372 }, { "epoch": 0.1649651715623042, "grad_norm": 0.35631942588047266, "learning_rate": 5.478645066273933e-06, "loss": 0.41, "step": 373 }, { "epoch": 0.16540743743780636, "grad_norm": 0.4384413299339308, "learning_rate": 5.49337260677467e-06, "loss": 0.4233, "step": 374 }, { "epoch": 0.1658497033133085, "grad_norm": 0.30311431510273373, "learning_rate": 5.508100147275406e-06, "loss": 0.4029, "step": 375 }, { "epoch": 0.16629196918881067, "grad_norm": 0.3861890401878624, "learning_rate": 5.522827687776142e-06, "loss": 0.381, "step": 376 }, { "epoch": 0.16673423506431284, "grad_norm": 0.3359459978208452, "learning_rate": 5.537555228276879e-06, "loss": 0.3978, "step": 377 }, { "epoch": 0.16717650093981498, "grad_norm": 0.3628216878388851, "learning_rate": 5.552282768777615e-06, "loss": 0.3863, "step": 378 }, { "epoch": 0.16761876681531715, "grad_norm": 0.3324924044012746, "learning_rate": 5.567010309278351e-06, "loss": 0.3639, "step": 379 }, { "epoch": 0.1680610326908193, "grad_norm": 0.31661187130506785, "learning_rate": 5.581737849779087e-06, "loss": 0.3867, "step": 380 }, { "epoch": 0.16850329856632146, "grad_norm": 0.32414694038071895, "learning_rate": 5.596465390279824e-06, "loss": 0.3813, "step": 381 }, { "epoch": 0.1689455644418236, "grad_norm": 0.4017483359650772, "learning_rate": 5.6111929307805605e-06, "loss": 0.3844, "step": 382 }, { "epoch": 0.16938783031732577, "grad_norm": 0.3254782091808309, "learning_rate": 5.625920471281296e-06, "loss": 0.4108, "step": 383 }, { "epoch": 0.16983009619282793, "grad_norm": 0.3273233595163591, "learning_rate": 5.640648011782033e-06, "loss": 0.4149, "step": 384 }, { "epoch": 0.17027236206833007, "grad_norm": 0.3800847124454333, "learning_rate": 5.655375552282769e-06, "loss": 0.361, "step": 385 }, { "epoch": 0.17071462794383224, "grad_norm": 0.31930596968696834, "learning_rate": 5.670103092783505e-06, "loss": 0.3763, "step": 386 }, { "epoch": 0.17115689381933438, "grad_norm": 0.320923350955301, "learning_rate": 5.684830633284242e-06, "loss": 0.4009, "step": 387 }, { "epoch": 0.17159915969483655, "grad_norm": 0.33544111851008, "learning_rate": 5.699558173784978e-06, "loss": 0.3831, "step": 388 }, { "epoch": 0.1720414255703387, "grad_norm": 0.3004229886168698, "learning_rate": 5.7142857142857145e-06, "loss": 0.3706, "step": 389 }, { "epoch": 0.17248369144584086, "grad_norm": 0.33667085699276134, "learning_rate": 5.729013254786451e-06, "loss": 0.3971, "step": 390 }, { "epoch": 0.17292595732134303, "grad_norm": 0.3200735676074629, "learning_rate": 5.743740795287187e-06, "loss": 0.3677, "step": 391 }, { "epoch": 0.17336822319684517, "grad_norm": 0.302078344951202, "learning_rate": 5.758468335787924e-06, "loss": 0.3553, "step": 392 }, { "epoch": 0.17381048907234734, "grad_norm": 0.35168586166105936, "learning_rate": 5.7731958762886594e-06, "loss": 0.3984, "step": 393 }, { "epoch": 0.17425275494784948, "grad_norm": 0.379024189319952, "learning_rate": 5.787923416789397e-06, "loss": 0.3994, "step": 394 }, { "epoch": 0.17469502082335164, "grad_norm": 0.33359927959826513, "learning_rate": 5.802650957290134e-06, "loss": 0.3679, "step": 395 }, { "epoch": 0.17513728669885378, "grad_norm": 0.32583143138244974, "learning_rate": 5.81737849779087e-06, "loss": 0.3973, "step": 396 }, { "epoch": 0.17557955257435595, "grad_norm": 0.33203354736256646, "learning_rate": 5.832106038291606e-06, "loss": 0.3775, "step": 397 }, { "epoch": 0.17602181844985812, "grad_norm": 0.34821268611815365, "learning_rate": 5.846833578792343e-06, "loss": 0.3759, "step": 398 }, { "epoch": 0.17646408432536026, "grad_norm": 0.44894556565633464, "learning_rate": 5.8615611192930785e-06, "loss": 0.3975, "step": 399 }, { "epoch": 0.17690635020086243, "grad_norm": 0.34975683063203844, "learning_rate": 5.876288659793815e-06, "loss": 0.3874, "step": 400 }, { "epoch": 0.17734861607636457, "grad_norm": 0.34944081408240724, "learning_rate": 5.891016200294552e-06, "loss": 0.3857, "step": 401 }, { "epoch": 0.17779088195186674, "grad_norm": 0.39142208429290587, "learning_rate": 5.905743740795288e-06, "loss": 0.4073, "step": 402 }, { "epoch": 0.17823314782736888, "grad_norm": 0.3422803383096213, "learning_rate": 5.920471281296024e-06, "loss": 0.409, "step": 403 }, { "epoch": 0.17867541370287104, "grad_norm": 0.37896706254029494, "learning_rate": 5.93519882179676e-06, "loss": 0.4068, "step": 404 }, { "epoch": 0.1791176795783732, "grad_norm": 0.36414387786619473, "learning_rate": 5.949926362297497e-06, "loss": 0.4064, "step": 405 }, { "epoch": 0.17955994545387535, "grad_norm": 0.3402866648286493, "learning_rate": 5.964653902798233e-06, "loss": 0.4021, "step": 406 }, { "epoch": 0.18000221132937752, "grad_norm": 0.44243364644491745, "learning_rate": 5.979381443298969e-06, "loss": 0.4111, "step": 407 }, { "epoch": 0.18044447720487966, "grad_norm": 0.3700875698304009, "learning_rate": 5.994108983799706e-06, "loss": 0.4104, "step": 408 }, { "epoch": 0.18088674308038183, "grad_norm": 0.37028321436262757, "learning_rate": 6.0088365243004425e-06, "loss": 0.3903, "step": 409 }, { "epoch": 0.18132900895588397, "grad_norm": 0.3643053169097023, "learning_rate": 6.023564064801178e-06, "loss": 0.3716, "step": 410 }, { "epoch": 0.18177127483138614, "grad_norm": 0.36377625496753757, "learning_rate": 6.038291605301915e-06, "loss": 0.3526, "step": 411 }, { "epoch": 0.18221354070688828, "grad_norm": 0.35867719411424726, "learning_rate": 6.053019145802651e-06, "loss": 0.3908, "step": 412 }, { "epoch": 0.18265580658239045, "grad_norm": 0.3453300990924209, "learning_rate": 6.067746686303387e-06, "loss": 0.3399, "step": 413 }, { "epoch": 0.1830980724578926, "grad_norm": 0.41783003964044396, "learning_rate": 6.082474226804124e-06, "loss": 0.437, "step": 414 }, { "epoch": 0.18354033833339475, "grad_norm": 0.33591820764943736, "learning_rate": 6.0972017673048616e-06, "loss": 0.3801, "step": 415 }, { "epoch": 0.18398260420889692, "grad_norm": 0.4077502361528675, "learning_rate": 6.111929307805597e-06, "loss": 0.3709, "step": 416 }, { "epoch": 0.18442487008439906, "grad_norm": 0.3163347709839168, "learning_rate": 6.126656848306334e-06, "loss": 0.355, "step": 417 }, { "epoch": 0.18486713595990123, "grad_norm": 0.3580156740704523, "learning_rate": 6.14138438880707e-06, "loss": 0.3625, "step": 418 }, { "epoch": 0.18530940183540337, "grad_norm": 0.4041192489567811, "learning_rate": 6.1561119293078065e-06, "loss": 0.3968, "step": 419 }, { "epoch": 0.18575166771090554, "grad_norm": 0.32484546131824166, "learning_rate": 6.170839469808543e-06, "loss": 0.3772, "step": 420 }, { "epoch": 0.1861939335864077, "grad_norm": 0.3579007083174091, "learning_rate": 6.185567010309279e-06, "loss": 0.4034, "step": 421 }, { "epoch": 0.18663619946190985, "grad_norm": 0.32449363004363513, "learning_rate": 6.200294550810016e-06, "loss": 0.3457, "step": 422 }, { "epoch": 0.18707846533741201, "grad_norm": 0.31568451096674205, "learning_rate": 6.215022091310751e-06, "loss": 0.3926, "step": 423 }, { "epoch": 0.18752073121291415, "grad_norm": 0.3324026881882254, "learning_rate": 6.229749631811488e-06, "loss": 0.3934, "step": 424 }, { "epoch": 0.18796299708841632, "grad_norm": 0.3071965926262146, "learning_rate": 6.244477172312225e-06, "loss": 0.34, "step": 425 }, { "epoch": 0.18840526296391846, "grad_norm": 0.37478049782475054, "learning_rate": 6.2592047128129605e-06, "loss": 0.3835, "step": 426 }, { "epoch": 0.18884752883942063, "grad_norm": 0.32395435626035946, "learning_rate": 6.273932253313697e-06, "loss": 0.4027, "step": 427 }, { "epoch": 0.1892897947149228, "grad_norm": 0.42461458727762497, "learning_rate": 6.288659793814433e-06, "loss": 0.4388, "step": 428 }, { "epoch": 0.18973206059042494, "grad_norm": 0.32878819057369596, "learning_rate": 6.30338733431517e-06, "loss": 0.3974, "step": 429 }, { "epoch": 0.1901743264659271, "grad_norm": 0.3343995915905678, "learning_rate": 6.318114874815906e-06, "loss": 0.401, "step": 430 }, { "epoch": 0.19061659234142925, "grad_norm": 0.3188093347259377, "learning_rate": 6.332842415316642e-06, "loss": 0.3431, "step": 431 }, { "epoch": 0.19105885821693142, "grad_norm": 0.3244782394634724, "learning_rate": 6.347569955817379e-06, "loss": 0.3743, "step": 432 }, { "epoch": 0.19150112409243356, "grad_norm": 0.31414934115426973, "learning_rate": 6.362297496318115e-06, "loss": 0.3855, "step": 433 }, { "epoch": 0.19194338996793572, "grad_norm": 0.32074033237774624, "learning_rate": 6.377025036818851e-06, "loss": 0.4313, "step": 434 }, { "epoch": 0.1923856558434379, "grad_norm": 0.3573373812634971, "learning_rate": 6.391752577319588e-06, "loss": 0.4161, "step": 435 }, { "epoch": 0.19282792171894003, "grad_norm": 0.3870374408224551, "learning_rate": 6.406480117820325e-06, "loss": 0.4372, "step": 436 }, { "epoch": 0.1932701875944422, "grad_norm": 0.30011201064519283, "learning_rate": 6.421207658321061e-06, "loss": 0.3392, "step": 437 }, { "epoch": 0.19371245346994434, "grad_norm": 0.29095375446833305, "learning_rate": 6.435935198821798e-06, "loss": 0.3862, "step": 438 }, { "epoch": 0.1941547193454465, "grad_norm": 0.39035454083242505, "learning_rate": 6.4506627393225344e-06, "loss": 0.4213, "step": 439 }, { "epoch": 0.19459698522094865, "grad_norm": 0.6244913820188169, "learning_rate": 6.46539027982327e-06, "loss": 0.3958, "step": 440 }, { "epoch": 0.19503925109645082, "grad_norm": 0.3338882485091999, "learning_rate": 6.480117820324007e-06, "loss": 0.3862, "step": 441 }, { "epoch": 0.19548151697195298, "grad_norm": 0.39010940389368637, "learning_rate": 6.494845360824743e-06, "loss": 0.429, "step": 442 }, { "epoch": 0.19592378284745512, "grad_norm": 0.37584587992461077, "learning_rate": 6.509572901325479e-06, "loss": 0.4096, "step": 443 }, { "epoch": 0.1963660487229573, "grad_norm": 0.3484894734191946, "learning_rate": 6.524300441826216e-06, "loss": 0.3717, "step": 444 }, { "epoch": 0.19680831459845943, "grad_norm": 0.44833895000809465, "learning_rate": 6.539027982326952e-06, "loss": 0.3789, "step": 445 }, { "epoch": 0.1972505804739616, "grad_norm": 0.515104916512036, "learning_rate": 6.5537555228276885e-06, "loss": 0.4795, "step": 446 }, { "epoch": 0.19769284634946374, "grad_norm": 0.36580702731940595, "learning_rate": 6.568483063328424e-06, "loss": 0.4081, "step": 447 }, { "epoch": 0.1981351122249659, "grad_norm": 0.36962430621779385, "learning_rate": 6.583210603829161e-06, "loss": 0.4519, "step": 448 }, { "epoch": 0.19857737810046808, "grad_norm": 0.590336387102248, "learning_rate": 6.597938144329898e-06, "loss": 0.3627, "step": 449 }, { "epoch": 0.19901964397597022, "grad_norm": 0.3642345246236517, "learning_rate": 6.612665684830633e-06, "loss": 0.3564, "step": 450 }, { "epoch": 0.19946190985147239, "grad_norm": 0.35119750269551486, "learning_rate": 6.62739322533137e-06, "loss": 0.3912, "step": 451 }, { "epoch": 0.19990417572697453, "grad_norm": 0.29069916673307306, "learning_rate": 6.642120765832107e-06, "loss": 0.3877, "step": 452 }, { "epoch": 0.2003464416024767, "grad_norm": 0.28948299088853024, "learning_rate": 6.6568483063328425e-06, "loss": 0.3742, "step": 453 }, { "epoch": 0.20078870747797883, "grad_norm": 0.36648642009705545, "learning_rate": 6.671575846833579e-06, "loss": 0.3796, "step": 454 }, { "epoch": 0.201230973353481, "grad_norm": 0.35445363308076605, "learning_rate": 6.686303387334315e-06, "loss": 0.4018, "step": 455 }, { "epoch": 0.20167323922898317, "grad_norm": 0.4344650019218321, "learning_rate": 6.701030927835052e-06, "loss": 0.4355, "step": 456 }, { "epoch": 0.2021155051044853, "grad_norm": 0.3882883193704053, "learning_rate": 6.715758468335788e-06, "loss": 0.4368, "step": 457 }, { "epoch": 0.20255777097998748, "grad_norm": 0.29834837964940364, "learning_rate": 6.730486008836526e-06, "loss": 0.3993, "step": 458 }, { "epoch": 0.20300003685548962, "grad_norm": 0.3600585483265971, "learning_rate": 6.7452135493372616e-06, "loss": 0.3883, "step": 459 }, { "epoch": 0.2034423027309918, "grad_norm": 0.3199408964234363, "learning_rate": 6.759941089837998e-06, "loss": 0.3722, "step": 460 }, { "epoch": 0.20388456860649393, "grad_norm": 0.37275783494681153, "learning_rate": 6.774668630338734e-06, "loss": 0.365, "step": 461 }, { "epoch": 0.2043268344819961, "grad_norm": 0.32024930638944854, "learning_rate": 6.789396170839471e-06, "loss": 0.4053, "step": 462 }, { "epoch": 0.20476910035749826, "grad_norm": 0.4078809342019072, "learning_rate": 6.804123711340207e-06, "loss": 0.4104, "step": 463 }, { "epoch": 0.2052113662330004, "grad_norm": 0.3591169884493893, "learning_rate": 6.818851251840943e-06, "loss": 0.3717, "step": 464 }, { "epoch": 0.20565363210850257, "grad_norm": 0.40594072552009464, "learning_rate": 6.83357879234168e-06, "loss": 0.4263, "step": 465 }, { "epoch": 0.2060958979840047, "grad_norm": 0.319030404015539, "learning_rate": 6.848306332842416e-06, "loss": 0.3583, "step": 466 }, { "epoch": 0.20653816385950688, "grad_norm": 0.32500915485356086, "learning_rate": 6.863033873343152e-06, "loss": 0.3989, "step": 467 }, { "epoch": 0.20698042973500902, "grad_norm": 0.31973434377671134, "learning_rate": 6.877761413843889e-06, "loss": 0.3786, "step": 468 }, { "epoch": 0.2074226956105112, "grad_norm": 0.35995407745014874, "learning_rate": 6.892488954344625e-06, "loss": 0.3939, "step": 469 }, { "epoch": 0.20786496148601336, "grad_norm": 0.30505449490429964, "learning_rate": 6.907216494845361e-06, "loss": 0.3756, "step": 470 }, { "epoch": 0.2083072273615155, "grad_norm": 0.30505892437639415, "learning_rate": 6.921944035346098e-06, "loss": 0.3849, "step": 471 }, { "epoch": 0.20874949323701766, "grad_norm": 0.4706104491022883, "learning_rate": 6.936671575846834e-06, "loss": 0.4239, "step": 472 }, { "epoch": 0.2091917591125198, "grad_norm": 0.3107739269607226, "learning_rate": 6.9513991163475705e-06, "loss": 0.3645, "step": 473 }, { "epoch": 0.20963402498802197, "grad_norm": 0.489800375487845, "learning_rate": 6.966126656848306e-06, "loss": 0.3917, "step": 474 }, { "epoch": 0.2100762908635241, "grad_norm": 0.2984024651567133, "learning_rate": 6.980854197349043e-06, "loss": 0.3845, "step": 475 }, { "epoch": 0.21051855673902628, "grad_norm": 0.3571788265946514, "learning_rate": 6.99558173784978e-06, "loss": 0.361, "step": 476 }, { "epoch": 0.21096082261452842, "grad_norm": 0.38879329405614954, "learning_rate": 7.010309278350515e-06, "loss": 0.3758, "step": 477 }, { "epoch": 0.2114030884900306, "grad_norm": 0.3241163636633769, "learning_rate": 7.025036818851252e-06, "loss": 0.3814, "step": 478 }, { "epoch": 0.21184535436553276, "grad_norm": 0.4101592303606582, "learning_rate": 7.0397643593519895e-06, "loss": 0.3902, "step": 479 }, { "epoch": 0.2122876202410349, "grad_norm": 0.3987450807507568, "learning_rate": 7.054491899852725e-06, "loss": 0.3444, "step": 480 }, { "epoch": 0.21272988611653706, "grad_norm": 0.33614177928537825, "learning_rate": 7.069219440353462e-06, "loss": 0.3961, "step": 481 }, { "epoch": 0.2131721519920392, "grad_norm": 0.2860254762468595, "learning_rate": 7.083946980854199e-06, "loss": 0.382, "step": 482 }, { "epoch": 0.21361441786754137, "grad_norm": 0.37552868121632077, "learning_rate": 7.0986745213549345e-06, "loss": 0.4139, "step": 483 }, { "epoch": 0.2140566837430435, "grad_norm": 0.28559802207718454, "learning_rate": 7.113402061855671e-06, "loss": 0.3678, "step": 484 }, { "epoch": 0.21449894961854568, "grad_norm": 0.3143575302015516, "learning_rate": 7.128129602356407e-06, "loss": 0.3975, "step": 485 }, { "epoch": 0.21494121549404785, "grad_norm": 0.33053355031140047, "learning_rate": 7.1428571428571436e-06, "loss": 0.3992, "step": 486 }, { "epoch": 0.21538348136955, "grad_norm": 0.8487302476383153, "learning_rate": 7.15758468335788e-06, "loss": 0.4041, "step": 487 }, { "epoch": 0.21582574724505216, "grad_norm": 0.42279649468466984, "learning_rate": 7.172312223858616e-06, "loss": 0.4139, "step": 488 }, { "epoch": 0.2162680131205543, "grad_norm": 0.31171579543743627, "learning_rate": 7.187039764359353e-06, "loss": 0.3747, "step": 489 }, { "epoch": 0.21671027899605647, "grad_norm": 0.3620420879730953, "learning_rate": 7.201767304860089e-06, "loss": 0.39, "step": 490 }, { "epoch": 0.2171525448715586, "grad_norm": 0.3286488737963793, "learning_rate": 7.216494845360825e-06, "loss": 0.345, "step": 491 }, { "epoch": 0.21759481074706077, "grad_norm": 0.33062156114185437, "learning_rate": 7.231222385861562e-06, "loss": 0.3971, "step": 492 }, { "epoch": 0.21803707662256294, "grad_norm": 0.3877074758071456, "learning_rate": 7.245949926362298e-06, "loss": 0.3939, "step": 493 }, { "epoch": 0.21847934249806508, "grad_norm": 0.34007616089676357, "learning_rate": 7.260677466863034e-06, "loss": 0.3358, "step": 494 }, { "epoch": 0.21892160837356725, "grad_norm": 0.35884519927501196, "learning_rate": 7.275405007363771e-06, "loss": 0.3745, "step": 495 }, { "epoch": 0.2193638742490694, "grad_norm": 0.39685730583303386, "learning_rate": 7.290132547864507e-06, "loss": 0.4379, "step": 496 }, { "epoch": 0.21980614012457156, "grad_norm": 0.32321166235394805, "learning_rate": 7.304860088365243e-06, "loss": 0.3727, "step": 497 }, { "epoch": 0.2202484060000737, "grad_norm": 0.31071119012341386, "learning_rate": 7.319587628865979e-06, "loss": 0.383, "step": 498 }, { "epoch": 0.22069067187557587, "grad_norm": 0.41746134613109837, "learning_rate": 7.334315169366716e-06, "loss": 0.4094, "step": 499 }, { "epoch": 0.22113293775107803, "grad_norm": 0.4542508545119592, "learning_rate": 7.349042709867453e-06, "loss": 0.4062, "step": 500 }, { "epoch": 0.22157520362658017, "grad_norm": 0.3995675348810038, "learning_rate": 7.36377025036819e-06, "loss": 0.3977, "step": 501 }, { "epoch": 0.22201746950208234, "grad_norm": 0.38676548671423366, "learning_rate": 7.378497790868926e-06, "loss": 0.428, "step": 502 }, { "epoch": 0.22245973537758448, "grad_norm": 0.32948518802279647, "learning_rate": 7.393225331369662e-06, "loss": 0.3797, "step": 503 }, { "epoch": 0.22290200125308665, "grad_norm": 0.37974612046701056, "learning_rate": 7.407952871870398e-06, "loss": 0.4105, "step": 504 }, { "epoch": 0.2233442671285888, "grad_norm": 0.365496016998349, "learning_rate": 7.422680412371135e-06, "loss": 0.3824, "step": 505 }, { "epoch": 0.22378653300409096, "grad_norm": 0.33651172695385084, "learning_rate": 7.4374079528718715e-06, "loss": 0.3951, "step": 506 }, { "epoch": 0.22422879887959313, "grad_norm": 0.4008706627447371, "learning_rate": 7.452135493372607e-06, "loss": 0.3616, "step": 507 }, { "epoch": 0.22467106475509527, "grad_norm": 0.33113363831338033, "learning_rate": 7.466863033873344e-06, "loss": 0.4003, "step": 508 }, { "epoch": 0.22511333063059744, "grad_norm": 0.30916364155377024, "learning_rate": 7.48159057437408e-06, "loss": 0.4094, "step": 509 }, { "epoch": 0.22555559650609958, "grad_norm": 0.3625630762924631, "learning_rate": 7.4963181148748165e-06, "loss": 0.4269, "step": 510 }, { "epoch": 0.22599786238160174, "grad_norm": 0.36020228804410265, "learning_rate": 7.511045655375553e-06, "loss": 0.3962, "step": 511 }, { "epoch": 0.22644012825710388, "grad_norm": 0.36552141622036305, "learning_rate": 7.525773195876289e-06, "loss": 0.4177, "step": 512 }, { "epoch": 0.22688239413260605, "grad_norm": 0.3623022952030413, "learning_rate": 7.5405007363770256e-06, "loss": 0.4231, "step": 513 }, { "epoch": 0.22732466000810822, "grad_norm": 0.3291005336774654, "learning_rate": 7.555228276877762e-06, "loss": 0.3902, "step": 514 }, { "epoch": 0.22776692588361036, "grad_norm": 0.30911097402235105, "learning_rate": 7.569955817378498e-06, "loss": 0.4045, "step": 515 }, { "epoch": 0.22820919175911253, "grad_norm": 0.32061010960014946, "learning_rate": 7.584683357879235e-06, "loss": 0.3708, "step": 516 }, { "epoch": 0.22865145763461467, "grad_norm": 0.32658893283722684, "learning_rate": 7.5994108983799705e-06, "loss": 0.3883, "step": 517 }, { "epoch": 0.22909372351011684, "grad_norm": 0.3211637499567747, "learning_rate": 7.614138438880707e-06, "loss": 0.392, "step": 518 }, { "epoch": 0.22953598938561898, "grad_norm": 0.3637672951552815, "learning_rate": 7.628865979381444e-06, "loss": 0.3929, "step": 519 }, { "epoch": 0.22997825526112115, "grad_norm": 0.4220330179144857, "learning_rate": 7.64359351988218e-06, "loss": 0.489, "step": 520 }, { "epoch": 0.2304205211366233, "grad_norm": 0.4770220051882468, "learning_rate": 7.658321060382917e-06, "loss": 0.4577, "step": 521 }, { "epoch": 0.23086278701212545, "grad_norm": 0.3980216422741699, "learning_rate": 7.673048600883654e-06, "loss": 0.3946, "step": 522 }, { "epoch": 0.23130505288762762, "grad_norm": 0.38392586290141173, "learning_rate": 7.68777614138439e-06, "loss": 0.397, "step": 523 }, { "epoch": 0.23174731876312976, "grad_norm": 0.32956159125469264, "learning_rate": 7.702503681885127e-06, "loss": 0.3582, "step": 524 }, { "epoch": 0.23218958463863193, "grad_norm": 0.3692166638248333, "learning_rate": 7.717231222385862e-06, "loss": 0.4174, "step": 525 }, { "epoch": 0.23263185051413407, "grad_norm": 0.3224035204862082, "learning_rate": 7.731958762886599e-06, "loss": 0.4153, "step": 526 }, { "epoch": 0.23307411638963624, "grad_norm": 0.40258529048093855, "learning_rate": 7.746686303387335e-06, "loss": 0.4007, "step": 527 }, { "epoch": 0.2335163822651384, "grad_norm": 0.35194952780260663, "learning_rate": 7.761413843888072e-06, "loss": 0.3994, "step": 528 }, { "epoch": 0.23395864814064055, "grad_norm": 0.310914805941736, "learning_rate": 7.776141384388809e-06, "loss": 0.4039, "step": 529 }, { "epoch": 0.23440091401614271, "grad_norm": 0.35424445962978857, "learning_rate": 7.790868924889544e-06, "loss": 0.4018, "step": 530 }, { "epoch": 0.23484317989164485, "grad_norm": 0.3257987718156192, "learning_rate": 7.80559646539028e-06, "loss": 0.3743, "step": 531 }, { "epoch": 0.23528544576714702, "grad_norm": 0.3334402292708255, "learning_rate": 7.820324005891017e-06, "loss": 0.3665, "step": 532 }, { "epoch": 0.23572771164264916, "grad_norm": 0.3175387274321486, "learning_rate": 7.835051546391754e-06, "loss": 0.3952, "step": 533 }, { "epoch": 0.23616997751815133, "grad_norm": 0.3838459070987746, "learning_rate": 7.84977908689249e-06, "loss": 0.3877, "step": 534 }, { "epoch": 0.2366122433936535, "grad_norm": 0.33875520270837867, "learning_rate": 7.864506627393225e-06, "loss": 0.4193, "step": 535 }, { "epoch": 0.23705450926915564, "grad_norm": 0.39581347984766296, "learning_rate": 7.879234167893962e-06, "loss": 0.407, "step": 536 }, { "epoch": 0.2374967751446578, "grad_norm": 0.5547704957925942, "learning_rate": 7.893961708394698e-06, "loss": 0.4522, "step": 537 }, { "epoch": 0.23793904102015995, "grad_norm": 0.3521921131116238, "learning_rate": 7.908689248895435e-06, "loss": 0.3769, "step": 538 }, { "epoch": 0.23838130689566212, "grad_norm": 0.35920892965614254, "learning_rate": 7.923416789396172e-06, "loss": 0.4091, "step": 539 }, { "epoch": 0.23882357277116426, "grad_norm": 0.34859437001242094, "learning_rate": 7.938144329896907e-06, "loss": 0.3943, "step": 540 }, { "epoch": 0.23926583864666642, "grad_norm": 0.31101393353998147, "learning_rate": 7.952871870397643e-06, "loss": 0.3554, "step": 541 }, { "epoch": 0.23970810452216856, "grad_norm": 0.3756085078536371, "learning_rate": 7.96759941089838e-06, "loss": 0.349, "step": 542 }, { "epoch": 0.24015037039767073, "grad_norm": 0.44323324438364164, "learning_rate": 7.982326951399117e-06, "loss": 0.4681, "step": 543 }, { "epoch": 0.2405926362731729, "grad_norm": 0.36155916501779545, "learning_rate": 7.997054491899853e-06, "loss": 0.4156, "step": 544 }, { "epoch": 0.24103490214867504, "grad_norm": 0.36118138630990654, "learning_rate": 8.01178203240059e-06, "loss": 0.3729, "step": 545 }, { "epoch": 0.2414771680241772, "grad_norm": 0.3714254869496448, "learning_rate": 8.026509572901327e-06, "loss": 0.3963, "step": 546 }, { "epoch": 0.24191943389967935, "grad_norm": 0.33232408521248913, "learning_rate": 8.041237113402063e-06, "loss": 0.391, "step": 547 }, { "epoch": 0.24236169977518152, "grad_norm": 0.268929056925799, "learning_rate": 8.0559646539028e-06, "loss": 0.372, "step": 548 }, { "epoch": 0.24280396565068366, "grad_norm": 0.3362067025111874, "learning_rate": 8.070692194403535e-06, "loss": 0.3838, "step": 549 }, { "epoch": 0.24324623152618582, "grad_norm": 0.34153814159278567, "learning_rate": 8.085419734904272e-06, "loss": 0.4148, "step": 550 }, { "epoch": 0.243688497401688, "grad_norm": 0.30463388411969633, "learning_rate": 8.100147275405008e-06, "loss": 0.3603, "step": 551 }, { "epoch": 0.24413076327719013, "grad_norm": 0.3329945926890606, "learning_rate": 8.114874815905745e-06, "loss": 0.3917, "step": 552 }, { "epoch": 0.2445730291526923, "grad_norm": 0.36933174221865966, "learning_rate": 8.129602356406482e-06, "loss": 0.4242, "step": 553 }, { "epoch": 0.24501529502819444, "grad_norm": 0.3456035959956852, "learning_rate": 8.144329896907216e-06, "loss": 0.3816, "step": 554 }, { "epoch": 0.2454575609036966, "grad_norm": 0.322634557502108, "learning_rate": 8.159057437407953e-06, "loss": 0.389, "step": 555 }, { "epoch": 0.24589982677919875, "grad_norm": 0.3183891103192985, "learning_rate": 8.17378497790869e-06, "loss": 0.3771, "step": 556 }, { "epoch": 0.24634209265470092, "grad_norm": 0.3131720536893327, "learning_rate": 8.188512518409426e-06, "loss": 0.4376, "step": 557 }, { "epoch": 0.24678435853020309, "grad_norm": 0.3362210346435818, "learning_rate": 8.203240058910163e-06, "loss": 0.3876, "step": 558 }, { "epoch": 0.24722662440570523, "grad_norm": 0.29429419311605387, "learning_rate": 8.217967599410898e-06, "loss": 0.395, "step": 559 }, { "epoch": 0.2476688902812074, "grad_norm": 0.28818627467936175, "learning_rate": 8.232695139911635e-06, "loss": 0.3604, "step": 560 }, { "epoch": 0.24811115615670953, "grad_norm": 0.3612794413365004, "learning_rate": 8.247422680412371e-06, "loss": 0.4015, "step": 561 }, { "epoch": 0.2485534220322117, "grad_norm": 0.325364799491205, "learning_rate": 8.262150220913108e-06, "loss": 0.3512, "step": 562 }, { "epoch": 0.24899568790771384, "grad_norm": 0.3532004992279409, "learning_rate": 8.276877761413845e-06, "loss": 0.395, "step": 563 }, { "epoch": 0.249437953783216, "grad_norm": 0.32537977286267405, "learning_rate": 8.291605301914581e-06, "loss": 0.3982, "step": 564 }, { "epoch": 0.24988021965871818, "grad_norm": 0.3694511403685403, "learning_rate": 8.306332842415318e-06, "loss": 0.3959, "step": 565 }, { "epoch": 0.2503224855342203, "grad_norm": 0.3014250848794977, "learning_rate": 8.321060382916055e-06, "loss": 0.3845, "step": 566 }, { "epoch": 0.2507647514097225, "grad_norm": 0.518179774455641, "learning_rate": 8.335787923416791e-06, "loss": 0.3868, "step": 567 }, { "epoch": 0.25120701728522465, "grad_norm": 0.37130149462997236, "learning_rate": 8.350515463917526e-06, "loss": 0.4, "step": 568 }, { "epoch": 0.25164928316072677, "grad_norm": 0.33634882242390063, "learning_rate": 8.365243004418263e-06, "loss": 0.4306, "step": 569 }, { "epoch": 0.25209154903622893, "grad_norm": 0.3801330419423097, "learning_rate": 8.379970544919e-06, "loss": 0.3999, "step": 570 }, { "epoch": 0.2525338149117311, "grad_norm": 0.4039441593000365, "learning_rate": 8.394698085419736e-06, "loss": 0.4189, "step": 571 }, { "epoch": 0.25297608078723327, "grad_norm": 0.3322545942088769, "learning_rate": 8.409425625920473e-06, "loss": 0.3663, "step": 572 }, { "epoch": 0.25341834666273544, "grad_norm": 0.36012048314498424, "learning_rate": 8.424153166421208e-06, "loss": 0.4087, "step": 573 }, { "epoch": 0.25386061253823755, "grad_norm": 0.30936269915247183, "learning_rate": 8.438880706921944e-06, "loss": 0.3846, "step": 574 }, { "epoch": 0.2543028784137397, "grad_norm": 0.3413243721433314, "learning_rate": 8.453608247422681e-06, "loss": 0.4199, "step": 575 }, { "epoch": 0.2547451442892419, "grad_norm": 0.3302170626738759, "learning_rate": 8.468335787923418e-06, "loss": 0.3677, "step": 576 }, { "epoch": 0.25518741016474406, "grad_norm": 0.3147730466234293, "learning_rate": 8.483063328424154e-06, "loss": 0.3825, "step": 577 }, { "epoch": 0.25562967604024617, "grad_norm": 0.35635635498607937, "learning_rate": 8.49779086892489e-06, "loss": 0.3779, "step": 578 }, { "epoch": 0.25607194191574834, "grad_norm": 0.4297026167828229, "learning_rate": 8.512518409425626e-06, "loss": 0.4384, "step": 579 }, { "epoch": 0.2565142077912505, "grad_norm": 0.35109340169952885, "learning_rate": 8.527245949926363e-06, "loss": 0.4041, "step": 580 }, { "epoch": 0.25695647366675267, "grad_norm": 0.3157088249539107, "learning_rate": 8.5419734904271e-06, "loss": 0.4076, "step": 581 }, { "epoch": 0.25739873954225484, "grad_norm": 0.3849833201364007, "learning_rate": 8.556701030927836e-06, "loss": 0.3606, "step": 582 }, { "epoch": 0.25784100541775695, "grad_norm": 0.3647744350210299, "learning_rate": 8.571428571428571e-06, "loss": 0.3765, "step": 583 }, { "epoch": 0.2582832712932591, "grad_norm": 0.3169102161618409, "learning_rate": 8.586156111929308e-06, "loss": 0.3692, "step": 584 }, { "epoch": 0.2587255371687613, "grad_norm": 0.8858333809754846, "learning_rate": 8.600883652430046e-06, "loss": 0.3862, "step": 585 }, { "epoch": 0.25916780304426346, "grad_norm": 0.3251494812973131, "learning_rate": 8.615611192930783e-06, "loss": 0.3987, "step": 586 }, { "epoch": 0.2596100689197656, "grad_norm": 0.3620380248437641, "learning_rate": 8.630338733431518e-06, "loss": 0.3914, "step": 587 }, { "epoch": 0.26005233479526774, "grad_norm": 0.37454460475905793, "learning_rate": 8.645066273932254e-06, "loss": 0.4446, "step": 588 }, { "epoch": 0.2604946006707699, "grad_norm": 0.35995888056238956, "learning_rate": 8.65979381443299e-06, "loss": 0.3891, "step": 589 }, { "epoch": 0.2609368665462721, "grad_norm": 0.32180681054320504, "learning_rate": 8.674521354933727e-06, "loss": 0.4018, "step": 590 }, { "epoch": 0.26137913242177424, "grad_norm": 0.4218624966732303, "learning_rate": 8.689248895434464e-06, "loss": 0.3975, "step": 591 }, { "epoch": 0.26182139829727635, "grad_norm": 0.4126660947564939, "learning_rate": 8.703976435935199e-06, "loss": 0.3766, "step": 592 }, { "epoch": 0.2622636641727785, "grad_norm": 0.2894818771236959, "learning_rate": 8.718703976435936e-06, "loss": 0.3911, "step": 593 }, { "epoch": 0.2627059300482807, "grad_norm": 0.31840077400567496, "learning_rate": 8.733431516936672e-06, "loss": 0.3818, "step": 594 }, { "epoch": 0.26314819592378286, "grad_norm": 0.36403941370953247, "learning_rate": 8.748159057437409e-06, "loss": 0.4203, "step": 595 }, { "epoch": 0.263590461799285, "grad_norm": 0.31773748156616655, "learning_rate": 8.762886597938146e-06, "loss": 0.3908, "step": 596 }, { "epoch": 0.26403272767478714, "grad_norm": 0.4181070183440365, "learning_rate": 8.77761413843888e-06, "loss": 0.4416, "step": 597 }, { "epoch": 0.2644749935502893, "grad_norm": 0.3498212378271477, "learning_rate": 8.792341678939617e-06, "loss": 0.3881, "step": 598 }, { "epoch": 0.2649172594257915, "grad_norm": 0.3339621766255387, "learning_rate": 8.807069219440354e-06, "loss": 0.3594, "step": 599 }, { "epoch": 0.26535952530129364, "grad_norm": 0.35395755975219373, "learning_rate": 8.82179675994109e-06, "loss": 0.3883, "step": 600 }, { "epoch": 0.2658017911767958, "grad_norm": 0.34500710283586955, "learning_rate": 8.836524300441827e-06, "loss": 0.3702, "step": 601 }, { "epoch": 0.2662440570522979, "grad_norm": 0.3223104140035005, "learning_rate": 8.851251840942562e-06, "loss": 0.3907, "step": 602 }, { "epoch": 0.2666863229278001, "grad_norm": 0.3214737518391533, "learning_rate": 8.865979381443299e-06, "loss": 0.3555, "step": 603 }, { "epoch": 0.26712858880330226, "grad_norm": 0.3056593475434351, "learning_rate": 8.880706921944036e-06, "loss": 0.3921, "step": 604 }, { "epoch": 0.2675708546788044, "grad_norm": 0.39278357786093127, "learning_rate": 8.895434462444772e-06, "loss": 0.3828, "step": 605 }, { "epoch": 0.26801312055430654, "grad_norm": 0.3648014827866533, "learning_rate": 8.910162002945509e-06, "loss": 0.386, "step": 606 }, { "epoch": 0.2684553864298087, "grad_norm": 0.30846595197373233, "learning_rate": 8.924889543446246e-06, "loss": 0.367, "step": 607 }, { "epoch": 0.2688976523053109, "grad_norm": 0.3241044239996195, "learning_rate": 8.939617083946982e-06, "loss": 0.3715, "step": 608 }, { "epoch": 0.26933991818081304, "grad_norm": 0.3570721642739329, "learning_rate": 8.954344624447719e-06, "loss": 0.3695, "step": 609 }, { "epoch": 0.2697821840563152, "grad_norm": 0.3102579557091555, "learning_rate": 8.969072164948455e-06, "loss": 0.4014, "step": 610 }, { "epoch": 0.2702244499318173, "grad_norm": 0.4804881757982265, "learning_rate": 8.98379970544919e-06, "loss": 0.4044, "step": 611 }, { "epoch": 0.2706667158073195, "grad_norm": 0.37598108802402763, "learning_rate": 8.998527245949927e-06, "loss": 0.4092, "step": 612 }, { "epoch": 0.27110898168282166, "grad_norm": 0.35113989541372687, "learning_rate": 9.013254786450664e-06, "loss": 0.3798, "step": 613 }, { "epoch": 0.2715512475583238, "grad_norm": 0.3060062721012405, "learning_rate": 9.0279823269514e-06, "loss": 0.388, "step": 614 }, { "epoch": 0.271993513433826, "grad_norm": 0.3711299406863348, "learning_rate": 9.042709867452137e-06, "loss": 0.4251, "step": 615 }, { "epoch": 0.2724357793093281, "grad_norm": 0.3054900175572718, "learning_rate": 9.057437407952872e-06, "loss": 0.3748, "step": 616 }, { "epoch": 0.2728780451848303, "grad_norm": 0.3399978014811487, "learning_rate": 9.072164948453609e-06, "loss": 0.3738, "step": 617 }, { "epoch": 0.27332031106033244, "grad_norm": 0.45787946375059463, "learning_rate": 9.086892488954345e-06, "loss": 0.4354, "step": 618 }, { "epoch": 0.2737625769358346, "grad_norm": 0.33090534541876226, "learning_rate": 9.101620029455082e-06, "loss": 0.3971, "step": 619 }, { "epoch": 0.2742048428113367, "grad_norm": 0.3330247341959805, "learning_rate": 9.116347569955819e-06, "loss": 0.3962, "step": 620 }, { "epoch": 0.2746471086868389, "grad_norm": 0.3234680616943663, "learning_rate": 9.131075110456554e-06, "loss": 0.3806, "step": 621 }, { "epoch": 0.27508937456234106, "grad_norm": 0.30588230711045017, "learning_rate": 9.14580265095729e-06, "loss": 0.3898, "step": 622 }, { "epoch": 0.27553164043784323, "grad_norm": 0.30664449768800334, "learning_rate": 9.160530191458027e-06, "loss": 0.391, "step": 623 }, { "epoch": 0.2759739063133454, "grad_norm": 0.3801877680105698, "learning_rate": 9.175257731958764e-06, "loss": 0.4062, "step": 624 }, { "epoch": 0.2764161721888475, "grad_norm": 0.29328065991277097, "learning_rate": 9.1899852724595e-06, "loss": 0.3644, "step": 625 }, { "epoch": 0.2768584380643497, "grad_norm": 0.3430435640397146, "learning_rate": 9.204712812960235e-06, "loss": 0.3697, "step": 626 }, { "epoch": 0.27730070393985184, "grad_norm": 0.3470632838463265, "learning_rate": 9.219440353460973e-06, "loss": 0.3896, "step": 627 }, { "epoch": 0.277742969815354, "grad_norm": 0.3606643958030464, "learning_rate": 9.23416789396171e-06, "loss": 0.3895, "step": 628 }, { "epoch": 0.2781852356908562, "grad_norm": 0.37385868279905665, "learning_rate": 9.248895434462447e-06, "loss": 0.409, "step": 629 }, { "epoch": 0.2786275015663583, "grad_norm": 0.3906067509549741, "learning_rate": 9.263622974963182e-06, "loss": 0.3915, "step": 630 }, { "epoch": 0.27906976744186046, "grad_norm": 0.30408617095808227, "learning_rate": 9.278350515463918e-06, "loss": 0.3676, "step": 631 }, { "epoch": 0.27951203331736263, "grad_norm": 0.3524201400322172, "learning_rate": 9.293078055964655e-06, "loss": 0.415, "step": 632 }, { "epoch": 0.2799542991928648, "grad_norm": 0.4273178269088461, "learning_rate": 9.307805596465392e-06, "loss": 0.3935, "step": 633 }, { "epoch": 0.2803965650683669, "grad_norm": 0.31866556337983254, "learning_rate": 9.322533136966128e-06, "loss": 0.3408, "step": 634 }, { "epoch": 0.2808388309438691, "grad_norm": 0.3503807929590369, "learning_rate": 9.337260677466863e-06, "loss": 0.3607, "step": 635 }, { "epoch": 0.28128109681937125, "grad_norm": 0.4050301462178413, "learning_rate": 9.3519882179676e-06, "loss": 0.4343, "step": 636 }, { "epoch": 0.2817233626948734, "grad_norm": 0.34681620710660255, "learning_rate": 9.366715758468337e-06, "loss": 0.4278, "step": 637 }, { "epoch": 0.2821656285703756, "grad_norm": 0.31843476089460004, "learning_rate": 9.381443298969073e-06, "loss": 0.3767, "step": 638 }, { "epoch": 0.2826078944458777, "grad_norm": 0.3434571424312571, "learning_rate": 9.39617083946981e-06, "loss": 0.3591, "step": 639 }, { "epoch": 0.28305016032137986, "grad_norm": 0.3008969951860365, "learning_rate": 9.410898379970545e-06, "loss": 0.3538, "step": 640 }, { "epoch": 0.28349242619688203, "grad_norm": 0.398109843126312, "learning_rate": 9.425625920471282e-06, "loss": 0.4081, "step": 641 }, { "epoch": 0.2839346920723842, "grad_norm": 0.3087192400370694, "learning_rate": 9.440353460972018e-06, "loss": 0.3702, "step": 642 }, { "epoch": 0.2843769579478863, "grad_norm": 0.32528916433111105, "learning_rate": 9.455081001472755e-06, "loss": 0.4185, "step": 643 }, { "epoch": 0.2848192238233885, "grad_norm": 0.3207151529451909, "learning_rate": 9.469808541973491e-06, "loss": 0.4051, "step": 644 }, { "epoch": 0.28526148969889065, "grad_norm": 0.44677304999816253, "learning_rate": 9.484536082474226e-06, "loss": 0.4093, "step": 645 }, { "epoch": 0.2857037555743928, "grad_norm": 0.33280865606910404, "learning_rate": 9.499263622974963e-06, "loss": 0.397, "step": 646 }, { "epoch": 0.286146021449895, "grad_norm": 0.3985246095328315, "learning_rate": 9.5139911634757e-06, "loss": 0.4103, "step": 647 }, { "epoch": 0.2865882873253971, "grad_norm": 0.3540089860963544, "learning_rate": 9.528718703976436e-06, "loss": 0.4042, "step": 648 }, { "epoch": 0.28703055320089926, "grad_norm": 0.36345693037755583, "learning_rate": 9.543446244477173e-06, "loss": 0.4, "step": 649 }, { "epoch": 0.28747281907640143, "grad_norm": 0.350823625972751, "learning_rate": 9.55817378497791e-06, "loss": 0.4079, "step": 650 }, { "epoch": 0.2879150849519036, "grad_norm": 0.3267966982586813, "learning_rate": 9.572901325478646e-06, "loss": 0.3849, "step": 651 }, { "epoch": 0.28835735082740577, "grad_norm": 0.44034993347282336, "learning_rate": 9.587628865979383e-06, "loss": 0.4347, "step": 652 }, { "epoch": 0.2887996167029079, "grad_norm": 0.32886778400122274, "learning_rate": 9.60235640648012e-06, "loss": 0.4062, "step": 653 }, { "epoch": 0.28924188257841005, "grad_norm": 0.30444782439407736, "learning_rate": 9.617083946980855e-06, "loss": 0.3589, "step": 654 }, { "epoch": 0.2896841484539122, "grad_norm": 0.297718836421567, "learning_rate": 9.631811487481591e-06, "loss": 0.3839, "step": 655 }, { "epoch": 0.2901264143294144, "grad_norm": 0.3396097992646443, "learning_rate": 9.646539027982328e-06, "loss": 0.4105, "step": 656 }, { "epoch": 0.2905686802049165, "grad_norm": 0.3012804057062133, "learning_rate": 9.661266568483065e-06, "loss": 0.3442, "step": 657 }, { "epoch": 0.29101094608041866, "grad_norm": 0.3228061891950649, "learning_rate": 9.675994108983801e-06, "loss": 0.3646, "step": 658 }, { "epoch": 0.29145321195592083, "grad_norm": 0.3251852707871598, "learning_rate": 9.690721649484536e-06, "loss": 0.4063, "step": 659 }, { "epoch": 0.291895477831423, "grad_norm": 0.3475937251018205, "learning_rate": 9.705449189985273e-06, "loss": 0.4113, "step": 660 }, { "epoch": 0.29233774370692517, "grad_norm": 0.3513234376179405, "learning_rate": 9.72017673048601e-06, "loss": 0.367, "step": 661 }, { "epoch": 0.2927800095824273, "grad_norm": 0.31461891108670637, "learning_rate": 9.734904270986746e-06, "loss": 0.3858, "step": 662 }, { "epoch": 0.29322227545792945, "grad_norm": 0.3270392915013356, "learning_rate": 9.749631811487483e-06, "loss": 0.3835, "step": 663 }, { "epoch": 0.2936645413334316, "grad_norm": 0.340519438202178, "learning_rate": 9.764359351988218e-06, "loss": 0.3506, "step": 664 }, { "epoch": 0.2941068072089338, "grad_norm": 0.34646400859784776, "learning_rate": 9.779086892488954e-06, "loss": 0.4124, "step": 665 }, { "epoch": 0.29454907308443595, "grad_norm": 0.29292729761448527, "learning_rate": 9.793814432989691e-06, "loss": 0.3884, "step": 666 }, { "epoch": 0.29499133895993807, "grad_norm": 0.3194406721102042, "learning_rate": 9.808541973490428e-06, "loss": 0.3976, "step": 667 }, { "epoch": 0.29543360483544023, "grad_norm": 0.38303533012518975, "learning_rate": 9.823269513991164e-06, "loss": 0.4056, "step": 668 }, { "epoch": 0.2958758707109424, "grad_norm": 0.3149135932139269, "learning_rate": 9.8379970544919e-06, "loss": 0.4061, "step": 669 }, { "epoch": 0.29631813658644457, "grad_norm": 0.3541528129772728, "learning_rate": 9.852724594992638e-06, "loss": 0.419, "step": 670 }, { "epoch": 0.2967604024619467, "grad_norm": 0.327482905968914, "learning_rate": 9.867452135493374e-06, "loss": 0.408, "step": 671 }, { "epoch": 0.29720266833744885, "grad_norm": 0.3505251198680731, "learning_rate": 9.882179675994111e-06, "loss": 0.4107, "step": 672 }, { "epoch": 0.297644934212951, "grad_norm": 0.323540975863573, "learning_rate": 9.896907216494846e-06, "loss": 0.375, "step": 673 }, { "epoch": 0.2980872000884532, "grad_norm": 0.33682981096165787, "learning_rate": 9.911634756995583e-06, "loss": 0.3867, "step": 674 }, { "epoch": 0.29852946596395535, "grad_norm": 0.3069887615250735, "learning_rate": 9.92636229749632e-06, "loss": 0.4124, "step": 675 }, { "epoch": 0.29897173183945747, "grad_norm": 0.3394628670286458, "learning_rate": 9.941089837997056e-06, "loss": 0.4024, "step": 676 }, { "epoch": 0.29941399771495963, "grad_norm": 0.4515394319443443, "learning_rate": 9.955817378497793e-06, "loss": 0.435, "step": 677 }, { "epoch": 0.2998562635904618, "grad_norm": 0.3343627824765761, "learning_rate": 9.970544918998528e-06, "loss": 0.3874, "step": 678 }, { "epoch": 0.30029852946596397, "grad_norm": 0.33346693598788946, "learning_rate": 9.985272459499264e-06, "loss": 0.3857, "step": 679 }, { "epoch": 0.30074079534146614, "grad_norm": 0.33881954675073794, "learning_rate": 1e-05, "loss": 0.419, "step": 680 }, { "epoch": 0.30118306121696825, "grad_norm": 0.3433421790828746, "learning_rate": 9.99999933841766e-06, "loss": 0.3758, "step": 681 }, { "epoch": 0.3016253270924704, "grad_norm": 0.3569843645207418, "learning_rate": 9.999997353670812e-06, "loss": 0.4224, "step": 682 }, { "epoch": 0.3020675929679726, "grad_norm": 0.3261582721345569, "learning_rate": 9.999994045759982e-06, "loss": 0.4, "step": 683 }, { "epoch": 0.30250985884347475, "grad_norm": 0.5109550886525084, "learning_rate": 9.999989414686046e-06, "loss": 0.381, "step": 684 }, { "epoch": 0.30295212471897687, "grad_norm": 0.33835945941740236, "learning_rate": 9.999983460450228e-06, "loss": 0.4091, "step": 685 }, { "epoch": 0.30339439059447904, "grad_norm": 0.333878619577278, "learning_rate": 9.999976183054108e-06, "loss": 0.3775, "step": 686 }, { "epoch": 0.3038366564699812, "grad_norm": 0.2980730687467199, "learning_rate": 9.999967582499605e-06, "loss": 0.3745, "step": 687 }, { "epoch": 0.30427892234548337, "grad_norm": 0.30937531853307954, "learning_rate": 9.999957658789e-06, "loss": 0.412, "step": 688 }, { "epoch": 0.30472118822098554, "grad_norm": 0.3813206082269362, "learning_rate": 9.99994641192492e-06, "loss": 0.4568, "step": 689 }, { "epoch": 0.30516345409648765, "grad_norm": 0.30931149138367986, "learning_rate": 9.999933841910337e-06, "loss": 0.4081, "step": 690 }, { "epoch": 0.3056057199719898, "grad_norm": 0.35686289736884175, "learning_rate": 9.99991994874858e-06, "loss": 0.3942, "step": 691 }, { "epoch": 0.306047985847492, "grad_norm": 0.3441697292694062, "learning_rate": 9.999904732443326e-06, "loss": 0.3848, "step": 692 }, { "epoch": 0.30649025172299416, "grad_norm": 0.3175531372950004, "learning_rate": 9.9998881929986e-06, "loss": 0.3774, "step": 693 }, { "epoch": 0.30693251759849627, "grad_norm": 0.3221353206941767, "learning_rate": 9.99987033041878e-06, "loss": 0.4032, "step": 694 }, { "epoch": 0.30737478347399844, "grad_norm": 0.34931007291157645, "learning_rate": 9.999851144708593e-06, "loss": 0.4206, "step": 695 }, { "epoch": 0.3078170493495006, "grad_norm": 0.3049632896741952, "learning_rate": 9.999830635873116e-06, "loss": 0.3558, "step": 696 }, { "epoch": 0.3082593152250028, "grad_norm": 0.3798392131936684, "learning_rate": 9.999808803917775e-06, "loss": 0.4352, "step": 697 }, { "epoch": 0.30870158110050494, "grad_norm": 0.30553382337358437, "learning_rate": 9.99978564884835e-06, "loss": 0.367, "step": 698 }, { "epoch": 0.30914384697600705, "grad_norm": 0.3546160106158521, "learning_rate": 9.999761170670968e-06, "loss": 0.4007, "step": 699 }, { "epoch": 0.3095861128515092, "grad_norm": 0.3538316547965842, "learning_rate": 9.999735369392104e-06, "loss": 0.3984, "step": 700 }, { "epoch": 0.3100283787270114, "grad_norm": 0.30882120312155176, "learning_rate": 9.999708245018588e-06, "loss": 0.3916, "step": 701 }, { "epoch": 0.31047064460251356, "grad_norm": 0.34011948606276254, "learning_rate": 9.9996797975576e-06, "loss": 0.4356, "step": 702 }, { "epoch": 0.3109129104780157, "grad_norm": 0.38011166029779914, "learning_rate": 9.999650027016665e-06, "loss": 0.4224, "step": 703 }, { "epoch": 0.31135517635351784, "grad_norm": 0.3540726199381365, "learning_rate": 9.999618933403663e-06, "loss": 0.4125, "step": 704 }, { "epoch": 0.31179744222902, "grad_norm": 0.35402573752213407, "learning_rate": 9.999586516726822e-06, "loss": 0.3887, "step": 705 }, { "epoch": 0.3122397081045222, "grad_norm": 0.34570027332939446, "learning_rate": 9.999552776994719e-06, "loss": 0.3748, "step": 706 }, { "epoch": 0.31268197398002434, "grad_norm": 0.35582364501556674, "learning_rate": 9.999517714216285e-06, "loss": 0.4091, "step": 707 }, { "epoch": 0.31312423985552645, "grad_norm": 0.28511104429826495, "learning_rate": 9.999481328400797e-06, "loss": 0.4023, "step": 708 }, { "epoch": 0.3135665057310286, "grad_norm": 0.28341865017385987, "learning_rate": 9.999443619557884e-06, "loss": 0.3599, "step": 709 }, { "epoch": 0.3140087716065308, "grad_norm": 0.34259843763842585, "learning_rate": 9.999404587697526e-06, "loss": 0.4102, "step": 710 }, { "epoch": 0.31445103748203296, "grad_norm": 0.3252714509842128, "learning_rate": 9.999364232830053e-06, "loss": 0.4096, "step": 711 }, { "epoch": 0.3148933033575351, "grad_norm": 0.329465724748232, "learning_rate": 9.99932255496614e-06, "loss": 0.4019, "step": 712 }, { "epoch": 0.31533556923303724, "grad_norm": 0.39770904010770963, "learning_rate": 9.999279554116821e-06, "loss": 0.3971, "step": 713 }, { "epoch": 0.3157778351085394, "grad_norm": 0.3748178618416407, "learning_rate": 9.999235230293474e-06, "loss": 0.4218, "step": 714 }, { "epoch": 0.3162201009840416, "grad_norm": 0.3222086517995307, "learning_rate": 9.999189583507828e-06, "loss": 0.3726, "step": 715 }, { "epoch": 0.31666236685954374, "grad_norm": 0.3227178129135171, "learning_rate": 9.999142613771965e-06, "loss": 0.3772, "step": 716 }, { "epoch": 0.3171046327350459, "grad_norm": 0.4101372484784259, "learning_rate": 9.999094321098309e-06, "loss": 0.428, "step": 717 }, { "epoch": 0.317546898610548, "grad_norm": 0.37101865877851187, "learning_rate": 9.999044705499645e-06, "loss": 0.3866, "step": 718 }, { "epoch": 0.3179891644860502, "grad_norm": 0.30109896544776454, "learning_rate": 9.9989937669891e-06, "loss": 0.3823, "step": 719 }, { "epoch": 0.31843143036155236, "grad_norm": 0.32804298094677203, "learning_rate": 9.998941505580157e-06, "loss": 0.4227, "step": 720 }, { "epoch": 0.3188736962370545, "grad_norm": 0.3521046021371402, "learning_rate": 9.998887921286643e-06, "loss": 0.404, "step": 721 }, { "epoch": 0.31931596211255664, "grad_norm": 0.36289031615419404, "learning_rate": 9.998833014122741e-06, "loss": 0.3926, "step": 722 }, { "epoch": 0.3197582279880588, "grad_norm": 0.40888438565977336, "learning_rate": 9.998776784102981e-06, "loss": 0.3618, "step": 723 }, { "epoch": 0.320200493863561, "grad_norm": 0.3288746595982331, "learning_rate": 9.998719231242241e-06, "loss": 0.3972, "step": 724 }, { "epoch": 0.32064275973906314, "grad_norm": 0.41267254070231607, "learning_rate": 9.998660355555751e-06, "loss": 0.4181, "step": 725 }, { "epoch": 0.3210850256145653, "grad_norm": 0.28935692360748877, "learning_rate": 9.998600157059095e-06, "loss": 0.4101, "step": 726 }, { "epoch": 0.3215272914900674, "grad_norm": 1.1996519258287452, "learning_rate": 9.998538635768202e-06, "loss": 0.4011, "step": 727 }, { "epoch": 0.3219695573655696, "grad_norm": 0.33060196262219865, "learning_rate": 9.99847579169935e-06, "loss": 0.3986, "step": 728 }, { "epoch": 0.32241182324107176, "grad_norm": 0.7004003259457516, "learning_rate": 9.998411624869172e-06, "loss": 0.4122, "step": 729 }, { "epoch": 0.32285408911657393, "grad_norm": 0.3389288640357919, "learning_rate": 9.99834613529465e-06, "loss": 0.3739, "step": 730 }, { "epoch": 0.3232963549920761, "grad_norm": 0.31998775370463634, "learning_rate": 9.998279322993112e-06, "loss": 0.4136, "step": 731 }, { "epoch": 0.3237386208675782, "grad_norm": 0.3450530106068192, "learning_rate": 9.99821118798224e-06, "loss": 0.4244, "step": 732 }, { "epoch": 0.3241808867430804, "grad_norm": 0.34407669030140414, "learning_rate": 9.998141730280064e-06, "loss": 0.377, "step": 733 }, { "epoch": 0.32462315261858254, "grad_norm": 0.37847513127757315, "learning_rate": 9.998070949904966e-06, "loss": 0.4033, "step": 734 }, { "epoch": 0.3250654184940847, "grad_norm": 0.34482517910421095, "learning_rate": 9.997998846875676e-06, "loss": 0.3822, "step": 735 }, { "epoch": 0.3255076843695868, "grad_norm": 0.32035556169507706, "learning_rate": 9.997925421211275e-06, "loss": 0.3731, "step": 736 }, { "epoch": 0.325949950245089, "grad_norm": 0.3228492114797286, "learning_rate": 9.997850672931193e-06, "loss": 0.4153, "step": 737 }, { "epoch": 0.32639221612059116, "grad_norm": 0.33412425521584355, "learning_rate": 9.997774602055213e-06, "loss": 0.386, "step": 738 }, { "epoch": 0.32683448199609333, "grad_norm": 0.330548649665821, "learning_rate": 9.997697208603464e-06, "loss": 0.3913, "step": 739 }, { "epoch": 0.3272767478715955, "grad_norm": 0.2942797742647829, "learning_rate": 9.997618492596425e-06, "loss": 0.352, "step": 740 }, { "epoch": 0.3277190137470976, "grad_norm": 0.34285872394761235, "learning_rate": 9.997538454054933e-06, "loss": 0.4211, "step": 741 }, { "epoch": 0.3281612796225998, "grad_norm": 0.34600161425421133, "learning_rate": 9.997457093000165e-06, "loss": 0.4085, "step": 742 }, { "epoch": 0.32860354549810195, "grad_norm": 0.3905506793651762, "learning_rate": 9.997374409453649e-06, "loss": 0.4359, "step": 743 }, { "epoch": 0.3290458113736041, "grad_norm": 0.3349389030533602, "learning_rate": 9.99729040343727e-06, "loss": 0.3644, "step": 744 }, { "epoch": 0.3294880772491063, "grad_norm": 0.34988537240993356, "learning_rate": 9.99720507497326e-06, "loss": 0.3992, "step": 745 }, { "epoch": 0.3299303431246084, "grad_norm": 0.3796869098190786, "learning_rate": 9.997118424084193e-06, "loss": 0.3924, "step": 746 }, { "epoch": 0.33037260900011056, "grad_norm": 0.31341694117406355, "learning_rate": 9.997030450793008e-06, "loss": 0.3543, "step": 747 }, { "epoch": 0.33081487487561273, "grad_norm": 0.36996831434219835, "learning_rate": 9.99694115512298e-06, "loss": 0.3941, "step": 748 }, { "epoch": 0.3312571407511149, "grad_norm": 0.3294148500467849, "learning_rate": 9.996850537097738e-06, "loss": 0.3604, "step": 749 }, { "epoch": 0.331699406626617, "grad_norm": 0.3959362563487315, "learning_rate": 9.99675859674127e-06, "loss": 0.4028, "step": 750 }, { "epoch": 0.3321416725021192, "grad_norm": 0.33263711008775326, "learning_rate": 9.9966653340779e-06, "loss": 0.3546, "step": 751 }, { "epoch": 0.33258393837762135, "grad_norm": 0.3562469574172677, "learning_rate": 9.99657074913231e-06, "loss": 0.4143, "step": 752 }, { "epoch": 0.3330262042531235, "grad_norm": 0.43308232694453, "learning_rate": 9.99647484192953e-06, "loss": 0.4357, "step": 753 }, { "epoch": 0.3334684701286257, "grad_norm": 0.4344662442218764, "learning_rate": 9.996377612494944e-06, "loss": 0.4249, "step": 754 }, { "epoch": 0.3339107360041278, "grad_norm": 0.3688787700045291, "learning_rate": 9.996279060854276e-06, "loss": 0.4031, "step": 755 }, { "epoch": 0.33435300187962996, "grad_norm": 0.33180965263113155, "learning_rate": 9.99617918703361e-06, "loss": 0.3564, "step": 756 }, { "epoch": 0.33479526775513213, "grad_norm": 0.37216517354039447, "learning_rate": 9.996077991059375e-06, "loss": 0.3956, "step": 757 }, { "epoch": 0.3352375336306343, "grad_norm": 0.3279990213429732, "learning_rate": 9.99597547295835e-06, "loss": 0.3999, "step": 758 }, { "epoch": 0.3356797995061364, "grad_norm": 0.649816537717823, "learning_rate": 9.995871632757668e-06, "loss": 0.4128, "step": 759 }, { "epoch": 0.3361220653816386, "grad_norm": 0.3442810120432562, "learning_rate": 9.995766470484802e-06, "loss": 0.3951, "step": 760 }, { "epoch": 0.33656433125714075, "grad_norm": 0.3684456910775568, "learning_rate": 9.995659986167588e-06, "loss": 0.4105, "step": 761 }, { "epoch": 0.3370065971326429, "grad_norm": 0.3736403997099007, "learning_rate": 9.995552179834202e-06, "loss": 0.4081, "step": 762 }, { "epoch": 0.3374488630081451, "grad_norm": 0.34016445670316175, "learning_rate": 9.995443051513175e-06, "loss": 0.3544, "step": 763 }, { "epoch": 0.3378911288836472, "grad_norm": 1.5857703107866403, "learning_rate": 9.995332601233383e-06, "loss": 0.4364, "step": 764 }, { "epoch": 0.33833339475914936, "grad_norm": 0.3503435258177327, "learning_rate": 9.995220829024056e-06, "loss": 0.3834, "step": 765 }, { "epoch": 0.33877566063465153, "grad_norm": 0.32673664139690095, "learning_rate": 9.995107734914773e-06, "loss": 0.3743, "step": 766 }, { "epoch": 0.3392179265101537, "grad_norm": 0.3744434702949884, "learning_rate": 9.994993318935465e-06, "loss": 0.4047, "step": 767 }, { "epoch": 0.33966019238565587, "grad_norm": 0.32105161623364775, "learning_rate": 9.994877581116405e-06, "loss": 0.407, "step": 768 }, { "epoch": 0.340102458261158, "grad_norm": 0.38330714910498176, "learning_rate": 9.994760521488225e-06, "loss": 0.3911, "step": 769 }, { "epoch": 0.34054472413666015, "grad_norm": 0.39374555795394955, "learning_rate": 9.994642140081901e-06, "loss": 0.4096, "step": 770 }, { "epoch": 0.3409869900121623, "grad_norm": 0.3212504014497009, "learning_rate": 9.99452243692876e-06, "loss": 0.3786, "step": 771 }, { "epoch": 0.3414292558876645, "grad_norm": 0.34499873993048086, "learning_rate": 9.994401412060483e-06, "loss": 0.4506, "step": 772 }, { "epoch": 0.3418715217631666, "grad_norm": 0.28954021674954533, "learning_rate": 9.994279065509094e-06, "loss": 0.3867, "step": 773 }, { "epoch": 0.34231378763866877, "grad_norm": 0.30934177067160223, "learning_rate": 9.99415539730697e-06, "loss": 0.3946, "step": 774 }, { "epoch": 0.34275605351417093, "grad_norm": 0.3189132173267416, "learning_rate": 9.99403040748684e-06, "loss": 0.3638, "step": 775 }, { "epoch": 0.3431983193896731, "grad_norm": 0.3258916283622027, "learning_rate": 9.993904096081775e-06, "loss": 0.382, "step": 776 }, { "epoch": 0.34364058526517527, "grad_norm": 0.3383402974205202, "learning_rate": 9.993776463125208e-06, "loss": 0.3875, "step": 777 }, { "epoch": 0.3440828511406774, "grad_norm": 0.3235275768482762, "learning_rate": 9.99364750865091e-06, "loss": 0.3842, "step": 778 }, { "epoch": 0.34452511701617955, "grad_norm": 0.33299176308605927, "learning_rate": 9.99351723269301e-06, "loss": 0.3723, "step": 779 }, { "epoch": 0.3449673828916817, "grad_norm": 0.33790101443591464, "learning_rate": 9.993385635285982e-06, "loss": 0.3796, "step": 780 }, { "epoch": 0.3454096487671839, "grad_norm": 0.31307786137909743, "learning_rate": 9.99325271646465e-06, "loss": 0.356, "step": 781 }, { "epoch": 0.34585191464268605, "grad_norm": 0.3795221617686431, "learning_rate": 9.99311847626419e-06, "loss": 0.3795, "step": 782 }, { "epoch": 0.34629418051818817, "grad_norm": 0.34411499115875666, "learning_rate": 9.992982914720124e-06, "loss": 0.4089, "step": 783 }, { "epoch": 0.34673644639369033, "grad_norm": 0.31696553937719635, "learning_rate": 9.99284603186833e-06, "loss": 0.4273, "step": 784 }, { "epoch": 0.3471787122691925, "grad_norm": 0.31651008145536363, "learning_rate": 9.992707827745029e-06, "loss": 0.4205, "step": 785 }, { "epoch": 0.34762097814469467, "grad_norm": 0.332676316429931, "learning_rate": 9.992568302386794e-06, "loss": 0.4094, "step": 786 }, { "epoch": 0.3480632440201968, "grad_norm": 0.36233385185330436, "learning_rate": 9.992427455830548e-06, "loss": 0.3705, "step": 787 }, { "epoch": 0.34850550989569895, "grad_norm": 0.29839750621659156, "learning_rate": 9.992285288113568e-06, "loss": 0.3511, "step": 788 }, { "epoch": 0.3489477757712011, "grad_norm": 0.3227347015221889, "learning_rate": 9.99214179927347e-06, "loss": 0.3836, "step": 789 }, { "epoch": 0.3493900416467033, "grad_norm": 0.30995044151924955, "learning_rate": 9.991996989348229e-06, "loss": 0.3873, "step": 790 }, { "epoch": 0.34983230752220545, "grad_norm": 0.2894770429962307, "learning_rate": 9.991850858376167e-06, "loss": 0.3378, "step": 791 }, { "epoch": 0.35027457339770757, "grad_norm": 0.31020068845351717, "learning_rate": 9.991703406395953e-06, "loss": 0.3958, "step": 792 }, { "epoch": 0.35071683927320974, "grad_norm": 0.33869034620900884, "learning_rate": 9.991554633446612e-06, "loss": 0.4144, "step": 793 }, { "epoch": 0.3511591051487119, "grad_norm": 0.3096972320347006, "learning_rate": 9.991404539567509e-06, "loss": 0.3877, "step": 794 }, { "epoch": 0.35160137102421407, "grad_norm": 0.3432015335739856, "learning_rate": 9.991253124798365e-06, "loss": 0.4044, "step": 795 }, { "epoch": 0.35204363689971624, "grad_norm": 0.3762776215503244, "learning_rate": 9.991100389179252e-06, "loss": 0.4276, "step": 796 }, { "epoch": 0.35248590277521835, "grad_norm": 0.31815529374221585, "learning_rate": 9.990946332750587e-06, "loss": 0.375, "step": 797 }, { "epoch": 0.3529281686507205, "grad_norm": 0.3615081944163721, "learning_rate": 9.990790955553137e-06, "loss": 0.3985, "step": 798 }, { "epoch": 0.3533704345262227, "grad_norm": 0.31468474090149534, "learning_rate": 9.990634257628022e-06, "loss": 0.3437, "step": 799 }, { "epoch": 0.35381270040172486, "grad_norm": 0.3122366088504658, "learning_rate": 9.990476239016709e-06, "loss": 0.3755, "step": 800 }, { "epoch": 0.35425496627722697, "grad_norm": 0.3346854970943513, "learning_rate": 9.990316899761016e-06, "loss": 0.3812, "step": 801 }, { "epoch": 0.35469723215272914, "grad_norm": 0.3360368026204711, "learning_rate": 9.990156239903106e-06, "loss": 0.411, "step": 802 }, { "epoch": 0.3551394980282313, "grad_norm": 0.28674521783439205, "learning_rate": 9.989994259485499e-06, "loss": 0.395, "step": 803 }, { "epoch": 0.35558176390373347, "grad_norm": 0.324371481083809, "learning_rate": 9.989830958551058e-06, "loss": 0.3806, "step": 804 }, { "epoch": 0.35602402977923564, "grad_norm": 0.32609555964808423, "learning_rate": 9.989666337142998e-06, "loss": 0.3807, "step": 805 }, { "epoch": 0.35646629565473775, "grad_norm": 0.3051400756515235, "learning_rate": 9.989500395304882e-06, "loss": 0.3707, "step": 806 }, { "epoch": 0.3569085615302399, "grad_norm": 0.3607392297466933, "learning_rate": 9.989333133080627e-06, "loss": 0.4523, "step": 807 }, { "epoch": 0.3573508274057421, "grad_norm": 0.3407888133416416, "learning_rate": 9.989164550514496e-06, "loss": 0.4221, "step": 808 }, { "epoch": 0.35779309328124426, "grad_norm": 0.32937807116621604, "learning_rate": 9.988994647651097e-06, "loss": 0.3873, "step": 809 }, { "epoch": 0.3582353591567464, "grad_norm": 0.3497906348705442, "learning_rate": 9.988823424535396e-06, "loss": 0.4047, "step": 810 }, { "epoch": 0.35867762503224854, "grad_norm": 0.34142123103946626, "learning_rate": 9.988650881212703e-06, "loss": 0.3743, "step": 811 }, { "epoch": 0.3591198909077507, "grad_norm": 0.29461935768688186, "learning_rate": 9.988477017728677e-06, "loss": 0.373, "step": 812 }, { "epoch": 0.3595621567832529, "grad_norm": 0.35273494848956194, "learning_rate": 9.98830183412933e-06, "loss": 0.4011, "step": 813 }, { "epoch": 0.36000442265875504, "grad_norm": 0.3243987776948259, "learning_rate": 9.988125330461023e-06, "loss": 0.3672, "step": 814 }, { "epoch": 0.36044668853425715, "grad_norm": 0.34588930017853087, "learning_rate": 9.987947506770461e-06, "loss": 0.3908, "step": 815 }, { "epoch": 0.3608889544097593, "grad_norm": 0.3415881169886628, "learning_rate": 9.987768363104703e-06, "loss": 0.3932, "step": 816 }, { "epoch": 0.3613312202852615, "grad_norm": 0.33097688434970307, "learning_rate": 9.987587899511158e-06, "loss": 0.3779, "step": 817 }, { "epoch": 0.36177348616076366, "grad_norm": 0.30770861919407005, "learning_rate": 9.987406116037581e-06, "loss": 0.3711, "step": 818 }, { "epoch": 0.3622157520362658, "grad_norm": 0.3385476959146475, "learning_rate": 9.987223012732079e-06, "loss": 0.375, "step": 819 }, { "epoch": 0.36265801791176794, "grad_norm": 0.2962525108746534, "learning_rate": 9.987038589643108e-06, "loss": 0.3784, "step": 820 }, { "epoch": 0.3631002837872701, "grad_norm": 0.3702475823738672, "learning_rate": 9.986852846819467e-06, "loss": 0.4177, "step": 821 }, { "epoch": 0.3635425496627723, "grad_norm": 0.41944379989359704, "learning_rate": 9.986665784310318e-06, "loss": 0.3896, "step": 822 }, { "epoch": 0.36398481553827444, "grad_norm": 0.28969454018317153, "learning_rate": 9.986477402165157e-06, "loss": 0.3978, "step": 823 }, { "epoch": 0.36442708141377655, "grad_norm": 0.37679663539537445, "learning_rate": 9.98628770043384e-06, "loss": 0.4068, "step": 824 }, { "epoch": 0.3648693472892787, "grad_norm": 0.35881493836695, "learning_rate": 9.986096679166569e-06, "loss": 0.4209, "step": 825 }, { "epoch": 0.3653116131647809, "grad_norm": 0.2868733983820795, "learning_rate": 9.985904338413889e-06, "loss": 0.3628, "step": 826 }, { "epoch": 0.36575387904028306, "grad_norm": 0.32030119320766104, "learning_rate": 9.985710678226705e-06, "loss": 0.3974, "step": 827 }, { "epoch": 0.3661961449157852, "grad_norm": 0.35009864255652356, "learning_rate": 9.985515698656264e-06, "loss": 0.4217, "step": 828 }, { "epoch": 0.36663841079128734, "grad_norm": 0.32453224291068117, "learning_rate": 9.985319399754163e-06, "loss": 0.3535, "step": 829 }, { "epoch": 0.3670806766667895, "grad_norm": 0.3353696129328265, "learning_rate": 9.985121781572352e-06, "loss": 0.4228, "step": 830 }, { "epoch": 0.3675229425422917, "grad_norm": 0.33154707917826465, "learning_rate": 9.984922844163125e-06, "loss": 0.3722, "step": 831 }, { "epoch": 0.36796520841779384, "grad_norm": 0.467004271244488, "learning_rate": 9.984722587579127e-06, "loss": 0.4009, "step": 832 }, { "epoch": 0.368407474293296, "grad_norm": 0.32718043268639363, "learning_rate": 9.984521011873355e-06, "loss": 0.4099, "step": 833 }, { "epoch": 0.3688497401687981, "grad_norm": 0.30143821893865014, "learning_rate": 9.98431811709915e-06, "loss": 0.3954, "step": 834 }, { "epoch": 0.3692920060443003, "grad_norm": 0.3680008181105228, "learning_rate": 9.984113903310206e-06, "loss": 0.4377, "step": 835 }, { "epoch": 0.36973427191980246, "grad_norm": 0.2895059438870427, "learning_rate": 9.983908370560565e-06, "loss": 0.3439, "step": 836 }, { "epoch": 0.3701765377953046, "grad_norm": 0.7691623164239542, "learning_rate": 9.983701518904616e-06, "loss": 0.5311, "step": 837 }, { "epoch": 0.37061880367080674, "grad_norm": 0.3667583318322003, "learning_rate": 9.9834933483971e-06, "loss": 0.4047, "step": 838 }, { "epoch": 0.3710610695463089, "grad_norm": 0.32276967962298175, "learning_rate": 9.983283859093107e-06, "loss": 0.4389, "step": 839 }, { "epoch": 0.3715033354218111, "grad_norm": 0.5081272593922255, "learning_rate": 9.983073051048072e-06, "loss": 0.3828, "step": 840 }, { "epoch": 0.37194560129731324, "grad_norm": 0.48436618191303615, "learning_rate": 9.982860924317785e-06, "loss": 0.4236, "step": 841 }, { "epoch": 0.3723878671728154, "grad_norm": 0.38216213627917556, "learning_rate": 9.982647478958378e-06, "loss": 0.4061, "step": 842 }, { "epoch": 0.3728301330483175, "grad_norm": 0.27485094383061565, "learning_rate": 9.982432715026337e-06, "loss": 0.3807, "step": 843 }, { "epoch": 0.3732723989238197, "grad_norm": 0.34483817216621077, "learning_rate": 9.982216632578499e-06, "loss": 0.4194, "step": 844 }, { "epoch": 0.37371466479932186, "grad_norm": 0.2866210841526716, "learning_rate": 9.981999231672042e-06, "loss": 0.3931, "step": 845 }, { "epoch": 0.37415693067482403, "grad_norm": 0.3588331942155727, "learning_rate": 9.981780512364499e-06, "loss": 0.4437, "step": 846 }, { "epoch": 0.3745991965503262, "grad_norm": 0.34149719428950626, "learning_rate": 9.98156047471375e-06, "loss": 0.3712, "step": 847 }, { "epoch": 0.3750414624258283, "grad_norm": 0.31937004319042955, "learning_rate": 9.981339118778026e-06, "loss": 0.3772, "step": 848 }, { "epoch": 0.3754837283013305, "grad_norm": 0.483816351833687, "learning_rate": 9.981116444615904e-06, "loss": 0.3909, "step": 849 }, { "epoch": 0.37592599417683265, "grad_norm": 0.35880631032960003, "learning_rate": 9.98089245228631e-06, "loss": 0.3784, "step": 850 }, { "epoch": 0.3763682600523348, "grad_norm": 0.3149615761208219, "learning_rate": 9.980667141848519e-06, "loss": 0.4102, "step": 851 }, { "epoch": 0.3768105259278369, "grad_norm": 0.34207246923327633, "learning_rate": 9.98044051336216e-06, "loss": 0.4022, "step": 852 }, { "epoch": 0.3772527918033391, "grad_norm": 0.35054055633276615, "learning_rate": 9.980212566887199e-06, "loss": 0.4193, "step": 853 }, { "epoch": 0.37769505767884126, "grad_norm": 0.30935792198649037, "learning_rate": 9.979983302483966e-06, "loss": 0.3974, "step": 854 }, { "epoch": 0.37813732355434343, "grad_norm": 0.3277991108825115, "learning_rate": 9.979752720213127e-06, "loss": 0.3776, "step": 855 }, { "epoch": 0.3785795894298456, "grad_norm": 0.3500715944861986, "learning_rate": 9.979520820135704e-06, "loss": 0.3865, "step": 856 }, { "epoch": 0.3790218553053477, "grad_norm": 1.023821250605801, "learning_rate": 9.979287602313064e-06, "loss": 0.3398, "step": 857 }, { "epoch": 0.3794641211808499, "grad_norm": 0.3292692987693669, "learning_rate": 9.979053066806926e-06, "loss": 0.3715, "step": 858 }, { "epoch": 0.37990638705635205, "grad_norm": 0.31498279387692574, "learning_rate": 9.978817213679353e-06, "loss": 0.4218, "step": 859 }, { "epoch": 0.3803486529318542, "grad_norm": 0.49747662754125316, "learning_rate": 9.97858004299276e-06, "loss": 0.4561, "step": 860 }, { "epoch": 0.3807909188073564, "grad_norm": 0.30349225762511833, "learning_rate": 9.978341554809912e-06, "loss": 0.3882, "step": 861 }, { "epoch": 0.3812331846828585, "grad_norm": 0.33983845003393603, "learning_rate": 9.978101749193921e-06, "loss": 0.3995, "step": 862 }, { "epoch": 0.38167545055836066, "grad_norm": 0.4246435074937266, "learning_rate": 9.977860626208246e-06, "loss": 0.4195, "step": 863 }, { "epoch": 0.38211771643386283, "grad_norm": 0.3457744895420125, "learning_rate": 9.977618185916695e-06, "loss": 0.4045, "step": 864 }, { "epoch": 0.382559982309365, "grad_norm": 0.2774782379184088, "learning_rate": 9.977374428383428e-06, "loss": 0.3761, "step": 865 }, { "epoch": 0.3830022481848671, "grad_norm": 0.40634091831897906, "learning_rate": 9.977129353672951e-06, "loss": 0.4213, "step": 866 }, { "epoch": 0.3834445140603693, "grad_norm": 0.31594795441934914, "learning_rate": 9.976882961850118e-06, "loss": 0.366, "step": 867 }, { "epoch": 0.38388677993587145, "grad_norm": 0.3109882081855986, "learning_rate": 9.976635252980131e-06, "loss": 0.3666, "step": 868 }, { "epoch": 0.3843290458113736, "grad_norm": 0.32199273604474005, "learning_rate": 9.976386227128546e-06, "loss": 0.4132, "step": 869 }, { "epoch": 0.3847713116868758, "grad_norm": 0.3505674437011217, "learning_rate": 9.97613588436126e-06, "loss": 0.3714, "step": 870 }, { "epoch": 0.3852135775623779, "grad_norm": 0.38900985163782187, "learning_rate": 9.975884224744522e-06, "loss": 0.4599, "step": 871 }, { "epoch": 0.38565584343788006, "grad_norm": 0.3065812136689132, "learning_rate": 9.975631248344932e-06, "loss": 0.3801, "step": 872 }, { "epoch": 0.38609810931338223, "grad_norm": 0.35959313955911293, "learning_rate": 9.975376955229435e-06, "loss": 0.3817, "step": 873 }, { "epoch": 0.3865403751888844, "grad_norm": 0.3643566131574229, "learning_rate": 9.975121345465321e-06, "loss": 0.4166, "step": 874 }, { "epoch": 0.38698264106438657, "grad_norm": 0.3195031157795866, "learning_rate": 9.97486441912024e-06, "loss": 0.4048, "step": 875 }, { "epoch": 0.3874249069398887, "grad_norm": 0.3194136926093858, "learning_rate": 9.974606176262178e-06, "loss": 0.37, "step": 876 }, { "epoch": 0.38786717281539085, "grad_norm": 0.32334194324414756, "learning_rate": 9.974346616959476e-06, "loss": 0.3936, "step": 877 }, { "epoch": 0.388309438690893, "grad_norm": 0.32227730130902094, "learning_rate": 9.974085741280823e-06, "loss": 0.3683, "step": 878 }, { "epoch": 0.3887517045663952, "grad_norm": 0.30628454027351526, "learning_rate": 9.973823549295254e-06, "loss": 0.4035, "step": 879 }, { "epoch": 0.3891939704418973, "grad_norm": 0.29959861168803803, "learning_rate": 9.973560041072153e-06, "loss": 0.3757, "step": 880 }, { "epoch": 0.38963623631739946, "grad_norm": 0.2971516518073857, "learning_rate": 9.973295216681256e-06, "loss": 0.3708, "step": 881 }, { "epoch": 0.39007850219290163, "grad_norm": 0.4499659987222703, "learning_rate": 9.97302907619264e-06, "loss": 0.4452, "step": 882 }, { "epoch": 0.3905207680684038, "grad_norm": 0.3025092238475454, "learning_rate": 9.972761619676738e-06, "loss": 0.3638, "step": 883 }, { "epoch": 0.39096303394390597, "grad_norm": 0.3579942145052595, "learning_rate": 9.972492847204325e-06, "loss": 0.4433, "step": 884 }, { "epoch": 0.3914052998194081, "grad_norm": 0.349991860152228, "learning_rate": 9.97222275884653e-06, "loss": 0.3969, "step": 885 }, { "epoch": 0.39184756569491025, "grad_norm": 0.46553329496697454, "learning_rate": 9.971951354674826e-06, "loss": 0.3943, "step": 886 }, { "epoch": 0.3922898315704124, "grad_norm": 0.3493670532044394, "learning_rate": 9.971678634761035e-06, "loss": 0.3684, "step": 887 }, { "epoch": 0.3927320974459146, "grad_norm": 0.32412843707997674, "learning_rate": 9.971404599177327e-06, "loss": 0.3617, "step": 888 }, { "epoch": 0.3931743633214167, "grad_norm": 0.3182767916698276, "learning_rate": 9.971129247996223e-06, "loss": 0.3791, "step": 889 }, { "epoch": 0.39361662919691887, "grad_norm": 0.4100598564264477, "learning_rate": 9.970852581290589e-06, "loss": 0.3951, "step": 890 }, { "epoch": 0.39405889507242103, "grad_norm": 0.32540641327853304, "learning_rate": 9.97057459913364e-06, "loss": 0.3547, "step": 891 }, { "epoch": 0.3945011609479232, "grad_norm": 0.40190970820937366, "learning_rate": 9.970295301598939e-06, "loss": 0.3953, "step": 892 }, { "epoch": 0.39494342682342537, "grad_norm": 0.34655368230021927, "learning_rate": 9.970014688760397e-06, "loss": 0.3833, "step": 893 }, { "epoch": 0.3953856926989275, "grad_norm": 0.2955129951590949, "learning_rate": 9.969732760692275e-06, "loss": 0.3718, "step": 894 }, { "epoch": 0.39582795857442965, "grad_norm": 0.31305047305676165, "learning_rate": 9.969449517469179e-06, "loss": 0.3911, "step": 895 }, { "epoch": 0.3962702244499318, "grad_norm": 0.36022702379392973, "learning_rate": 9.969164959166063e-06, "loss": 0.4339, "step": 896 }, { "epoch": 0.396712490325434, "grad_norm": 0.26400099511091846, "learning_rate": 9.968879085858234e-06, "loss": 0.3401, "step": 897 }, { "epoch": 0.39715475620093615, "grad_norm": 0.3511190053106137, "learning_rate": 9.968591897621343e-06, "loss": 0.3991, "step": 898 }, { "epoch": 0.39759702207643827, "grad_norm": 0.3451279981474987, "learning_rate": 9.968303394531386e-06, "loss": 0.3688, "step": 899 }, { "epoch": 0.39803928795194043, "grad_norm": 0.3094653940724691, "learning_rate": 9.968013576664715e-06, "loss": 0.3787, "step": 900 }, { "epoch": 0.3984815538274426, "grad_norm": 0.3080916555156346, "learning_rate": 9.967722444098022e-06, "loss": 0.381, "step": 901 }, { "epoch": 0.39892381970294477, "grad_norm": 0.31898553702781135, "learning_rate": 9.967429996908351e-06, "loss": 0.3945, "step": 902 }, { "epoch": 0.3993660855784469, "grad_norm": 0.3255541128249518, "learning_rate": 9.967136235173094e-06, "loss": 0.377, "step": 903 }, { "epoch": 0.39980835145394905, "grad_norm": 0.38482661881631053, "learning_rate": 9.96684115896999e-06, "loss": 0.3846, "step": 904 }, { "epoch": 0.4002506173294512, "grad_norm": 0.32410634381012254, "learning_rate": 9.966544768377124e-06, "loss": 0.3716, "step": 905 }, { "epoch": 0.4006928832049534, "grad_norm": 0.2868510015567104, "learning_rate": 9.966247063472934e-06, "loss": 0.3872, "step": 906 }, { "epoch": 0.40113514908045556, "grad_norm": 0.31515766140774715, "learning_rate": 9.965948044336198e-06, "loss": 0.3984, "step": 907 }, { "epoch": 0.40157741495595767, "grad_norm": 0.32554899646947855, "learning_rate": 9.965647711046051e-06, "loss": 0.3828, "step": 908 }, { "epoch": 0.40201968083145984, "grad_norm": 0.35184888734189024, "learning_rate": 9.965346063681968e-06, "loss": 0.4077, "step": 909 }, { "epoch": 0.402461946706962, "grad_norm": 0.36939276612420074, "learning_rate": 9.965043102323777e-06, "loss": 0.4074, "step": 910 }, { "epoch": 0.40290421258246417, "grad_norm": 0.33985072091131013, "learning_rate": 9.96473882705165e-06, "loss": 0.3925, "step": 911 }, { "epoch": 0.40334647845796634, "grad_norm": 0.3837589267760571, "learning_rate": 9.96443323794611e-06, "loss": 0.3789, "step": 912 }, { "epoch": 0.40378874433346845, "grad_norm": 0.3158333232739621, "learning_rate": 9.964126335088022e-06, "loss": 0.3815, "step": 913 }, { "epoch": 0.4042310102089706, "grad_norm": 0.3762526063581946, "learning_rate": 9.963818118558607e-06, "loss": 0.4006, "step": 914 }, { "epoch": 0.4046732760844728, "grad_norm": 0.3406514802935423, "learning_rate": 9.963508588439429e-06, "loss": 0.3842, "step": 915 }, { "epoch": 0.40511554195997496, "grad_norm": 0.3958709028557135, "learning_rate": 9.963197744812396e-06, "loss": 0.429, "step": 916 }, { "epoch": 0.40555780783547707, "grad_norm": 0.28018991584122144, "learning_rate": 9.96288558775977e-06, "loss": 0.3659, "step": 917 }, { "epoch": 0.40600007371097924, "grad_norm": 0.32108529203514713, "learning_rate": 9.962572117364157e-06, "loss": 0.384, "step": 918 }, { "epoch": 0.4064423395864814, "grad_norm": 0.3660394397045396, "learning_rate": 9.962257333708515e-06, "loss": 0.4079, "step": 919 }, { "epoch": 0.4068846054619836, "grad_norm": 0.3309301727242721, "learning_rate": 9.961941236876142e-06, "loss": 0.3937, "step": 920 }, { "epoch": 0.40732687133748574, "grad_norm": 0.29026098502561337, "learning_rate": 9.96162382695069e-06, "loss": 0.3704, "step": 921 }, { "epoch": 0.40776913721298785, "grad_norm": 0.3321590621290582, "learning_rate": 9.961305104016155e-06, "loss": 0.4035, "step": 922 }, { "epoch": 0.40821140308849, "grad_norm": 0.33168696305250195, "learning_rate": 9.960985068156881e-06, "loss": 0.3919, "step": 923 }, { "epoch": 0.4086536689639922, "grad_norm": 1.1492426378070062, "learning_rate": 9.960663719457563e-06, "loss": 0.3752, "step": 924 }, { "epoch": 0.40909593483949436, "grad_norm": 0.33133779213800635, "learning_rate": 9.960341058003236e-06, "loss": 0.372, "step": 925 }, { "epoch": 0.4095382007149965, "grad_norm": 0.31754539281654187, "learning_rate": 9.96001708387929e-06, "loss": 0.3812, "step": 926 }, { "epoch": 0.40998046659049864, "grad_norm": 0.32192298822655124, "learning_rate": 9.959691797171461e-06, "loss": 0.386, "step": 927 }, { "epoch": 0.4104227324660008, "grad_norm": 0.31743884762389124, "learning_rate": 9.959365197965824e-06, "loss": 0.3772, "step": 928 }, { "epoch": 0.410864998341503, "grad_norm": 0.3038057867919222, "learning_rate": 9.959037286348814e-06, "loss": 0.3829, "step": 929 }, { "epoch": 0.41130726421700514, "grad_norm": 0.3160594129520878, "learning_rate": 9.958708062407206e-06, "loss": 0.4052, "step": 930 }, { "epoch": 0.41174953009250725, "grad_norm": 0.3582866591513575, "learning_rate": 9.95837752622812e-06, "loss": 0.4178, "step": 931 }, { "epoch": 0.4121917959680094, "grad_norm": 0.4089135632316364, "learning_rate": 9.958045677899033e-06, "loss": 0.3889, "step": 932 }, { "epoch": 0.4126340618435116, "grad_norm": 0.31215382249020185, "learning_rate": 9.957712517507756e-06, "loss": 0.4427, "step": 933 }, { "epoch": 0.41307632771901376, "grad_norm": 0.42171304178247765, "learning_rate": 9.95737804514246e-06, "loss": 0.4021, "step": 934 }, { "epoch": 0.4135185935945159, "grad_norm": 0.38561777383207174, "learning_rate": 9.957042260891655e-06, "loss": 0.415, "step": 935 }, { "epoch": 0.41396085947001804, "grad_norm": 0.29061394261792683, "learning_rate": 9.9567051648442e-06, "loss": 0.354, "step": 936 }, { "epoch": 0.4144031253455202, "grad_norm": 0.3262187518553034, "learning_rate": 9.956366757089302e-06, "loss": 0.3852, "step": 937 }, { "epoch": 0.4148453912210224, "grad_norm": 0.3339698557850085, "learning_rate": 9.956027037716515e-06, "loss": 0.4223, "step": 938 }, { "epoch": 0.41528765709652454, "grad_norm": 0.3261160031505173, "learning_rate": 9.95568600681574e-06, "loss": 0.3624, "step": 939 }, { "epoch": 0.4157299229720267, "grad_norm": 0.3682901504578785, "learning_rate": 9.955343664477225e-06, "loss": 0.3855, "step": 940 }, { "epoch": 0.4161721888475288, "grad_norm": 0.33941542071584624, "learning_rate": 9.955000010791566e-06, "loss": 0.3943, "step": 941 }, { "epoch": 0.416614454723031, "grad_norm": 0.3560079673136296, "learning_rate": 9.954655045849705e-06, "loss": 0.3877, "step": 942 }, { "epoch": 0.41705672059853316, "grad_norm": 0.3104266212013735, "learning_rate": 9.954308769742928e-06, "loss": 0.3731, "step": 943 }, { "epoch": 0.4174989864740353, "grad_norm": 0.4035194876270946, "learning_rate": 9.953961182562876e-06, "loss": 0.3911, "step": 944 }, { "epoch": 0.41794125234953744, "grad_norm": 0.34530942628923544, "learning_rate": 9.953612284401527e-06, "loss": 0.4081, "step": 945 }, { "epoch": 0.4183835182250396, "grad_norm": 0.36796529623310237, "learning_rate": 9.953262075351215e-06, "loss": 0.4144, "step": 946 }, { "epoch": 0.4188257841005418, "grad_norm": 0.33770524785449907, "learning_rate": 9.952910555504615e-06, "loss": 0.391, "step": 947 }, { "epoch": 0.41926804997604394, "grad_norm": 0.6496509664636239, "learning_rate": 9.95255772495475e-06, "loss": 0.4262, "step": 948 }, { "epoch": 0.4197103158515461, "grad_norm": 0.3311671114652738, "learning_rate": 9.952203583794993e-06, "loss": 0.3533, "step": 949 }, { "epoch": 0.4201525817270482, "grad_norm": 0.30503279033925024, "learning_rate": 9.95184813211906e-06, "loss": 0.3563, "step": 950 }, { "epoch": 0.4205948476025504, "grad_norm": 0.38971933079718457, "learning_rate": 9.951491370021013e-06, "loss": 0.4271, "step": 951 }, { "epoch": 0.42103711347805256, "grad_norm": 0.3603431112641763, "learning_rate": 9.951133297595267e-06, "loss": 0.4245, "step": 952 }, { "epoch": 0.42147937935355473, "grad_norm": 0.3016527183279576, "learning_rate": 9.950773914936578e-06, "loss": 0.3706, "step": 953 }, { "epoch": 0.42192164522905684, "grad_norm": 0.44522882364744976, "learning_rate": 9.95041322214005e-06, "loss": 0.4199, "step": 954 }, { "epoch": 0.422363911104559, "grad_norm": 0.36049916366514617, "learning_rate": 9.950051219301134e-06, "loss": 0.4174, "step": 955 }, { "epoch": 0.4228061769800612, "grad_norm": 0.30383032226454715, "learning_rate": 9.949687906515628e-06, "loss": 0.3802, "step": 956 }, { "epoch": 0.42324844285556334, "grad_norm": 0.3526420936935153, "learning_rate": 9.949323283879678e-06, "loss": 0.3932, "step": 957 }, { "epoch": 0.4236907087310655, "grad_norm": 0.4625484740526671, "learning_rate": 9.948957351489774e-06, "loss": 0.389, "step": 958 }, { "epoch": 0.4241329746065676, "grad_norm": 0.3278200029908199, "learning_rate": 9.948590109442755e-06, "loss": 0.3919, "step": 959 }, { "epoch": 0.4245752404820698, "grad_norm": 0.3136908078695708, "learning_rate": 9.948221557835802e-06, "loss": 0.4161, "step": 960 }, { "epoch": 0.42501750635757196, "grad_norm": 0.3354053826335984, "learning_rate": 9.947851696766449e-06, "loss": 0.3559, "step": 961 }, { "epoch": 0.42545977223307413, "grad_norm": 0.33362272084981065, "learning_rate": 9.947480526332573e-06, "loss": 0.384, "step": 962 }, { "epoch": 0.4259020381085763, "grad_norm": 0.3014561715145963, "learning_rate": 9.9471080466324e-06, "loss": 0.363, "step": 963 }, { "epoch": 0.4263443039840784, "grad_norm": 0.2953045193882688, "learning_rate": 9.946734257764495e-06, "loss": 0.4167, "step": 964 }, { "epoch": 0.4267865698595806, "grad_norm": 0.33649248727816, "learning_rate": 9.946359159827776e-06, "loss": 0.3848, "step": 965 }, { "epoch": 0.42722883573508275, "grad_norm": 0.29118955213730613, "learning_rate": 9.94598275292151e-06, "loss": 0.368, "step": 966 }, { "epoch": 0.4276711016105849, "grad_norm": 0.3062131005414721, "learning_rate": 9.945605037145306e-06, "loss": 0.3693, "step": 967 }, { "epoch": 0.428113367486087, "grad_norm": 0.3332929827346834, "learning_rate": 9.945226012599117e-06, "loss": 0.3783, "step": 968 }, { "epoch": 0.4285556333615892, "grad_norm": 0.3463248608060258, "learning_rate": 9.944845679383248e-06, "loss": 0.4313, "step": 969 }, { "epoch": 0.42899789923709136, "grad_norm": 0.31486747115470687, "learning_rate": 9.944464037598347e-06, "loss": 0.3785, "step": 970 }, { "epoch": 0.42944016511259353, "grad_norm": 0.3649902197335104, "learning_rate": 9.944081087345407e-06, "loss": 0.394, "step": 971 }, { "epoch": 0.4298824309880957, "grad_norm": 0.39599999168068056, "learning_rate": 9.943696828725774e-06, "loss": 0.386, "step": 972 }, { "epoch": 0.4303246968635978, "grad_norm": 0.3604767570938464, "learning_rate": 9.94331126184113e-06, "loss": 0.3801, "step": 973 }, { "epoch": 0.4307669627391, "grad_norm": 0.39024141765717535, "learning_rate": 9.942924386793513e-06, "loss": 0.4089, "step": 974 }, { "epoch": 0.43120922861460215, "grad_norm": 0.3235820977382176, "learning_rate": 9.9425362036853e-06, "loss": 0.435, "step": 975 }, { "epoch": 0.4316514944901043, "grad_norm": 0.36077923536001555, "learning_rate": 9.942146712619217e-06, "loss": 0.4208, "step": 976 }, { "epoch": 0.4320937603656065, "grad_norm": 0.32318963117614097, "learning_rate": 9.941755913698339e-06, "loss": 0.4138, "step": 977 }, { "epoch": 0.4325360262411086, "grad_norm": 0.32356542973591007, "learning_rate": 9.941363807026082e-06, "loss": 0.372, "step": 978 }, { "epoch": 0.43297829211661076, "grad_norm": 0.3536255159289332, "learning_rate": 9.94097039270621e-06, "loss": 0.3619, "step": 979 }, { "epoch": 0.43342055799211293, "grad_norm": 0.2925562651865244, "learning_rate": 9.940575670842836e-06, "loss": 0.3557, "step": 980 }, { "epoch": 0.4338628238676151, "grad_norm": 0.2900013186130819, "learning_rate": 9.940179641540414e-06, "loss": 0.3873, "step": 981 }, { "epoch": 0.4343050897431172, "grad_norm": 0.3967901065463964, "learning_rate": 9.939782304903746e-06, "loss": 0.4157, "step": 982 }, { "epoch": 0.4347473556186194, "grad_norm": 0.30426279564430037, "learning_rate": 9.939383661037983e-06, "loss": 0.3802, "step": 983 }, { "epoch": 0.43518962149412155, "grad_norm": 0.37379535588195895, "learning_rate": 9.938983710048617e-06, "loss": 0.4455, "step": 984 }, { "epoch": 0.4356318873696237, "grad_norm": 0.41171863614822496, "learning_rate": 9.93858245204149e-06, "loss": 0.407, "step": 985 }, { "epoch": 0.4360741532451259, "grad_norm": 0.30912681961681127, "learning_rate": 9.938179887122785e-06, "loss": 0.3958, "step": 986 }, { "epoch": 0.436516419120628, "grad_norm": 0.396055420823425, "learning_rate": 9.93777601539904e-06, "loss": 0.4633, "step": 987 }, { "epoch": 0.43695868499613016, "grad_norm": 0.3023752924432356, "learning_rate": 9.937370836977125e-06, "loss": 0.3805, "step": 988 }, { "epoch": 0.43740095087163233, "grad_norm": 0.3328521331314838, "learning_rate": 9.936964351964266e-06, "loss": 0.3861, "step": 989 }, { "epoch": 0.4378432167471345, "grad_norm": 0.31843019392357463, "learning_rate": 9.936556560468037e-06, "loss": 0.3958, "step": 990 }, { "epoch": 0.43828548262263667, "grad_norm": 0.35472443370193524, "learning_rate": 9.936147462596348e-06, "loss": 0.4322, "step": 991 }, { "epoch": 0.4387277484981388, "grad_norm": 0.3195855814179417, "learning_rate": 9.935737058457462e-06, "loss": 0.3827, "step": 992 }, { "epoch": 0.43917001437364095, "grad_norm": 0.3097662206686766, "learning_rate": 9.935325348159985e-06, "loss": 0.3604, "step": 993 }, { "epoch": 0.4396122802491431, "grad_norm": 0.2706177157428936, "learning_rate": 9.93491233181287e-06, "loss": 0.3856, "step": 994 }, { "epoch": 0.4400545461246453, "grad_norm": 0.4042577704213826, "learning_rate": 9.934498009525411e-06, "loss": 0.39, "step": 995 }, { "epoch": 0.4404968120001474, "grad_norm": 0.3996282993085792, "learning_rate": 9.934082381407257e-06, "loss": 0.4353, "step": 996 }, { "epoch": 0.44093907787564957, "grad_norm": 0.5040434894154513, "learning_rate": 9.933665447568393e-06, "loss": 0.4262, "step": 997 }, { "epoch": 0.44138134375115173, "grad_norm": 0.35382327252230444, "learning_rate": 9.933247208119151e-06, "loss": 0.4139, "step": 998 }, { "epoch": 0.4418236096266539, "grad_norm": 0.31138649616146313, "learning_rate": 9.932827663170219e-06, "loss": 0.3844, "step": 999 }, { "epoch": 0.44226587550215607, "grad_norm": 0.3024943799190856, "learning_rate": 9.932406812832616e-06, "loss": 0.3947, "step": 1000 }, { "epoch": 0.4427081413776582, "grad_norm": 0.3037490094535797, "learning_rate": 9.931984657217714e-06, "loss": 0.379, "step": 1001 }, { "epoch": 0.44315040725316035, "grad_norm": 0.2841134445029086, "learning_rate": 9.93156119643723e-06, "loss": 0.3963, "step": 1002 }, { "epoch": 0.4435926731286625, "grad_norm": 0.34692280193898567, "learning_rate": 9.931136430603226e-06, "loss": 0.3985, "step": 1003 }, { "epoch": 0.4440349390041647, "grad_norm": 0.33595571706019395, "learning_rate": 9.930710359828109e-06, "loss": 0.386, "step": 1004 }, { "epoch": 0.44447720487966685, "grad_norm": 0.3214949221295631, "learning_rate": 9.93028298422463e-06, "loss": 0.3683, "step": 1005 }, { "epoch": 0.44491947075516897, "grad_norm": 0.34801833270816074, "learning_rate": 9.929854303905888e-06, "loss": 0.3602, "step": 1006 }, { "epoch": 0.44536173663067113, "grad_norm": 0.3234916847455004, "learning_rate": 9.929424318985326e-06, "loss": 0.4162, "step": 1007 }, { "epoch": 0.4458040025061733, "grad_norm": 0.32120157041824954, "learning_rate": 9.92899302957673e-06, "loss": 0.4019, "step": 1008 }, { "epoch": 0.44624626838167547, "grad_norm": 0.33374524815896384, "learning_rate": 9.928560435794237e-06, "loss": 0.4183, "step": 1009 }, { "epoch": 0.4466885342571776, "grad_norm": 0.32744539715709503, "learning_rate": 9.928126537752323e-06, "loss": 0.3878, "step": 1010 }, { "epoch": 0.44713080013267975, "grad_norm": 0.2986528384711046, "learning_rate": 9.927691335565812e-06, "loss": 0.3842, "step": 1011 }, { "epoch": 0.4475730660081819, "grad_norm": 0.3231145965840068, "learning_rate": 9.927254829349872e-06, "loss": 0.385, "step": 1012 }, { "epoch": 0.4480153318836841, "grad_norm": 0.34998563227916657, "learning_rate": 9.92681701922002e-06, "loss": 0.3943, "step": 1013 }, { "epoch": 0.44845759775918626, "grad_norm": 0.3067780307683369, "learning_rate": 9.926377905292113e-06, "loss": 0.4163, "step": 1014 }, { "epoch": 0.44889986363468837, "grad_norm": 0.4163480417432856, "learning_rate": 9.925937487682354e-06, "loss": 0.4176, "step": 1015 }, { "epoch": 0.44934212951019054, "grad_norm": 0.34095590076141263, "learning_rate": 9.925495766507293e-06, "loss": 0.4052, "step": 1016 }, { "epoch": 0.4497843953856927, "grad_norm": 0.3781438652190227, "learning_rate": 9.925052741883825e-06, "loss": 0.3843, "step": 1017 }, { "epoch": 0.45022666126119487, "grad_norm": 0.3200643137927971, "learning_rate": 9.924608413929189e-06, "loss": 0.4291, "step": 1018 }, { "epoch": 0.450668927136697, "grad_norm": 0.3663223037023744, "learning_rate": 9.924162782760965e-06, "loss": 0.4433, "step": 1019 }, { "epoch": 0.45111119301219915, "grad_norm": 0.37653482970304086, "learning_rate": 9.923715848497087e-06, "loss": 0.4098, "step": 1020 }, { "epoch": 0.4515534588877013, "grad_norm": 0.35158245444111774, "learning_rate": 9.923267611255824e-06, "loss": 0.3827, "step": 1021 }, { "epoch": 0.4519957247632035, "grad_norm": 0.3520278786563293, "learning_rate": 9.922818071155797e-06, "loss": 0.3379, "step": 1022 }, { "epoch": 0.45243799063870566, "grad_norm": 0.34232977750638083, "learning_rate": 9.922367228315969e-06, "loss": 0.3838, "step": 1023 }, { "epoch": 0.45288025651420777, "grad_norm": 0.368125510273477, "learning_rate": 9.921915082855647e-06, "loss": 0.4052, "step": 1024 }, { "epoch": 0.45332252238970994, "grad_norm": 0.3132002536946878, "learning_rate": 9.921461634894484e-06, "loss": 0.3886, "step": 1025 }, { "epoch": 0.4537647882652121, "grad_norm": 0.3242402906118547, "learning_rate": 9.921006884552477e-06, "loss": 0.3934, "step": 1026 }, { "epoch": 0.4542070541407143, "grad_norm": 0.3013768061684586, "learning_rate": 9.920550831949966e-06, "loss": 0.3569, "step": 1027 }, { "epoch": 0.45464932001621644, "grad_norm": 0.35065366093944816, "learning_rate": 9.920093477207643e-06, "loss": 0.3994, "step": 1028 }, { "epoch": 0.45509158589171855, "grad_norm": 0.4066526755360713, "learning_rate": 9.919634820446535e-06, "loss": 0.3967, "step": 1029 }, { "epoch": 0.4555338517672207, "grad_norm": 0.3023970039369827, "learning_rate": 9.919174861788018e-06, "loss": 0.418, "step": 1030 }, { "epoch": 0.4559761176427229, "grad_norm": 0.36341797485156424, "learning_rate": 9.918713601353812e-06, "loss": 0.3966, "step": 1031 }, { "epoch": 0.45641838351822506, "grad_norm": 0.3298629753860437, "learning_rate": 9.918251039265983e-06, "loss": 0.3737, "step": 1032 }, { "epoch": 0.45686064939372717, "grad_norm": 0.33353042941514466, "learning_rate": 9.917787175646938e-06, "loss": 0.3707, "step": 1033 }, { "epoch": 0.45730291526922934, "grad_norm": 0.3053882054396294, "learning_rate": 9.917322010619434e-06, "loss": 0.3797, "step": 1034 }, { "epoch": 0.4577451811447315, "grad_norm": 0.339613920935648, "learning_rate": 9.916855544306565e-06, "loss": 0.3328, "step": 1035 }, { "epoch": 0.4581874470202337, "grad_norm": 0.28605229511610175, "learning_rate": 9.916387776831778e-06, "loss": 0.3564, "step": 1036 }, { "epoch": 0.45862971289573584, "grad_norm": 0.34415303523901597, "learning_rate": 9.915918708318856e-06, "loss": 0.3925, "step": 1037 }, { "epoch": 0.45907197877123795, "grad_norm": 0.5218267768892777, "learning_rate": 9.91544833889193e-06, "loss": 0.4077, "step": 1038 }, { "epoch": 0.4595142446467401, "grad_norm": 0.6294332072884748, "learning_rate": 9.914976668675478e-06, "loss": 0.4027, "step": 1039 }, { "epoch": 0.4599565105222423, "grad_norm": 0.3105097122106025, "learning_rate": 9.914503697794317e-06, "loss": 0.4177, "step": 1040 }, { "epoch": 0.46039877639774446, "grad_norm": 0.31761141963434275, "learning_rate": 9.914029426373612e-06, "loss": 0.3676, "step": 1041 }, { "epoch": 0.4608410422732466, "grad_norm": 0.2831879453279003, "learning_rate": 9.91355385453887e-06, "loss": 0.4176, "step": 1042 }, { "epoch": 0.46128330814874874, "grad_norm": 0.3604489887659748, "learning_rate": 9.913076982415944e-06, "loss": 0.3791, "step": 1043 }, { "epoch": 0.4617255740242509, "grad_norm": 0.3895784783030283, "learning_rate": 9.912598810131027e-06, "loss": 0.4207, "step": 1044 }, { "epoch": 0.4621678398997531, "grad_norm": 0.35901594689614347, "learning_rate": 9.912119337810664e-06, "loss": 0.395, "step": 1045 }, { "epoch": 0.46261010577525524, "grad_norm": 0.4150240647659328, "learning_rate": 9.911638565581736e-06, "loss": 0.4008, "step": 1046 }, { "epoch": 0.46305237165075736, "grad_norm": 0.30527350620715743, "learning_rate": 9.91115649357147e-06, "loss": 0.3745, "step": 1047 }, { "epoch": 0.4634946375262595, "grad_norm": 0.3486786764458767, "learning_rate": 9.910673121907443e-06, "loss": 0.3686, "step": 1048 }, { "epoch": 0.4639369034017617, "grad_norm": 0.31766466313059194, "learning_rate": 9.910188450717565e-06, "loss": 0.3834, "step": 1049 }, { "epoch": 0.46437916927726386, "grad_norm": 0.3428058216384875, "learning_rate": 9.9097024801301e-06, "loss": 0.4124, "step": 1050 }, { "epoch": 0.464821435152766, "grad_norm": 0.2957126884930687, "learning_rate": 9.909215210273653e-06, "loss": 0.3828, "step": 1051 }, { "epoch": 0.46526370102826814, "grad_norm": 0.33220786314630285, "learning_rate": 9.908726641277167e-06, "loss": 0.3837, "step": 1052 }, { "epoch": 0.4657059669037703, "grad_norm": 0.3294514073671839, "learning_rate": 9.908236773269936e-06, "loss": 0.3919, "step": 1053 }, { "epoch": 0.4661482327792725, "grad_norm": 0.3047610755299818, "learning_rate": 9.907745606381598e-06, "loss": 0.3531, "step": 1054 }, { "epoch": 0.46659049865477464, "grad_norm": 0.32940308070802604, "learning_rate": 9.907253140742126e-06, "loss": 0.387, "step": 1055 }, { "epoch": 0.4670327645302768, "grad_norm": 0.3609533488532848, "learning_rate": 9.906759376481846e-06, "loss": 0.4379, "step": 1056 }, { "epoch": 0.4674750304057789, "grad_norm": 0.3656182943815668, "learning_rate": 9.906264313731426e-06, "loss": 0.4059, "step": 1057 }, { "epoch": 0.4679172962812811, "grad_norm": 0.30682901670923945, "learning_rate": 9.905767952621871e-06, "loss": 0.3999, "step": 1058 }, { "epoch": 0.46835956215678326, "grad_norm": 0.29778021838377267, "learning_rate": 9.905270293284539e-06, "loss": 0.3685, "step": 1059 }, { "epoch": 0.46880182803228543, "grad_norm": 0.27678943995051153, "learning_rate": 9.904771335851126e-06, "loss": 0.4124, "step": 1060 }, { "epoch": 0.46924409390778754, "grad_norm": 0.3170287867960451, "learning_rate": 9.904271080453671e-06, "loss": 0.395, "step": 1061 }, { "epoch": 0.4696863597832897, "grad_norm": 0.38520370664787656, "learning_rate": 9.903769527224559e-06, "loss": 0.4091, "step": 1062 }, { "epoch": 0.4701286256587919, "grad_norm": 0.32728640375197976, "learning_rate": 9.903266676296517e-06, "loss": 0.3976, "step": 1063 }, { "epoch": 0.47057089153429404, "grad_norm": 0.3313112884211205, "learning_rate": 9.902762527802618e-06, "loss": 0.3783, "step": 1064 }, { "epoch": 0.4710131574097962, "grad_norm": 0.30808577125076414, "learning_rate": 9.902257081876274e-06, "loss": 0.439, "step": 1065 }, { "epoch": 0.4714554232852983, "grad_norm": 0.3370911878872846, "learning_rate": 9.901750338651244e-06, "loss": 0.3988, "step": 1066 }, { "epoch": 0.4718976891608005, "grad_norm": 0.9298663001529227, "learning_rate": 9.901242298261627e-06, "loss": 0.4338, "step": 1067 }, { "epoch": 0.47233995503630266, "grad_norm": 0.3419333062476995, "learning_rate": 9.90073296084187e-06, "loss": 0.4283, "step": 1068 }, { "epoch": 0.47278222091180483, "grad_norm": 0.307170026294877, "learning_rate": 9.900222326526758e-06, "loss": 0.4205, "step": 1069 }, { "epoch": 0.473224486787307, "grad_norm": 0.36998621555883376, "learning_rate": 9.899710395451423e-06, "loss": 0.4135, "step": 1070 }, { "epoch": 0.4736667526628091, "grad_norm": 0.3377470267582217, "learning_rate": 9.899197167751338e-06, "loss": 0.4166, "step": 1071 }, { "epoch": 0.4741090185383113, "grad_norm": 0.32388569991463606, "learning_rate": 9.898682643562321e-06, "loss": 0.3541, "step": 1072 }, { "epoch": 0.47455128441381345, "grad_norm": 0.4038383283965275, "learning_rate": 9.898166823020532e-06, "loss": 0.4309, "step": 1073 }, { "epoch": 0.4749935502893156, "grad_norm": 0.33074244664802915, "learning_rate": 9.897649706262474e-06, "loss": 0.395, "step": 1074 }, { "epoch": 0.4754358161648177, "grad_norm": 0.3136883294054388, "learning_rate": 9.897131293424992e-06, "loss": 0.423, "step": 1075 }, { "epoch": 0.4758780820403199, "grad_norm": 0.336426978227042, "learning_rate": 9.896611584645276e-06, "loss": 0.3763, "step": 1076 }, { "epoch": 0.47632034791582206, "grad_norm": 0.27884154506626996, "learning_rate": 9.896090580060857e-06, "loss": 0.3881, "step": 1077 }, { "epoch": 0.47676261379132423, "grad_norm": 0.2996102511442864, "learning_rate": 9.89556827980961e-06, "loss": 0.3823, "step": 1078 }, { "epoch": 0.4772048796668264, "grad_norm": 0.30046346024113013, "learning_rate": 9.895044684029757e-06, "loss": 0.3787, "step": 1079 }, { "epoch": 0.4776471455423285, "grad_norm": 0.3303916017731984, "learning_rate": 9.894519792859854e-06, "loss": 0.3694, "step": 1080 }, { "epoch": 0.4780894114178307, "grad_norm": 0.31379252017401, "learning_rate": 9.893993606438805e-06, "loss": 0.3971, "step": 1081 }, { "epoch": 0.47853167729333285, "grad_norm": 0.3870467623251512, "learning_rate": 9.893466124905857e-06, "loss": 0.3847, "step": 1082 }, { "epoch": 0.478973943168835, "grad_norm": 0.41695547655967563, "learning_rate": 9.8929373484006e-06, "loss": 0.3987, "step": 1083 }, { "epoch": 0.4794162090443371, "grad_norm": 0.37761108155903705, "learning_rate": 9.892407277062965e-06, "loss": 0.4142, "step": 1084 }, { "epoch": 0.4798584749198393, "grad_norm": 0.3165137331928078, "learning_rate": 9.891875911033226e-06, "loss": 0.4057, "step": 1085 }, { "epoch": 0.48030074079534146, "grad_norm": 0.34202623911314406, "learning_rate": 9.891343250452002e-06, "loss": 0.3895, "step": 1086 }, { "epoch": 0.48074300667084363, "grad_norm": 0.29727432502375106, "learning_rate": 9.890809295460249e-06, "loss": 0.3456, "step": 1087 }, { "epoch": 0.4811852725463458, "grad_norm": 0.31978530282621837, "learning_rate": 9.89027404619927e-06, "loss": 0.3858, "step": 1088 }, { "epoch": 0.4816275384218479, "grad_norm": 0.4536856331457971, "learning_rate": 9.88973750281071e-06, "loss": 0.444, "step": 1089 }, { "epoch": 0.4820698042973501, "grad_norm": 0.44231238119632427, "learning_rate": 9.889199665436557e-06, "loss": 0.4015, "step": 1090 }, { "epoch": 0.48251207017285225, "grad_norm": 0.34083051031233846, "learning_rate": 9.888660534219141e-06, "loss": 0.4352, "step": 1091 }, { "epoch": 0.4829543360483544, "grad_norm": 0.29262963967255373, "learning_rate": 9.88812010930113e-06, "loss": 0.3946, "step": 1092 }, { "epoch": 0.4833966019238566, "grad_norm": 0.34192760715502074, "learning_rate": 9.887578390825545e-06, "loss": 0.3663, "step": 1093 }, { "epoch": 0.4838388677993587, "grad_norm": 0.31848208363410385, "learning_rate": 9.887035378935737e-06, "loss": 0.4204, "step": 1094 }, { "epoch": 0.48428113367486086, "grad_norm": 0.463364084460376, "learning_rate": 9.886491073775406e-06, "loss": 0.4623, "step": 1095 }, { "epoch": 0.48472339955036303, "grad_norm": 0.33123446918665517, "learning_rate": 9.885945475488592e-06, "loss": 0.3686, "step": 1096 }, { "epoch": 0.4851656654258652, "grad_norm": 0.3475694785469819, "learning_rate": 9.88539858421968e-06, "loss": 0.4028, "step": 1097 }, { "epoch": 0.4856079313013673, "grad_norm": 0.3096338560188106, "learning_rate": 9.884850400113398e-06, "loss": 0.3676, "step": 1098 }, { "epoch": 0.4860501971768695, "grad_norm": 0.34263879109526485, "learning_rate": 9.884300923314806e-06, "loss": 0.4176, "step": 1099 }, { "epoch": 0.48649246305237165, "grad_norm": 0.3442176779312205, "learning_rate": 9.883750153969323e-06, "loss": 0.3969, "step": 1100 }, { "epoch": 0.4869347289278738, "grad_norm": 0.31767607057996716, "learning_rate": 9.883198092222694e-06, "loss": 0.4347, "step": 1101 }, { "epoch": 0.487376994803376, "grad_norm": 0.26527395218096667, "learning_rate": 9.882644738221014e-06, "loss": 0.3364, "step": 1102 }, { "epoch": 0.4878192606788781, "grad_norm": 0.347081753857588, "learning_rate": 9.88209009211072e-06, "loss": 0.427, "step": 1103 }, { "epoch": 0.48826152655438027, "grad_norm": 0.2886586340230586, "learning_rate": 9.881534154038591e-06, "loss": 0.3675, "step": 1104 }, { "epoch": 0.48870379242988243, "grad_norm": 0.34692534206387826, "learning_rate": 9.880976924151745e-06, "loss": 0.407, "step": 1105 }, { "epoch": 0.4891460583053846, "grad_norm": 0.3137894913198346, "learning_rate": 9.88041840259764e-06, "loss": 0.4045, "step": 1106 }, { "epoch": 0.48958832418088677, "grad_norm": 0.3283915513903897, "learning_rate": 9.879858589524083e-06, "loss": 0.3932, "step": 1107 }, { "epoch": 0.4900305900563889, "grad_norm": 0.3897675577250338, "learning_rate": 9.879297485079218e-06, "loss": 0.3793, "step": 1108 }, { "epoch": 0.49047285593189105, "grad_norm": 0.41562977843461524, "learning_rate": 9.878735089411532e-06, "loss": 0.4023, "step": 1109 }, { "epoch": 0.4909151218073932, "grad_norm": 0.32191388792619746, "learning_rate": 9.878171402669856e-06, "loss": 0.356, "step": 1110 }, { "epoch": 0.4913573876828954, "grad_norm": 0.33184309034614984, "learning_rate": 9.877606425003355e-06, "loss": 0.3854, "step": 1111 }, { "epoch": 0.4917996535583975, "grad_norm": 0.3106591506231308, "learning_rate": 9.877040156561543e-06, "loss": 0.4248, "step": 1112 }, { "epoch": 0.49224191943389967, "grad_norm": 0.30899482576156356, "learning_rate": 9.876472597494276e-06, "loss": 0.3691, "step": 1113 }, { "epoch": 0.49268418530940183, "grad_norm": 0.3595605542849125, "learning_rate": 9.875903747951742e-06, "loss": 0.3844, "step": 1114 }, { "epoch": 0.493126451184904, "grad_norm": 0.28688315235606354, "learning_rate": 9.875333608084486e-06, "loss": 0.4363, "step": 1115 }, { "epoch": 0.49356871706040617, "grad_norm": 0.38467066144858075, "learning_rate": 9.874762178043378e-06, "loss": 0.3667, "step": 1116 }, { "epoch": 0.4940109829359083, "grad_norm": 0.28529961290618355, "learning_rate": 9.874189457979641e-06, "loss": 0.3585, "step": 1117 }, { "epoch": 0.49445324881141045, "grad_norm": 0.3425200080377474, "learning_rate": 9.873615448044836e-06, "loss": 0.3977, "step": 1118 }, { "epoch": 0.4948955146869126, "grad_norm": 0.3632469460000522, "learning_rate": 9.873040148390865e-06, "loss": 0.3651, "step": 1119 }, { "epoch": 0.4953377805624148, "grad_norm": 0.5071412791983807, "learning_rate": 9.872463559169969e-06, "loss": 0.4395, "step": 1120 }, { "epoch": 0.49578004643791695, "grad_norm": 0.3401420001833221, "learning_rate": 9.871885680534734e-06, "loss": 0.3823, "step": 1121 }, { "epoch": 0.49622231231341907, "grad_norm": 0.3737509375114271, "learning_rate": 9.871306512638085e-06, "loss": 0.4061, "step": 1122 }, { "epoch": 0.49666457818892124, "grad_norm": 0.34658424486753503, "learning_rate": 9.87072605563329e-06, "loss": 0.4072, "step": 1123 }, { "epoch": 0.4971068440644234, "grad_norm": 0.292308954681471, "learning_rate": 9.870144309673956e-06, "loss": 0.3723, "step": 1124 }, { "epoch": 0.49754910993992557, "grad_norm": 0.39360417787575247, "learning_rate": 9.869561274914033e-06, "loss": 0.4014, "step": 1125 }, { "epoch": 0.4979913758154277, "grad_norm": 0.31338209449085724, "learning_rate": 9.86897695150781e-06, "loss": 0.3954, "step": 1126 }, { "epoch": 0.49843364169092985, "grad_norm": 0.32042584487145614, "learning_rate": 9.868391339609919e-06, "loss": 0.4497, "step": 1127 }, { "epoch": 0.498875907566432, "grad_norm": 0.3183858489982941, "learning_rate": 9.867804439375334e-06, "loss": 0.3793, "step": 1128 }, { "epoch": 0.4993181734419342, "grad_norm": 0.3170882787593715, "learning_rate": 9.867216250959367e-06, "loss": 0.3775, "step": 1129 }, { "epoch": 0.49976043931743636, "grad_norm": 0.30523117045694764, "learning_rate": 9.86662677451767e-06, "loss": 0.3854, "step": 1130 }, { "epoch": 0.5002027051929385, "grad_norm": 0.3350926979260162, "learning_rate": 9.866036010206237e-06, "loss": 0.3705, "step": 1131 }, { "epoch": 0.5006449710684406, "grad_norm": 0.2862824083067664, "learning_rate": 9.865443958181408e-06, "loss": 0.3236, "step": 1132 }, { "epoch": 0.5010872369439427, "grad_norm": 0.3365923588675633, "learning_rate": 9.864850618599858e-06, "loss": 0.3822, "step": 1133 }, { "epoch": 0.501529502819445, "grad_norm": 0.3106272405221044, "learning_rate": 9.864255991618603e-06, "loss": 0.3488, "step": 1134 }, { "epoch": 0.5019717686949471, "grad_norm": 0.3158682627766929, "learning_rate": 9.863660077395002e-06, "loss": 0.4015, "step": 1135 }, { "epoch": 0.5024140345704493, "grad_norm": 0.3173839927584357, "learning_rate": 9.863062876086753e-06, "loss": 0.3741, "step": 1136 }, { "epoch": 0.5028563004459514, "grad_norm": 0.3391597674497146, "learning_rate": 9.862464387851895e-06, "loss": 0.3958, "step": 1137 }, { "epoch": 0.5032985663214535, "grad_norm": 0.34430530622322947, "learning_rate": 9.861864612848808e-06, "loss": 0.4064, "step": 1138 }, { "epoch": 0.5037408321969558, "grad_norm": 0.34572996394619093, "learning_rate": 9.861263551236212e-06, "loss": 0.4139, "step": 1139 }, { "epoch": 0.5041830980724579, "grad_norm": 0.3360829380666617, "learning_rate": 9.860661203173167e-06, "loss": 0.3856, "step": 1140 }, { "epoch": 0.5046253639479601, "grad_norm": 0.35173653369299646, "learning_rate": 9.860057568819075e-06, "loss": 0.3818, "step": 1141 }, { "epoch": 0.5050676298234622, "grad_norm": 0.316918850456826, "learning_rate": 9.859452648333678e-06, "loss": 0.3828, "step": 1142 }, { "epoch": 0.5055098956989643, "grad_norm": 0.49735814872927836, "learning_rate": 9.858846441877057e-06, "loss": 0.3785, "step": 1143 }, { "epoch": 0.5059521615744665, "grad_norm": 0.34513971349578076, "learning_rate": 9.858238949609636e-06, "loss": 0.4126, "step": 1144 }, { "epoch": 0.5063944274499687, "grad_norm": 0.3127000280673388, "learning_rate": 9.857630171692175e-06, "loss": 0.4015, "step": 1145 }, { "epoch": 0.5068366933254709, "grad_norm": 0.3017162992719264, "learning_rate": 9.857020108285777e-06, "loss": 0.3713, "step": 1146 }, { "epoch": 0.507278959200973, "grad_norm": 0.41443057711583875, "learning_rate": 9.856408759551885e-06, "loss": 0.4503, "step": 1147 }, { "epoch": 0.5077212250764751, "grad_norm": 0.32664833607831184, "learning_rate": 9.855796125652283e-06, "loss": 0.399, "step": 1148 }, { "epoch": 0.5081634909519773, "grad_norm": 0.3256231998149263, "learning_rate": 9.855182206749095e-06, "loss": 0.4089, "step": 1149 }, { "epoch": 0.5086057568274794, "grad_norm": 0.36338202482483956, "learning_rate": 9.854567003004782e-06, "loss": 0.4204, "step": 1150 }, { "epoch": 0.5090480227029817, "grad_norm": 0.32213132006224376, "learning_rate": 9.853950514582149e-06, "loss": 0.3769, "step": 1151 }, { "epoch": 0.5094902885784838, "grad_norm": 0.33125975642299155, "learning_rate": 9.853332741644337e-06, "loss": 0.376, "step": 1152 }, { "epoch": 0.5099325544539859, "grad_norm": 0.34255061948919, "learning_rate": 9.85271368435483e-06, "loss": 0.4266, "step": 1153 }, { "epoch": 0.5103748203294881, "grad_norm": 0.32549620805156454, "learning_rate": 9.852093342877454e-06, "loss": 0.3774, "step": 1154 }, { "epoch": 0.5108170862049902, "grad_norm": 0.36597540974404563, "learning_rate": 9.851471717376367e-06, "loss": 0.3884, "step": 1155 }, { "epoch": 0.5112593520804923, "grad_norm": 0.3542625960022111, "learning_rate": 9.85084880801607e-06, "loss": 0.3753, "step": 1156 }, { "epoch": 0.5117016179559946, "grad_norm": 0.2845684434070106, "learning_rate": 9.850224614961415e-06, "loss": 0.3949, "step": 1157 }, { "epoch": 0.5121438838314967, "grad_norm": 0.3083409691979219, "learning_rate": 9.849599138377574e-06, "loss": 0.3836, "step": 1158 }, { "epoch": 0.5125861497069989, "grad_norm": 0.34809837818198186, "learning_rate": 9.848972378430074e-06, "loss": 0.3742, "step": 1159 }, { "epoch": 0.513028415582501, "grad_norm": 0.303936937354364, "learning_rate": 9.848344335284774e-06, "loss": 0.3834, "step": 1160 }, { "epoch": 0.5134706814580031, "grad_norm": 0.3146073654579281, "learning_rate": 9.847715009107876e-06, "loss": 0.3368, "step": 1161 }, { "epoch": 0.5139129473335053, "grad_norm": 0.2979074673677599, "learning_rate": 9.847084400065921e-06, "loss": 0.3791, "step": 1162 }, { "epoch": 0.5143552132090075, "grad_norm": 0.3363370049171941, "learning_rate": 9.846452508325788e-06, "loss": 0.4081, "step": 1163 }, { "epoch": 0.5147974790845097, "grad_norm": 0.32744203491319307, "learning_rate": 9.845819334054699e-06, "loss": 0.3979, "step": 1164 }, { "epoch": 0.5152397449600118, "grad_norm": 0.5575167012001903, "learning_rate": 9.845184877420207e-06, "loss": 0.4005, "step": 1165 }, { "epoch": 0.5156820108355139, "grad_norm": 0.33804229048306117, "learning_rate": 9.844549138590214e-06, "loss": 0.3671, "step": 1166 }, { "epoch": 0.5161242767110161, "grad_norm": 0.3255280522347831, "learning_rate": 9.843912117732957e-06, "loss": 0.3814, "step": 1167 }, { "epoch": 0.5165665425865182, "grad_norm": 0.3741733855550193, "learning_rate": 9.843273815017013e-06, "loss": 0.4324, "step": 1168 }, { "epoch": 0.5170088084620205, "grad_norm": 0.33778361217353325, "learning_rate": 9.842634230611297e-06, "loss": 0.4069, "step": 1169 }, { "epoch": 0.5174510743375226, "grad_norm": 0.3476856914304497, "learning_rate": 9.841993364685066e-06, "loss": 0.4112, "step": 1170 }, { "epoch": 0.5178933402130247, "grad_norm": 0.32616625512278374, "learning_rate": 9.84135121740791e-06, "loss": 0.4134, "step": 1171 }, { "epoch": 0.5183356060885269, "grad_norm": 0.3642281436316887, "learning_rate": 9.840707788949768e-06, "loss": 0.4236, "step": 1172 }, { "epoch": 0.518777871964029, "grad_norm": 0.3552493403579972, "learning_rate": 9.840063079480908e-06, "loss": 0.3879, "step": 1173 }, { "epoch": 0.5192201378395312, "grad_norm": 0.35822513401141326, "learning_rate": 9.839417089171943e-06, "loss": 0.4454, "step": 1174 }, { "epoch": 0.5196624037150334, "grad_norm": 0.30220411736668257, "learning_rate": 9.838769818193823e-06, "loss": 0.3814, "step": 1175 }, { "epoch": 0.5201046695905355, "grad_norm": 0.29575176331071906, "learning_rate": 9.83812126671784e-06, "loss": 0.3438, "step": 1176 }, { "epoch": 0.5205469354660377, "grad_norm": 0.33003801093689705, "learning_rate": 9.837471434915617e-06, "loss": 0.3493, "step": 1177 }, { "epoch": 0.5209892013415398, "grad_norm": 0.30123926073366736, "learning_rate": 9.836820322959124e-06, "loss": 0.3751, "step": 1178 }, { "epoch": 0.521431467217042, "grad_norm": 0.390421441961968, "learning_rate": 9.836167931020666e-06, "loss": 0.443, "step": 1179 }, { "epoch": 0.5218737330925441, "grad_norm": 0.3244608497779346, "learning_rate": 9.835514259272886e-06, "loss": 0.3981, "step": 1180 }, { "epoch": 0.5223159989680463, "grad_norm": 0.3300601839140197, "learning_rate": 9.834859307888771e-06, "loss": 0.3896, "step": 1181 }, { "epoch": 0.5227582648435485, "grad_norm": 0.31836130082718456, "learning_rate": 9.834203077041639e-06, "loss": 0.3694, "step": 1182 }, { "epoch": 0.5232005307190506, "grad_norm": 0.423984459447207, "learning_rate": 9.833545566905152e-06, "loss": 0.3795, "step": 1183 }, { "epoch": 0.5236427965945527, "grad_norm": 0.3251112985915475, "learning_rate": 9.832886777653307e-06, "loss": 0.4432, "step": 1184 }, { "epoch": 0.5240850624700549, "grad_norm": 0.3400820512968421, "learning_rate": 9.832226709460444e-06, "loss": 0.4508, "step": 1185 }, { "epoch": 0.524527328345557, "grad_norm": 0.33444187685570737, "learning_rate": 9.831565362501238e-06, "loss": 0.3581, "step": 1186 }, { "epoch": 0.5249695942210593, "grad_norm": 0.30030565761685296, "learning_rate": 9.830902736950702e-06, "loss": 0.4004, "step": 1187 }, { "epoch": 0.5254118600965614, "grad_norm": 0.4031044794201674, "learning_rate": 9.83023883298419e-06, "loss": 0.3899, "step": 1188 }, { "epoch": 0.5258541259720635, "grad_norm": 0.31237473075803224, "learning_rate": 9.829573650777391e-06, "loss": 0.3582, "step": 1189 }, { "epoch": 0.5262963918475657, "grad_norm": 0.371737017173343, "learning_rate": 9.828907190506335e-06, "loss": 0.4101, "step": 1190 }, { "epoch": 0.5267386577230678, "grad_norm": 0.3780384559986603, "learning_rate": 9.828239452347392e-06, "loss": 0.3964, "step": 1191 }, { "epoch": 0.52718092359857, "grad_norm": 0.3133707205072211, "learning_rate": 9.827570436477262e-06, "loss": 0.4265, "step": 1192 }, { "epoch": 0.5276231894740722, "grad_norm": 0.31603808423503316, "learning_rate": 9.826900143072993e-06, "loss": 0.3768, "step": 1193 }, { "epoch": 0.5280654553495743, "grad_norm": 0.320652394900035, "learning_rate": 9.826228572311966e-06, "loss": 0.3925, "step": 1194 }, { "epoch": 0.5285077212250765, "grad_norm": 0.31534290594871506, "learning_rate": 9.825555724371899e-06, "loss": 0.3578, "step": 1195 }, { "epoch": 0.5289499871005786, "grad_norm": 0.37846720733755423, "learning_rate": 9.82488159943085e-06, "loss": 0.3975, "step": 1196 }, { "epoch": 0.5293922529760808, "grad_norm": 0.32109145920708376, "learning_rate": 9.824206197667218e-06, "loss": 0.3585, "step": 1197 }, { "epoch": 0.529834518851583, "grad_norm": 0.3380341145766143, "learning_rate": 9.823529519259734e-06, "loss": 0.3772, "step": 1198 }, { "epoch": 0.5302767847270851, "grad_norm": 0.32697801732961, "learning_rate": 9.822851564387468e-06, "loss": 0.4073, "step": 1199 }, { "epoch": 0.5307190506025873, "grad_norm": 0.31673388737273167, "learning_rate": 9.82217233322983e-06, "loss": 0.3638, "step": 1200 }, { "epoch": 0.5311613164780894, "grad_norm": 0.28931701282310174, "learning_rate": 9.82149182596657e-06, "loss": 0.3575, "step": 1201 }, { "epoch": 0.5316035823535916, "grad_norm": 0.3126841323736245, "learning_rate": 9.82081004277777e-06, "loss": 0.3764, "step": 1202 }, { "epoch": 0.5320458482290937, "grad_norm": 0.2963434281573375, "learning_rate": 9.82012698384385e-06, "loss": 0.3586, "step": 1203 }, { "epoch": 0.5324881141045958, "grad_norm": 0.318234374669819, "learning_rate": 9.819442649345574e-06, "loss": 0.3844, "step": 1204 }, { "epoch": 0.5329303799800981, "grad_norm": 0.32707582343479835, "learning_rate": 9.818757039464038e-06, "loss": 0.3608, "step": 1205 }, { "epoch": 0.5333726458556002, "grad_norm": 0.31864483084300854, "learning_rate": 9.818070154380675e-06, "loss": 0.419, "step": 1206 }, { "epoch": 0.5338149117311023, "grad_norm": 0.3249900567186747, "learning_rate": 9.81738199427726e-06, "loss": 0.3951, "step": 1207 }, { "epoch": 0.5342571776066045, "grad_norm": 0.3289044598905229, "learning_rate": 9.816692559335903e-06, "loss": 0.4054, "step": 1208 }, { "epoch": 0.5346994434821066, "grad_norm": 0.3658065730344045, "learning_rate": 9.816001849739051e-06, "loss": 0.3898, "step": 1209 }, { "epoch": 0.5351417093576089, "grad_norm": 0.24963757966823474, "learning_rate": 9.815309865669487e-06, "loss": 0.4103, "step": 1210 }, { "epoch": 0.535583975233111, "grad_norm": 0.3432464247243486, "learning_rate": 9.814616607310334e-06, "loss": 0.396, "step": 1211 }, { "epoch": 0.5360262411086131, "grad_norm": 0.34379388871098043, "learning_rate": 9.813922074845049e-06, "loss": 0.3914, "step": 1212 }, { "epoch": 0.5364685069841153, "grad_norm": 0.40074751498241085, "learning_rate": 9.81322626845743e-06, "loss": 0.3866, "step": 1213 }, { "epoch": 0.5369107728596174, "grad_norm": 0.31558288777617516, "learning_rate": 9.812529188331612e-06, "loss": 0.3709, "step": 1214 }, { "epoch": 0.5373530387351196, "grad_norm": 0.30134010571957076, "learning_rate": 9.81183083465206e-06, "loss": 0.4207, "step": 1215 }, { "epoch": 0.5377953046106217, "grad_norm": 0.3949669609227973, "learning_rate": 9.811131207603587e-06, "loss": 0.4281, "step": 1216 }, { "epoch": 0.5382375704861239, "grad_norm": 0.3120466752499874, "learning_rate": 9.810430307371335e-06, "loss": 0.3567, "step": 1217 }, { "epoch": 0.5386798363616261, "grad_norm": 0.29696806726906533, "learning_rate": 9.809728134140786e-06, "loss": 0.3882, "step": 1218 }, { "epoch": 0.5391221022371282, "grad_norm": 0.3094012676564128, "learning_rate": 9.809024688097756e-06, "loss": 0.4034, "step": 1219 }, { "epoch": 0.5395643681126304, "grad_norm": 0.37061521037258055, "learning_rate": 9.808319969428404e-06, "loss": 0.3703, "step": 1220 }, { "epoch": 0.5400066339881325, "grad_norm": 0.2960355585251065, "learning_rate": 9.807613978319217e-06, "loss": 0.3643, "step": 1221 }, { "epoch": 0.5404488998636346, "grad_norm": 0.284788627815203, "learning_rate": 9.806906714957027e-06, "loss": 0.3712, "step": 1222 }, { "epoch": 0.5408911657391369, "grad_norm": 0.3125856959520357, "learning_rate": 9.806198179528996e-06, "loss": 0.3805, "step": 1223 }, { "epoch": 0.541333431614639, "grad_norm": 0.2960968450198522, "learning_rate": 9.80548837222263e-06, "loss": 0.3837, "step": 1224 }, { "epoch": 0.5417756974901412, "grad_norm": 0.311933853192585, "learning_rate": 9.804777293225764e-06, "loss": 0.4068, "step": 1225 }, { "epoch": 0.5422179633656433, "grad_norm": 0.29976180402875835, "learning_rate": 9.804064942726574e-06, "loss": 0.4141, "step": 1226 }, { "epoch": 0.5426602292411454, "grad_norm": 0.35564884864426344, "learning_rate": 9.803351320913573e-06, "loss": 0.4065, "step": 1227 }, { "epoch": 0.5431024951166477, "grad_norm": 0.3536213080144482, "learning_rate": 9.802636427975608e-06, "loss": 0.3843, "step": 1228 }, { "epoch": 0.5435447609921498, "grad_norm": 0.35644524421083085, "learning_rate": 9.801920264101859e-06, "loss": 0.3897, "step": 1229 }, { "epoch": 0.543987026867652, "grad_norm": 0.3049868162797628, "learning_rate": 9.801202829481853e-06, "loss": 0.368, "step": 1230 }, { "epoch": 0.5444292927431541, "grad_norm": 0.3390096737065107, "learning_rate": 9.800484124305443e-06, "loss": 0.3725, "step": 1231 }, { "epoch": 0.5448715586186562, "grad_norm": 0.3267033359702651, "learning_rate": 9.799764148762823e-06, "loss": 0.394, "step": 1232 }, { "epoch": 0.5453138244941584, "grad_norm": 0.29291633067524336, "learning_rate": 9.799042903044523e-06, "loss": 0.3694, "step": 1233 }, { "epoch": 0.5457560903696606, "grad_norm": 0.33401780519034135, "learning_rate": 9.798320387341406e-06, "loss": 0.391, "step": 1234 }, { "epoch": 0.5461983562451627, "grad_norm": 0.3434663128249078, "learning_rate": 9.797596601844676e-06, "loss": 0.4186, "step": 1235 }, { "epoch": 0.5466406221206649, "grad_norm": 0.30352849119446657, "learning_rate": 9.79687154674587e-06, "loss": 0.376, "step": 1236 }, { "epoch": 0.547082887996167, "grad_norm": 0.29223144530188144, "learning_rate": 9.79614522223686e-06, "loss": 0.4248, "step": 1237 }, { "epoch": 0.5475251538716692, "grad_norm": 0.29986911977048447, "learning_rate": 9.795417628509857e-06, "loss": 0.404, "step": 1238 }, { "epoch": 0.5479674197471713, "grad_norm": 0.5584456446934694, "learning_rate": 9.794688765757407e-06, "loss": 0.4586, "step": 1239 }, { "epoch": 0.5484096856226734, "grad_norm": 0.33941650984080096, "learning_rate": 9.793958634172388e-06, "loss": 0.3955, "step": 1240 }, { "epoch": 0.5488519514981757, "grad_norm": 0.5106776491407737, "learning_rate": 9.79322723394802e-06, "loss": 0.3725, "step": 1241 }, { "epoch": 0.5492942173736778, "grad_norm": 0.3119550431996564, "learning_rate": 9.792494565277853e-06, "loss": 0.3985, "step": 1242 }, { "epoch": 0.54973648324918, "grad_norm": 0.30623136063488954, "learning_rate": 9.791760628355775e-06, "loss": 0.3745, "step": 1243 }, { "epoch": 0.5501787491246821, "grad_norm": 0.3186638873098573, "learning_rate": 9.791025423376015e-06, "loss": 0.3692, "step": 1244 }, { "epoch": 0.5506210150001842, "grad_norm": 0.29810505448395747, "learning_rate": 9.790288950533127e-06, "loss": 0.3628, "step": 1245 }, { "epoch": 0.5510632808756865, "grad_norm": 0.3566278963195365, "learning_rate": 9.789551210022007e-06, "loss": 0.4491, "step": 1246 }, { "epoch": 0.5515055467511886, "grad_norm": 0.35041274976225273, "learning_rate": 9.788812202037888e-06, "loss": 0.4212, "step": 1247 }, { "epoch": 0.5519478126266908, "grad_norm": 0.371801647140559, "learning_rate": 9.788071926776333e-06, "loss": 0.4221, "step": 1248 }, { "epoch": 0.5523900785021929, "grad_norm": 0.3327427501107837, "learning_rate": 9.787330384433244e-06, "loss": 0.3577, "step": 1249 }, { "epoch": 0.552832344377695, "grad_norm": 0.3373915941922301, "learning_rate": 9.786587575204858e-06, "loss": 0.3639, "step": 1250 }, { "epoch": 0.5532746102531972, "grad_norm": 0.3802579722274637, "learning_rate": 9.785843499287746e-06, "loss": 0.4461, "step": 1251 }, { "epoch": 0.5537168761286994, "grad_norm": 0.3436173303918471, "learning_rate": 9.785098156878818e-06, "loss": 0.3762, "step": 1252 }, { "epoch": 0.5541591420042016, "grad_norm": 0.29108977490403826, "learning_rate": 9.784351548175312e-06, "loss": 0.3642, "step": 1253 }, { "epoch": 0.5546014078797037, "grad_norm": 0.34504410273877756, "learning_rate": 9.783603673374807e-06, "loss": 0.4312, "step": 1254 }, { "epoch": 0.5550436737552058, "grad_norm": 0.37843526669487093, "learning_rate": 9.782854532675214e-06, "loss": 0.4134, "step": 1255 }, { "epoch": 0.555485939630708, "grad_norm": 0.3715843299719479, "learning_rate": 9.782104126274786e-06, "loss": 0.4416, "step": 1256 }, { "epoch": 0.5559282055062101, "grad_norm": 0.3678918182979492, "learning_rate": 9.781352454372096e-06, "loss": 0.4058, "step": 1257 }, { "epoch": 0.5563704713817124, "grad_norm": 0.34949137533760016, "learning_rate": 9.780599517166069e-06, "loss": 0.3963, "step": 1258 }, { "epoch": 0.5568127372572145, "grad_norm": 0.3284281765244415, "learning_rate": 9.779845314855952e-06, "loss": 0.4231, "step": 1259 }, { "epoch": 0.5572550031327166, "grad_norm": 0.32238649644426043, "learning_rate": 9.779089847641335e-06, "loss": 0.3799, "step": 1260 }, { "epoch": 0.5576972690082188, "grad_norm": 0.2500238754918383, "learning_rate": 9.778333115722139e-06, "loss": 0.3602, "step": 1261 }, { "epoch": 0.5581395348837209, "grad_norm": 0.337607160506246, "learning_rate": 9.777575119298618e-06, "loss": 0.4142, "step": 1262 }, { "epoch": 0.558581800759223, "grad_norm": 0.3370686800752674, "learning_rate": 9.776815858571364e-06, "loss": 0.4087, "step": 1263 }, { "epoch": 0.5590240666347253, "grad_norm": 0.30825761878783725, "learning_rate": 9.776055333741303e-06, "loss": 0.4211, "step": 1264 }, { "epoch": 0.5594663325102274, "grad_norm": 0.29915638592141575, "learning_rate": 9.775293545009696e-06, "loss": 0.3868, "step": 1265 }, { "epoch": 0.5599085983857296, "grad_norm": 0.376322344998031, "learning_rate": 9.774530492578136e-06, "loss": 0.4196, "step": 1266 }, { "epoch": 0.5603508642612317, "grad_norm": 0.3704782257466805, "learning_rate": 9.77376617664855e-06, "loss": 0.4181, "step": 1267 }, { "epoch": 0.5607931301367338, "grad_norm": 0.3230900195682697, "learning_rate": 9.773000597423203e-06, "loss": 0.3614, "step": 1268 }, { "epoch": 0.561235396012236, "grad_norm": 0.3832809342677949, "learning_rate": 9.772233755104695e-06, "loss": 0.4128, "step": 1269 }, { "epoch": 0.5616776618877382, "grad_norm": 0.35055843561291417, "learning_rate": 9.771465649895953e-06, "loss": 0.4322, "step": 1270 }, { "epoch": 0.5621199277632404, "grad_norm": 0.31936558289254185, "learning_rate": 9.770696282000245e-06, "loss": 0.3594, "step": 1271 }, { "epoch": 0.5625621936387425, "grad_norm": 0.30159716320432095, "learning_rate": 9.769925651621172e-06, "loss": 0.3702, "step": 1272 }, { "epoch": 0.5630044595142446, "grad_norm": 0.3359730097397114, "learning_rate": 9.769153758962665e-06, "loss": 0.4069, "step": 1273 }, { "epoch": 0.5634467253897468, "grad_norm": 0.3159922039579333, "learning_rate": 9.768380604228998e-06, "loss": 0.3707, "step": 1274 }, { "epoch": 0.5638889912652489, "grad_norm": 0.345254380949788, "learning_rate": 9.76760618762477e-06, "loss": 0.3863, "step": 1275 }, { "epoch": 0.5643312571407512, "grad_norm": 0.45995879839411496, "learning_rate": 9.766830509354914e-06, "loss": 0.4268, "step": 1276 }, { "epoch": 0.5647735230162533, "grad_norm": 0.4112462523830985, "learning_rate": 9.766053569624705e-06, "loss": 0.429, "step": 1277 }, { "epoch": 0.5652157888917554, "grad_norm": 0.35251142089233, "learning_rate": 9.765275368639744e-06, "loss": 0.4345, "step": 1278 }, { "epoch": 0.5656580547672576, "grad_norm": 0.34883659036789444, "learning_rate": 9.76449590660597e-06, "loss": 0.3801, "step": 1279 }, { "epoch": 0.5661003206427597, "grad_norm": 0.38936944212774793, "learning_rate": 9.763715183729653e-06, "loss": 0.4593, "step": 1280 }, { "epoch": 0.566542586518262, "grad_norm": 0.334215333324977, "learning_rate": 9.762933200217399e-06, "loss": 0.3974, "step": 1281 }, { "epoch": 0.5669848523937641, "grad_norm": 0.3148780390638992, "learning_rate": 9.762149956276147e-06, "loss": 0.4038, "step": 1282 }, { "epoch": 0.5674271182692662, "grad_norm": 0.3284728725093526, "learning_rate": 9.761365452113167e-06, "loss": 0.4048, "step": 1283 }, { "epoch": 0.5678693841447684, "grad_norm": 0.32767229129770964, "learning_rate": 9.760579687936067e-06, "loss": 0.4558, "step": 1284 }, { "epoch": 0.5683116500202705, "grad_norm": 0.33271580744005885, "learning_rate": 9.759792663952785e-06, "loss": 0.425, "step": 1285 }, { "epoch": 0.5687539158957726, "grad_norm": 0.33095650298143614, "learning_rate": 9.759004380371595e-06, "loss": 0.4174, "step": 1286 }, { "epoch": 0.5691961817712748, "grad_norm": 0.33180136680855676, "learning_rate": 9.7582148374011e-06, "loss": 0.3904, "step": 1287 }, { "epoch": 0.569638447646777, "grad_norm": 0.35585791949254736, "learning_rate": 9.75742403525024e-06, "loss": 0.4145, "step": 1288 }, { "epoch": 0.5700807135222792, "grad_norm": 0.3033822206689683, "learning_rate": 9.75663197412829e-06, "loss": 0.3868, "step": 1289 }, { "epoch": 0.5705229793977813, "grad_norm": 0.7999936040460194, "learning_rate": 9.755838654244851e-06, "loss": 0.4473, "step": 1290 }, { "epoch": 0.5709652452732834, "grad_norm": 0.31033819853789735, "learning_rate": 9.755044075809865e-06, "loss": 0.3984, "step": 1291 }, { "epoch": 0.5714075111487856, "grad_norm": 0.3225542138748871, "learning_rate": 9.754248239033603e-06, "loss": 0.4016, "step": 1292 }, { "epoch": 0.5718497770242877, "grad_norm": 0.874145356443114, "learning_rate": 9.753451144126668e-06, "loss": 0.4169, "step": 1293 }, { "epoch": 0.57229204289979, "grad_norm": 0.5251562483207832, "learning_rate": 9.7526527913e-06, "loss": 0.4557, "step": 1294 }, { "epoch": 0.5727343087752921, "grad_norm": 0.32684467725006283, "learning_rate": 9.751853180764865e-06, "loss": 0.4246, "step": 1295 }, { "epoch": 0.5731765746507942, "grad_norm": 0.27987025952466227, "learning_rate": 9.751052312732871e-06, "loss": 0.4065, "step": 1296 }, { "epoch": 0.5736188405262964, "grad_norm": 0.36855975701509813, "learning_rate": 9.750250187415955e-06, "loss": 0.3636, "step": 1297 }, { "epoch": 0.5740611064017985, "grad_norm": 0.37513695429997124, "learning_rate": 9.74944680502638e-06, "loss": 0.4201, "step": 1298 }, { "epoch": 0.5745033722773008, "grad_norm": 0.30438777551947915, "learning_rate": 9.748642165776753e-06, "loss": 0.3623, "step": 1299 }, { "epoch": 0.5749456381528029, "grad_norm": 0.30710355983211546, "learning_rate": 9.747836269880005e-06, "loss": 0.3819, "step": 1300 }, { "epoch": 0.575387904028305, "grad_norm": 0.34121460101520706, "learning_rate": 9.747029117549402e-06, "loss": 0.4029, "step": 1301 }, { "epoch": 0.5758301699038072, "grad_norm": 0.28170654722660193, "learning_rate": 9.746220708998547e-06, "loss": 0.3431, "step": 1302 }, { "epoch": 0.5762724357793093, "grad_norm": 0.3907476570095945, "learning_rate": 9.745411044441366e-06, "loss": 0.4377, "step": 1303 }, { "epoch": 0.5767147016548115, "grad_norm": 0.29894933946621016, "learning_rate": 9.744600124092127e-06, "loss": 0.3926, "step": 1304 }, { "epoch": 0.5771569675303136, "grad_norm": 0.31909834225312933, "learning_rate": 9.743787948165426e-06, "loss": 0.3665, "step": 1305 }, { "epoch": 0.5775992334058158, "grad_norm": 0.3542829809352689, "learning_rate": 9.74297451687619e-06, "loss": 0.4265, "step": 1306 }, { "epoch": 0.578041499281318, "grad_norm": 0.368309844624887, "learning_rate": 9.74215983043968e-06, "loss": 0.3911, "step": 1307 }, { "epoch": 0.5784837651568201, "grad_norm": 0.35491142133082815, "learning_rate": 9.74134388907149e-06, "loss": 0.3874, "step": 1308 }, { "epoch": 0.5789260310323223, "grad_norm": 0.2902704909116982, "learning_rate": 9.740526692987541e-06, "loss": 0.3964, "step": 1309 }, { "epoch": 0.5793682969078244, "grad_norm": 0.3226077498788426, "learning_rate": 9.739708242404098e-06, "loss": 0.4031, "step": 1310 }, { "epoch": 0.5798105627833265, "grad_norm": 0.33324365698950975, "learning_rate": 9.738888537537743e-06, "loss": 0.431, "step": 1311 }, { "epoch": 0.5802528286588288, "grad_norm": 0.28220573089535217, "learning_rate": 9.738067578605398e-06, "loss": 0.3784, "step": 1312 }, { "epoch": 0.5806950945343309, "grad_norm": 0.310745327617424, "learning_rate": 9.737245365824318e-06, "loss": 0.4082, "step": 1313 }, { "epoch": 0.581137360409833, "grad_norm": 0.277262326162641, "learning_rate": 9.736421899412085e-06, "loss": 0.3789, "step": 1314 }, { "epoch": 0.5815796262853352, "grad_norm": 0.31366733145879555, "learning_rate": 9.735597179586618e-06, "loss": 0.3783, "step": 1315 }, { "epoch": 0.5820218921608373, "grad_norm": 0.3100840497802662, "learning_rate": 9.734771206566163e-06, "loss": 0.4083, "step": 1316 }, { "epoch": 0.5824641580363396, "grad_norm": 0.2791306602921892, "learning_rate": 9.7339439805693e-06, "loss": 0.3877, "step": 1317 }, { "epoch": 0.5829064239118417, "grad_norm": 0.36779886769395487, "learning_rate": 9.733115501814943e-06, "loss": 0.4312, "step": 1318 }, { "epoch": 0.5833486897873438, "grad_norm": 0.3132974384744281, "learning_rate": 9.73228577052233e-06, "loss": 0.3828, "step": 1319 }, { "epoch": 0.583790955662846, "grad_norm": 0.3125957629486097, "learning_rate": 9.731454786911039e-06, "loss": 0.3491, "step": 1320 }, { "epoch": 0.5842332215383481, "grad_norm": 0.6594492721137817, "learning_rate": 9.730622551200973e-06, "loss": 0.4104, "step": 1321 }, { "epoch": 0.5846754874138503, "grad_norm": 0.3162020776241055, "learning_rate": 9.729789063612372e-06, "loss": 0.3711, "step": 1322 }, { "epoch": 0.5851177532893524, "grad_norm": 0.2945514696107015, "learning_rate": 9.7289543243658e-06, "loss": 0.3765, "step": 1323 }, { "epoch": 0.5855600191648546, "grad_norm": 0.32514987097261155, "learning_rate": 9.728118333682162e-06, "loss": 0.3896, "step": 1324 }, { "epoch": 0.5860022850403568, "grad_norm": 0.3557987703842294, "learning_rate": 9.727281091782683e-06, "loss": 0.3694, "step": 1325 }, { "epoch": 0.5864445509158589, "grad_norm": 0.3795410189129997, "learning_rate": 9.726442598888928e-06, "loss": 0.4008, "step": 1326 }, { "epoch": 0.5868868167913611, "grad_norm": 0.8953549565652625, "learning_rate": 9.725602855222791e-06, "loss": 0.3941, "step": 1327 }, { "epoch": 0.5873290826668632, "grad_norm": 0.3492819398895392, "learning_rate": 9.724761861006493e-06, "loss": 0.3852, "step": 1328 }, { "epoch": 0.5877713485423653, "grad_norm": 0.3428728830601299, "learning_rate": 9.723919616462589e-06, "loss": 0.3673, "step": 1329 }, { "epoch": 0.5882136144178676, "grad_norm": 0.3733728515723948, "learning_rate": 9.723076121813968e-06, "loss": 0.4268, "step": 1330 }, { "epoch": 0.5886558802933697, "grad_norm": 0.37128092140146524, "learning_rate": 9.722231377283841e-06, "loss": 0.4566, "step": 1331 }, { "epoch": 0.5890981461688719, "grad_norm": 0.36977846831915523, "learning_rate": 9.72138538309576e-06, "loss": 0.3795, "step": 1332 }, { "epoch": 0.589540412044374, "grad_norm": 0.3846091919723403, "learning_rate": 9.720538139473602e-06, "loss": 0.4207, "step": 1333 }, { "epoch": 0.5899826779198761, "grad_norm": 0.26443142017290105, "learning_rate": 9.719689646641573e-06, "loss": 0.3706, "step": 1334 }, { "epoch": 0.5904249437953784, "grad_norm": 0.29135245165377405, "learning_rate": 9.718839904824216e-06, "loss": 0.4232, "step": 1335 }, { "epoch": 0.5908672096708805, "grad_norm": 0.40413251792974003, "learning_rate": 9.717988914246397e-06, "loss": 0.43, "step": 1336 }, { "epoch": 0.5913094755463826, "grad_norm": 0.34081843687630997, "learning_rate": 9.717136675133317e-06, "loss": 0.3862, "step": 1337 }, { "epoch": 0.5917517414218848, "grad_norm": 0.3084278214832102, "learning_rate": 9.716283187710508e-06, "loss": 0.391, "step": 1338 }, { "epoch": 0.5921940072973869, "grad_norm": 0.3421945122064139, "learning_rate": 9.71542845220383e-06, "loss": 0.3995, "step": 1339 }, { "epoch": 0.5926362731728891, "grad_norm": 0.2875253506681211, "learning_rate": 9.714572468839476e-06, "loss": 0.4015, "step": 1340 }, { "epoch": 0.5930785390483913, "grad_norm": 0.2916377090423078, "learning_rate": 9.713715237843964e-06, "loss": 0.3976, "step": 1341 }, { "epoch": 0.5935208049238934, "grad_norm": 0.33331874930244254, "learning_rate": 9.712856759444146e-06, "loss": 0.4103, "step": 1342 }, { "epoch": 0.5939630707993956, "grad_norm": 0.3547633833015756, "learning_rate": 9.711997033867206e-06, "loss": 0.3915, "step": 1343 }, { "epoch": 0.5944053366748977, "grad_norm": 0.36670190038074657, "learning_rate": 9.711136061340655e-06, "loss": 0.4009, "step": 1344 }, { "epoch": 0.5948476025503999, "grad_norm": 0.3099326574909379, "learning_rate": 9.710273842092335e-06, "loss": 0.3622, "step": 1345 }, { "epoch": 0.595289868425902, "grad_norm": 0.27861069035734665, "learning_rate": 9.709410376350416e-06, "loss": 0.3588, "step": 1346 }, { "epoch": 0.5957321343014041, "grad_norm": 0.31397565887654333, "learning_rate": 9.7085456643434e-06, "loss": 0.3986, "step": 1347 }, { "epoch": 0.5961744001769064, "grad_norm": 0.37440251972230554, "learning_rate": 9.70767970630012e-06, "loss": 0.3697, "step": 1348 }, { "epoch": 0.5966166660524085, "grad_norm": 0.34053015156417904, "learning_rate": 9.706812502449734e-06, "loss": 0.3928, "step": 1349 }, { "epoch": 0.5970589319279107, "grad_norm": 0.3121892876125679, "learning_rate": 9.705944053021733e-06, "loss": 0.414, "step": 1350 }, { "epoch": 0.5975011978034128, "grad_norm": 0.3078802607489687, "learning_rate": 9.705074358245943e-06, "loss": 0.3723, "step": 1351 }, { "epoch": 0.5979434636789149, "grad_norm": 0.3521422504935196, "learning_rate": 9.704203418352508e-06, "loss": 0.3936, "step": 1352 }, { "epoch": 0.5983857295544172, "grad_norm": 0.32817125435645084, "learning_rate": 9.703331233571907e-06, "loss": 0.3807, "step": 1353 }, { "epoch": 0.5988279954299193, "grad_norm": 0.28261505580088436, "learning_rate": 9.702457804134953e-06, "loss": 0.3633, "step": 1354 }, { "epoch": 0.5992702613054215, "grad_norm": 0.3221781383538362, "learning_rate": 9.70158313027278e-06, "loss": 0.3745, "step": 1355 }, { "epoch": 0.5997125271809236, "grad_norm": 0.3033830940611865, "learning_rate": 9.700707212216857e-06, "loss": 0.3749, "step": 1356 }, { "epoch": 0.6001547930564257, "grad_norm": 0.3078136862084654, "learning_rate": 9.699830050198984e-06, "loss": 0.3759, "step": 1357 }, { "epoch": 0.6005970589319279, "grad_norm": 0.36237977152687256, "learning_rate": 9.698951644451283e-06, "loss": 0.3829, "step": 1358 }, { "epoch": 0.60103932480743, "grad_norm": 0.3490762096149056, "learning_rate": 9.698071995206212e-06, "loss": 0.4242, "step": 1359 }, { "epoch": 0.6014815906829323, "grad_norm": 0.3423780895307991, "learning_rate": 9.697191102696551e-06, "loss": 0.3825, "step": 1360 }, { "epoch": 0.6019238565584344, "grad_norm": 0.32483806181208424, "learning_rate": 9.696308967155415e-06, "loss": 0.404, "step": 1361 }, { "epoch": 0.6023661224339365, "grad_norm": 0.29045593833789257, "learning_rate": 9.695425588816248e-06, "loss": 0.3936, "step": 1362 }, { "epoch": 0.6028083883094387, "grad_norm": 0.28518416631696913, "learning_rate": 9.69454096791282e-06, "loss": 0.3544, "step": 1363 }, { "epoch": 0.6032506541849408, "grad_norm": 0.30304612196326014, "learning_rate": 9.693655104679232e-06, "loss": 0.3923, "step": 1364 }, { "epoch": 0.603692920060443, "grad_norm": 0.3045125282061282, "learning_rate": 9.692767999349908e-06, "loss": 0.3805, "step": 1365 }, { "epoch": 0.6041351859359452, "grad_norm": 0.31590324688188554, "learning_rate": 9.69187965215961e-06, "loss": 0.3835, "step": 1366 }, { "epoch": 0.6045774518114473, "grad_norm": 0.372057979111024, "learning_rate": 9.690990063343423e-06, "loss": 0.4018, "step": 1367 }, { "epoch": 0.6050197176869495, "grad_norm": 0.30088573198920365, "learning_rate": 9.69009923313676e-06, "loss": 0.3785, "step": 1368 }, { "epoch": 0.6054619835624516, "grad_norm": 0.3275967886394411, "learning_rate": 9.689207161775366e-06, "loss": 0.3994, "step": 1369 }, { "epoch": 0.6059042494379537, "grad_norm": 0.3211254630195027, "learning_rate": 9.68831384949531e-06, "loss": 0.4175, "step": 1370 }, { "epoch": 0.606346515313456, "grad_norm": 0.3771126127014851, "learning_rate": 9.687419296532994e-06, "loss": 0.4312, "step": 1371 }, { "epoch": 0.6067887811889581, "grad_norm": 0.4731145500532118, "learning_rate": 9.686523503125145e-06, "loss": 0.3931, "step": 1372 }, { "epoch": 0.6072310470644603, "grad_norm": 0.2890132165605816, "learning_rate": 9.68562646950882e-06, "loss": 0.3663, "step": 1373 }, { "epoch": 0.6076733129399624, "grad_norm": 0.3124505221425591, "learning_rate": 9.684728195921405e-06, "loss": 0.3805, "step": 1374 }, { "epoch": 0.6081155788154645, "grad_norm": 0.34611254818990833, "learning_rate": 9.683828682600608e-06, "loss": 0.4169, "step": 1375 }, { "epoch": 0.6085578446909667, "grad_norm": 0.36878082110074833, "learning_rate": 9.682927929784475e-06, "loss": 0.3902, "step": 1376 }, { "epoch": 0.6090001105664689, "grad_norm": 0.34188365073917215, "learning_rate": 9.682025937711373e-06, "loss": 0.4141, "step": 1377 }, { "epoch": 0.6094423764419711, "grad_norm": 0.3149688280582913, "learning_rate": 9.681122706619999e-06, "loss": 0.3895, "step": 1378 }, { "epoch": 0.6098846423174732, "grad_norm": 0.2962439448778595, "learning_rate": 9.680218236749377e-06, "loss": 0.4001, "step": 1379 }, { "epoch": 0.6103269081929753, "grad_norm": 0.30734560315499276, "learning_rate": 9.67931252833886e-06, "loss": 0.3517, "step": 1380 }, { "epoch": 0.6107691740684775, "grad_norm": 0.3197918511976623, "learning_rate": 9.678405581628128e-06, "loss": 0.3795, "step": 1381 }, { "epoch": 0.6112114399439796, "grad_norm": 0.37105063259029386, "learning_rate": 9.67749739685719e-06, "loss": 0.4313, "step": 1382 }, { "epoch": 0.6116537058194819, "grad_norm": 0.34228340203661434, "learning_rate": 9.67658797426638e-06, "loss": 0.4119, "step": 1383 }, { "epoch": 0.612095971694984, "grad_norm": 0.3038516353430877, "learning_rate": 9.675677314096363e-06, "loss": 0.3424, "step": 1384 }, { "epoch": 0.6125382375704861, "grad_norm": 0.3274162000995715, "learning_rate": 9.674765416588127e-06, "loss": 0.4267, "step": 1385 }, { "epoch": 0.6129805034459883, "grad_norm": 0.345744581559566, "learning_rate": 9.673852281982993e-06, "loss": 0.3969, "step": 1386 }, { "epoch": 0.6134227693214904, "grad_norm": 0.2944893971856562, "learning_rate": 9.672937910522606e-06, "loss": 0.4011, "step": 1387 }, { "epoch": 0.6138650351969925, "grad_norm": 0.3165729093478836, "learning_rate": 9.672022302448938e-06, "loss": 0.4048, "step": 1388 }, { "epoch": 0.6143073010724948, "grad_norm": 0.3259100076273786, "learning_rate": 9.671105458004289e-06, "loss": 0.4157, "step": 1389 }, { "epoch": 0.6147495669479969, "grad_norm": 0.27724198609272255, "learning_rate": 9.670187377431285e-06, "loss": 0.3834, "step": 1390 }, { "epoch": 0.6151918328234991, "grad_norm": 0.31515605971472205, "learning_rate": 9.669268060972882e-06, "loss": 0.3811, "step": 1391 }, { "epoch": 0.6156340986990012, "grad_norm": 0.36417983567840556, "learning_rate": 9.668347508872362e-06, "loss": 0.4482, "step": 1392 }, { "epoch": 0.6160763645745033, "grad_norm": 0.3305234674628196, "learning_rate": 9.667425721373333e-06, "loss": 0.353, "step": 1393 }, { "epoch": 0.6165186304500055, "grad_norm": 0.29769472302635913, "learning_rate": 9.666502698719728e-06, "loss": 0.3942, "step": 1394 }, { "epoch": 0.6169608963255077, "grad_norm": 0.36402593232926966, "learning_rate": 9.665578441155814e-06, "loss": 0.4107, "step": 1395 }, { "epoch": 0.6174031622010099, "grad_norm": 0.3118721612813891, "learning_rate": 9.664652948926174e-06, "loss": 0.3935, "step": 1396 }, { "epoch": 0.617845428076512, "grad_norm": 0.34062297092790117, "learning_rate": 9.66372622227573e-06, "loss": 0.4136, "step": 1397 }, { "epoch": 0.6182876939520141, "grad_norm": 0.29853628730413917, "learning_rate": 9.662798261449717e-06, "loss": 0.4211, "step": 1398 }, { "epoch": 0.6187299598275163, "grad_norm": 0.35696624676927596, "learning_rate": 9.661869066693712e-06, "loss": 0.3826, "step": 1399 }, { "epoch": 0.6191722257030184, "grad_norm": 0.3459739166095715, "learning_rate": 9.660938638253605e-06, "loss": 0.395, "step": 1400 }, { "epoch": 0.6196144915785207, "grad_norm": 0.36998960984270035, "learning_rate": 9.66000697637562e-06, "loss": 0.4281, "step": 1401 }, { "epoch": 0.6200567574540228, "grad_norm": 0.3269488528880725, "learning_rate": 9.659074081306306e-06, "loss": 0.3926, "step": 1402 }, { "epoch": 0.6204990233295249, "grad_norm": 0.32154599618296004, "learning_rate": 9.658139953292536e-06, "loss": 0.3914, "step": 1403 }, { "epoch": 0.6209412892050271, "grad_norm": 0.3315917506076342, "learning_rate": 9.657204592581513e-06, "loss": 0.4407, "step": 1404 }, { "epoch": 0.6213835550805292, "grad_norm": 0.33525123804574164, "learning_rate": 9.65626799942076e-06, "loss": 0.3529, "step": 1405 }, { "epoch": 0.6218258209560314, "grad_norm": 0.30684576699485394, "learning_rate": 9.655330174058138e-06, "loss": 0.3933, "step": 1406 }, { "epoch": 0.6222680868315336, "grad_norm": 0.2896668830257268, "learning_rate": 9.654391116741819e-06, "loss": 0.3943, "step": 1407 }, { "epoch": 0.6227103527070357, "grad_norm": 0.37180395890873225, "learning_rate": 9.653450827720313e-06, "loss": 0.3662, "step": 1408 }, { "epoch": 0.6231526185825379, "grad_norm": 0.30697449868494886, "learning_rate": 9.65250930724245e-06, "loss": 0.4027, "step": 1409 }, { "epoch": 0.62359488445804, "grad_norm": 0.3207202061554816, "learning_rate": 9.651566555557388e-06, "loss": 0.4143, "step": 1410 }, { "epoch": 0.6240371503335422, "grad_norm": 0.32367681998981085, "learning_rate": 9.650622572914607e-06, "loss": 0.3916, "step": 1411 }, { "epoch": 0.6244794162090443, "grad_norm": 0.3076977207582505, "learning_rate": 9.64967735956392e-06, "loss": 0.414, "step": 1412 }, { "epoch": 0.6249216820845465, "grad_norm": 0.3016220003932714, "learning_rate": 9.648730915755461e-06, "loss": 0.3693, "step": 1413 }, { "epoch": 0.6253639479600487, "grad_norm": 0.31331178328269066, "learning_rate": 9.647783241739687e-06, "loss": 0.381, "step": 1414 }, { "epoch": 0.6258062138355508, "grad_norm": 0.3204150527372771, "learning_rate": 9.646834337767388e-06, "loss": 0.3768, "step": 1415 }, { "epoch": 0.6262484797110529, "grad_norm": 0.41180134177498484, "learning_rate": 9.645884204089673e-06, "loss": 0.3813, "step": 1416 }, { "epoch": 0.6266907455865551, "grad_norm": 0.33366643480657737, "learning_rate": 9.644932840957978e-06, "loss": 0.4045, "step": 1417 }, { "epoch": 0.6271330114620572, "grad_norm": 0.33668119065541924, "learning_rate": 9.643980248624067e-06, "loss": 0.385, "step": 1418 }, { "epoch": 0.6275752773375595, "grad_norm": 0.3239954123725251, "learning_rate": 9.643026427340025e-06, "loss": 0.357, "step": 1419 }, { "epoch": 0.6280175432130616, "grad_norm": 0.38328939862076117, "learning_rate": 9.642071377358266e-06, "loss": 0.4015, "step": 1420 }, { "epoch": 0.6284598090885637, "grad_norm": 0.3835320635707049, "learning_rate": 9.641115098931528e-06, "loss": 0.4149, "step": 1421 }, { "epoch": 0.6289020749640659, "grad_norm": 0.3237767435173494, "learning_rate": 9.640157592312873e-06, "loss": 0.3836, "step": 1422 }, { "epoch": 0.629344340839568, "grad_norm": 0.347735261691476, "learning_rate": 9.639198857755687e-06, "loss": 0.3876, "step": 1423 }, { "epoch": 0.6297866067150703, "grad_norm": 0.3189698571627338, "learning_rate": 9.638238895513687e-06, "loss": 0.4285, "step": 1424 }, { "epoch": 0.6302288725905724, "grad_norm": 0.30379127815509965, "learning_rate": 9.637277705840909e-06, "loss": 0.4006, "step": 1425 }, { "epoch": 0.6306711384660745, "grad_norm": 0.3424654043806267, "learning_rate": 9.636315288991713e-06, "loss": 0.455, "step": 1426 }, { "epoch": 0.6311134043415767, "grad_norm": 0.33504028775988665, "learning_rate": 9.635351645220787e-06, "loss": 0.3767, "step": 1427 }, { "epoch": 0.6315556702170788, "grad_norm": 0.31376480806875273, "learning_rate": 9.634386774783146e-06, "loss": 0.3914, "step": 1428 }, { "epoch": 0.631997936092581, "grad_norm": 0.27977629997727776, "learning_rate": 9.633420677934125e-06, "loss": 0.3973, "step": 1429 }, { "epoch": 0.6324402019680831, "grad_norm": 0.2881974573005097, "learning_rate": 9.632453354929381e-06, "loss": 0.3572, "step": 1430 }, { "epoch": 0.6328824678435853, "grad_norm": 0.29086530263700394, "learning_rate": 9.631484806024906e-06, "loss": 0.3837, "step": 1431 }, { "epoch": 0.6333247337190875, "grad_norm": 0.36138172125317247, "learning_rate": 9.630515031477006e-06, "loss": 0.3756, "step": 1432 }, { "epoch": 0.6337669995945896, "grad_norm": 0.4212522484996528, "learning_rate": 9.629544031542315e-06, "loss": 0.4046, "step": 1433 }, { "epoch": 0.6342092654700918, "grad_norm": 0.5754947115509093, "learning_rate": 9.628571806477794e-06, "loss": 0.3988, "step": 1434 }, { "epoch": 0.6346515313455939, "grad_norm": 0.28331556953617665, "learning_rate": 9.627598356540723e-06, "loss": 0.3595, "step": 1435 }, { "epoch": 0.635093797221096, "grad_norm": 0.3485262796197907, "learning_rate": 9.626623681988715e-06, "loss": 0.4049, "step": 1436 }, { "epoch": 0.6355360630965983, "grad_norm": 0.3199384039616983, "learning_rate": 9.625647783079691e-06, "loss": 0.4072, "step": 1437 }, { "epoch": 0.6359783289721004, "grad_norm": 0.30652018094593264, "learning_rate": 9.624670660071912e-06, "loss": 0.3675, "step": 1438 }, { "epoch": 0.6364205948476026, "grad_norm": 1.6054098915832395, "learning_rate": 9.623692313223958e-06, "loss": 0.3971, "step": 1439 }, { "epoch": 0.6368628607231047, "grad_norm": 0.36515562255769735, "learning_rate": 9.622712742794731e-06, "loss": 0.4021, "step": 1440 }, { "epoch": 0.6373051265986068, "grad_norm": 0.3393859389586723, "learning_rate": 9.621731949043456e-06, "loss": 0.3605, "step": 1441 }, { "epoch": 0.637747392474109, "grad_norm": 1.7105037930765452, "learning_rate": 9.620749932229683e-06, "loss": 0.4599, "step": 1442 }, { "epoch": 0.6381896583496112, "grad_norm": 0.31948468667229596, "learning_rate": 9.619766692613289e-06, "loss": 0.3821, "step": 1443 }, { "epoch": 0.6386319242251133, "grad_norm": 0.4959488932562768, "learning_rate": 9.618782230454466e-06, "loss": 0.345, "step": 1444 }, { "epoch": 0.6390741901006155, "grad_norm": 0.3016863479530887, "learning_rate": 9.617796546013742e-06, "loss": 0.3926, "step": 1445 }, { "epoch": 0.6395164559761176, "grad_norm": 0.6292033378587916, "learning_rate": 9.616809639551958e-06, "loss": 0.4011, "step": 1446 }, { "epoch": 0.6399587218516198, "grad_norm": 0.475334317192781, "learning_rate": 9.615821511330281e-06, "loss": 0.3527, "step": 1447 }, { "epoch": 0.640400987727122, "grad_norm": 0.34920640839130535, "learning_rate": 9.614832161610204e-06, "loss": 0.3904, "step": 1448 }, { "epoch": 0.6408432536026241, "grad_norm": 0.3096362930174555, "learning_rate": 9.613841590653541e-06, "loss": 0.3817, "step": 1449 }, { "epoch": 0.6412855194781263, "grad_norm": 0.2814405824402582, "learning_rate": 9.61284979872243e-06, "loss": 0.3611, "step": 1450 }, { "epoch": 0.6417277853536284, "grad_norm": 0.8828177790931324, "learning_rate": 9.611856786079331e-06, "loss": 0.3884, "step": 1451 }, { "epoch": 0.6421700512291306, "grad_norm": 0.3359610097154349, "learning_rate": 9.610862552987029e-06, "loss": 0.3759, "step": 1452 }, { "epoch": 0.6426123171046327, "grad_norm": 0.3746440607349514, "learning_rate": 9.60986709970863e-06, "loss": 0.4187, "step": 1453 }, { "epoch": 0.6430545829801348, "grad_norm": 0.34819453091578434, "learning_rate": 9.608870426507564e-06, "loss": 0.4173, "step": 1454 }, { "epoch": 0.6434968488556371, "grad_norm": 0.3407272323577818, "learning_rate": 9.607872533647584e-06, "loss": 0.3957, "step": 1455 }, { "epoch": 0.6439391147311392, "grad_norm": 0.3273616734477944, "learning_rate": 9.606873421392764e-06, "loss": 0.4019, "step": 1456 }, { "epoch": 0.6443813806066414, "grad_norm": 0.287959311538723, "learning_rate": 9.605873090007503e-06, "loss": 0.3946, "step": 1457 }, { "epoch": 0.6448236464821435, "grad_norm": 0.3105549762355174, "learning_rate": 9.604871539756522e-06, "loss": 0.3961, "step": 1458 }, { "epoch": 0.6452659123576456, "grad_norm": 0.3553512774844773, "learning_rate": 9.603868770904862e-06, "loss": 0.4126, "step": 1459 }, { "epoch": 0.6457081782331479, "grad_norm": 0.3641804090202093, "learning_rate": 9.602864783717894e-06, "loss": 0.3587, "step": 1460 }, { "epoch": 0.64615044410865, "grad_norm": 0.34321610540629754, "learning_rate": 9.601859578461298e-06, "loss": 0.3741, "step": 1461 }, { "epoch": 0.6465927099841522, "grad_norm": 0.2911985972592254, "learning_rate": 9.600853155401092e-06, "loss": 0.3738, "step": 1462 }, { "epoch": 0.6470349758596543, "grad_norm": 0.7638654007527621, "learning_rate": 9.599845514803605e-06, "loss": 0.3871, "step": 1463 }, { "epoch": 0.6474772417351564, "grad_norm": 0.32158679747775704, "learning_rate": 9.598836656935491e-06, "loss": 0.3777, "step": 1464 }, { "epoch": 0.6479195076106586, "grad_norm": 0.3383237703676603, "learning_rate": 9.597826582063729e-06, "loss": 0.3807, "step": 1465 }, { "epoch": 0.6483617734861608, "grad_norm": 0.3190505203976119, "learning_rate": 9.59681529045562e-06, "loss": 0.4181, "step": 1466 }, { "epoch": 0.6488040393616629, "grad_norm": 0.33651359429715055, "learning_rate": 9.59580278237878e-06, "loss": 0.3901, "step": 1467 }, { "epoch": 0.6492463052371651, "grad_norm": 0.3707254993598279, "learning_rate": 9.594789058101154e-06, "loss": 0.4048, "step": 1468 }, { "epoch": 0.6496885711126672, "grad_norm": 0.3280522383448747, "learning_rate": 9.593774117891008e-06, "loss": 0.4062, "step": 1469 }, { "epoch": 0.6501308369881694, "grad_norm": 0.3364081272365921, "learning_rate": 9.59275796201693e-06, "loss": 0.3469, "step": 1470 }, { "epoch": 0.6505731028636715, "grad_norm": 0.3014277525541911, "learning_rate": 9.591740590747823e-06, "loss": 0.3823, "step": 1471 }, { "epoch": 0.6510153687391736, "grad_norm": 0.3731140907631425, "learning_rate": 9.59072200435292e-06, "loss": 0.3686, "step": 1472 }, { "epoch": 0.6514576346146759, "grad_norm": 0.31327905114079285, "learning_rate": 9.589702203101776e-06, "loss": 0.3766, "step": 1473 }, { "epoch": 0.651899900490178, "grad_norm": 0.3534356077089583, "learning_rate": 9.58868118726426e-06, "loss": 0.3845, "step": 1474 }, { "epoch": 0.6523421663656802, "grad_norm": 0.35442432261897827, "learning_rate": 9.587658957110566e-06, "loss": 0.3947, "step": 1475 }, { "epoch": 0.6527844322411823, "grad_norm": 0.3487457431734744, "learning_rate": 9.586635512911213e-06, "loss": 0.3741, "step": 1476 }, { "epoch": 0.6532266981166844, "grad_norm": 0.2714663519491272, "learning_rate": 9.585610854937034e-06, "loss": 0.3368, "step": 1477 }, { "epoch": 0.6536689639921867, "grad_norm": 0.3304810816362886, "learning_rate": 9.58458498345919e-06, "loss": 0.4397, "step": 1478 }, { "epoch": 0.6541112298676888, "grad_norm": 0.3379371075895768, "learning_rate": 9.583557898749159e-06, "loss": 0.4261, "step": 1479 }, { "epoch": 0.654553495743191, "grad_norm": 0.34954782944074353, "learning_rate": 9.582529601078743e-06, "loss": 0.3758, "step": 1480 }, { "epoch": 0.6549957616186931, "grad_norm": 0.32307550299056487, "learning_rate": 9.581500090720062e-06, "loss": 0.3603, "step": 1481 }, { "epoch": 0.6554380274941952, "grad_norm": 0.3896412691369583, "learning_rate": 9.580469367945559e-06, "loss": 0.4291, "step": 1482 }, { "epoch": 0.6558802933696974, "grad_norm": 0.33304407776149103, "learning_rate": 9.579437433027996e-06, "loss": 0.4383, "step": 1483 }, { "epoch": 0.6563225592451996, "grad_norm": 0.2938084644066865, "learning_rate": 9.578404286240458e-06, "loss": 0.3988, "step": 1484 }, { "epoch": 0.6567648251207018, "grad_norm": 0.34013109861343366, "learning_rate": 9.577369927856353e-06, "loss": 0.4267, "step": 1485 }, { "epoch": 0.6572070909962039, "grad_norm": 0.3252693358430626, "learning_rate": 9.5763343581494e-06, "loss": 0.4181, "step": 1486 }, { "epoch": 0.657649356871706, "grad_norm": 0.3649961359493179, "learning_rate": 9.575297577393647e-06, "loss": 0.4215, "step": 1487 }, { "epoch": 0.6580916227472082, "grad_norm": 0.30982902747136803, "learning_rate": 9.574259585863464e-06, "loss": 0.3741, "step": 1488 }, { "epoch": 0.6585338886227103, "grad_norm": 0.2977227295544947, "learning_rate": 9.573220383833532e-06, "loss": 0.3771, "step": 1489 }, { "epoch": 0.6589761544982126, "grad_norm": 0.32899293040149274, "learning_rate": 9.572179971578863e-06, "loss": 0.3737, "step": 1490 }, { "epoch": 0.6594184203737147, "grad_norm": 0.2976142758687593, "learning_rate": 9.571138349374783e-06, "loss": 0.3894, "step": 1491 }, { "epoch": 0.6598606862492168, "grad_norm": 0.4353733981756302, "learning_rate": 9.570095517496937e-06, "loss": 0.3908, "step": 1492 }, { "epoch": 0.660302952124719, "grad_norm": 3.5974228096535854, "learning_rate": 9.569051476221295e-06, "loss": 0.3798, "step": 1493 }, { "epoch": 0.6607452180002211, "grad_norm": 0.3289575620155943, "learning_rate": 9.568006225824144e-06, "loss": 0.3978, "step": 1494 }, { "epoch": 0.6611874838757232, "grad_norm": 0.3312364589768437, "learning_rate": 9.566959766582093e-06, "loss": 0.3592, "step": 1495 }, { "epoch": 0.6616297497512255, "grad_norm": 0.30998142933945494, "learning_rate": 9.565912098772067e-06, "loss": 0.3754, "step": 1496 }, { "epoch": 0.6620720156267276, "grad_norm": 0.29953836900539704, "learning_rate": 9.564863222671317e-06, "loss": 0.3673, "step": 1497 }, { "epoch": 0.6625142815022298, "grad_norm": 0.3432867654060151, "learning_rate": 9.563813138557408e-06, "loss": 0.4101, "step": 1498 }, { "epoch": 0.6629565473777319, "grad_norm": 0.32493635428178713, "learning_rate": 9.562761846708225e-06, "loss": 0.3927, "step": 1499 }, { "epoch": 0.663398813253234, "grad_norm": 0.2818150354665978, "learning_rate": 9.561709347401978e-06, "loss": 0.349, "step": 1500 }, { "epoch": 0.6638410791287362, "grad_norm": 3.0942453841252866, "learning_rate": 9.56065564091719e-06, "loss": 0.3766, "step": 1501 }, { "epoch": 0.6642833450042384, "grad_norm": 0.36242862032379186, "learning_rate": 9.559600727532708e-06, "loss": 0.4027, "step": 1502 }, { "epoch": 0.6647256108797406, "grad_norm": 0.409486732388478, "learning_rate": 9.558544607527697e-06, "loss": 0.4218, "step": 1503 }, { "epoch": 0.6651678767552427, "grad_norm": 0.31599817830655835, "learning_rate": 9.55748728118164e-06, "loss": 0.3775, "step": 1504 }, { "epoch": 0.6656101426307448, "grad_norm": 0.38736674750242517, "learning_rate": 9.556428748774344e-06, "loss": 0.3874, "step": 1505 }, { "epoch": 0.666052408506247, "grad_norm": 0.30321356033518676, "learning_rate": 9.555369010585927e-06, "loss": 0.3745, "step": 1506 }, { "epoch": 0.6664946743817491, "grad_norm": 0.33829920843513234, "learning_rate": 9.554308066896832e-06, "loss": 0.4279, "step": 1507 }, { "epoch": 0.6669369402572514, "grad_norm": 0.3816492084625608, "learning_rate": 9.553245917987818e-06, "loss": 0.3911, "step": 1508 }, { "epoch": 0.6673792061327535, "grad_norm": 0.4019417075167777, "learning_rate": 9.55218256413997e-06, "loss": 0.4455, "step": 1509 }, { "epoch": 0.6678214720082556, "grad_norm": 0.32343622559766055, "learning_rate": 9.55111800563468e-06, "loss": 0.391, "step": 1510 }, { "epoch": 0.6682637378837578, "grad_norm": 0.3160581103624194, "learning_rate": 9.550052242753672e-06, "loss": 0.3629, "step": 1511 }, { "epoch": 0.6687060037592599, "grad_norm": 0.3156832264998888, "learning_rate": 9.548985275778974e-06, "loss": 0.395, "step": 1512 }, { "epoch": 0.6691482696347621, "grad_norm": 0.3040484445289212, "learning_rate": 9.547917104992948e-06, "loss": 0.3806, "step": 1513 }, { "epoch": 0.6695905355102643, "grad_norm": 0.3645896380078816, "learning_rate": 9.546847730678262e-06, "loss": 0.4269, "step": 1514 }, { "epoch": 0.6700328013857664, "grad_norm": 0.32571832887089225, "learning_rate": 9.545777153117912e-06, "loss": 0.3677, "step": 1515 }, { "epoch": 0.6704750672612686, "grad_norm": 0.352677689032271, "learning_rate": 9.544705372595203e-06, "loss": 0.4078, "step": 1516 }, { "epoch": 0.6709173331367707, "grad_norm": 0.32101307877962537, "learning_rate": 9.543632389393767e-06, "loss": 0.3836, "step": 1517 }, { "epoch": 0.6713595990122728, "grad_norm": 0.3002241951192633, "learning_rate": 9.542558203797551e-06, "loss": 0.3687, "step": 1518 }, { "epoch": 0.671801864887775, "grad_norm": 0.3444738405078688, "learning_rate": 9.541482816090818e-06, "loss": 0.4133, "step": 1519 }, { "epoch": 0.6722441307632772, "grad_norm": 0.336091426631974, "learning_rate": 9.540406226558153e-06, "loss": 0.368, "step": 1520 }, { "epoch": 0.6726863966387794, "grad_norm": 0.32781610993274174, "learning_rate": 9.539328435484453e-06, "loss": 0.3647, "step": 1521 }, { "epoch": 0.6731286625142815, "grad_norm": 0.298849067355614, "learning_rate": 9.538249443154941e-06, "loss": 0.3645, "step": 1522 }, { "epoch": 0.6735709283897836, "grad_norm": 0.3469443309447057, "learning_rate": 9.537169249855152e-06, "loss": 0.3787, "step": 1523 }, { "epoch": 0.6740131942652858, "grad_norm": 0.30851018036348166, "learning_rate": 9.536087855870943e-06, "loss": 0.4256, "step": 1524 }, { "epoch": 0.6744554601407879, "grad_norm": 0.27338091839167505, "learning_rate": 9.535005261488484e-06, "loss": 0.3895, "step": 1525 }, { "epoch": 0.6748977260162902, "grad_norm": 0.3123932456897019, "learning_rate": 9.533921466994268e-06, "loss": 0.3783, "step": 1526 }, { "epoch": 0.6753399918917923, "grad_norm": 0.3104994752097023, "learning_rate": 9.532836472675098e-06, "loss": 0.3833, "step": 1527 }, { "epoch": 0.6757822577672944, "grad_norm": 0.36935438961678607, "learning_rate": 9.531750278818102e-06, "loss": 0.4219, "step": 1528 }, { "epoch": 0.6762245236427966, "grad_norm": 0.29618377642029564, "learning_rate": 9.530662885710724e-06, "loss": 0.3636, "step": 1529 }, { "epoch": 0.6766667895182987, "grad_norm": 0.354466988635226, "learning_rate": 9.529574293640724e-06, "loss": 0.401, "step": 1530 }, { "epoch": 0.677109055393801, "grad_norm": 0.3240169478267007, "learning_rate": 9.528484502896176e-06, "loss": 0.3819, "step": 1531 }, { "epoch": 0.6775513212693031, "grad_norm": 0.2975797810081774, "learning_rate": 9.527393513765476e-06, "loss": 0.3764, "step": 1532 }, { "epoch": 0.6779935871448052, "grad_norm": 0.38811190519690847, "learning_rate": 9.526301326537337e-06, "loss": 0.4084, "step": 1533 }, { "epoch": 0.6784358530203074, "grad_norm": 0.2774262179081021, "learning_rate": 9.525207941500788e-06, "loss": 0.3373, "step": 1534 }, { "epoch": 0.6788781188958095, "grad_norm": 0.33084409253335473, "learning_rate": 9.524113358945173e-06, "loss": 0.3849, "step": 1535 }, { "epoch": 0.6793203847713117, "grad_norm": 0.32006445931005684, "learning_rate": 9.523017579160155e-06, "loss": 0.3842, "step": 1536 }, { "epoch": 0.6797626506468138, "grad_norm": 0.2624314887306127, "learning_rate": 9.521920602435713e-06, "loss": 0.363, "step": 1537 }, { "epoch": 0.680204916522316, "grad_norm": 0.2911496286814295, "learning_rate": 9.520822429062145e-06, "loss": 0.3499, "step": 1538 }, { "epoch": 0.6806471823978182, "grad_norm": 0.43773163067170917, "learning_rate": 9.519723059330061e-06, "loss": 0.3876, "step": 1539 }, { "epoch": 0.6810894482733203, "grad_norm": 0.2825934166510494, "learning_rate": 9.518622493530392e-06, "loss": 0.3824, "step": 1540 }, { "epoch": 0.6815317141488225, "grad_norm": 0.299624264187492, "learning_rate": 9.517520731954387e-06, "loss": 0.3836, "step": 1541 }, { "epoch": 0.6819739800243246, "grad_norm": 0.3429186192683476, "learning_rate": 9.516417774893603e-06, "loss": 0.3792, "step": 1542 }, { "epoch": 0.6824162458998267, "grad_norm": 0.31547084446021845, "learning_rate": 9.51531362263992e-06, "loss": 0.3439, "step": 1543 }, { "epoch": 0.682858511775329, "grad_norm": 0.3242642447986855, "learning_rate": 9.514208275485536e-06, "loss": 0.4163, "step": 1544 }, { "epoch": 0.6833007776508311, "grad_norm": 0.28286321295704764, "learning_rate": 9.513101733722961e-06, "loss": 0.3493, "step": 1545 }, { "epoch": 0.6837430435263332, "grad_norm": 0.2722606165584269, "learning_rate": 9.511993997645021e-06, "loss": 0.3722, "step": 1546 }, { "epoch": 0.6841853094018354, "grad_norm": 0.38063017416048983, "learning_rate": 9.51088506754486e-06, "loss": 0.4354, "step": 1547 }, { "epoch": 0.6846275752773375, "grad_norm": 0.30900011283765894, "learning_rate": 9.50977494371594e-06, "loss": 0.3797, "step": 1548 }, { "epoch": 0.6850698411528398, "grad_norm": 0.3764974582038516, "learning_rate": 9.508663626452032e-06, "loss": 0.4152, "step": 1549 }, { "epoch": 0.6855121070283419, "grad_norm": 0.26426250044933325, "learning_rate": 9.507551116047228e-06, "loss": 0.3759, "step": 1550 }, { "epoch": 0.685954372903844, "grad_norm": 0.2453267835900213, "learning_rate": 9.506437412795939e-06, "loss": 0.3729, "step": 1551 }, { "epoch": 0.6863966387793462, "grad_norm": 0.32423896760860876, "learning_rate": 9.505322516992884e-06, "loss": 0.3858, "step": 1552 }, { "epoch": 0.6868389046548483, "grad_norm": 0.27315607681055604, "learning_rate": 9.504206428933101e-06, "loss": 0.3907, "step": 1553 }, { "epoch": 0.6872811705303505, "grad_norm": 0.2949226532542921, "learning_rate": 9.503089148911947e-06, "loss": 0.3994, "step": 1554 }, { "epoch": 0.6877234364058527, "grad_norm": 0.28360029697939126, "learning_rate": 9.501970677225086e-06, "loss": 0.3979, "step": 1555 }, { "epoch": 0.6881657022813548, "grad_norm": 0.2835744540987448, "learning_rate": 9.500851014168505e-06, "loss": 0.3842, "step": 1556 }, { "epoch": 0.688607968156857, "grad_norm": 0.2866658575340847, "learning_rate": 9.499730160038505e-06, "loss": 0.3731, "step": 1557 }, { "epoch": 0.6890502340323591, "grad_norm": 0.3839332778583097, "learning_rate": 9.498608115131699e-06, "loss": 0.4042, "step": 1558 }, { "epoch": 0.6894924999078613, "grad_norm": 0.36734126941365897, "learning_rate": 9.497484879745019e-06, "loss": 0.4359, "step": 1559 }, { "epoch": 0.6899347657833634, "grad_norm": 0.3143686800958681, "learning_rate": 9.496360454175705e-06, "loss": 0.3691, "step": 1560 }, { "epoch": 0.6903770316588655, "grad_norm": 0.35807108561074297, "learning_rate": 9.495234838721326e-06, "loss": 0.3972, "step": 1561 }, { "epoch": 0.6908192975343678, "grad_norm": 0.2895949158250126, "learning_rate": 9.494108033679749e-06, "loss": 0.379, "step": 1562 }, { "epoch": 0.6912615634098699, "grad_norm": 0.3410089383906168, "learning_rate": 9.492980039349167e-06, "loss": 0.3984, "step": 1563 }, { "epoch": 0.6917038292853721, "grad_norm": 0.30866070988425715, "learning_rate": 9.491850856028084e-06, "loss": 0.3897, "step": 1564 }, { "epoch": 0.6921460951608742, "grad_norm": 0.31170509247254796, "learning_rate": 9.49072048401532e-06, "loss": 0.3607, "step": 1565 }, { "epoch": 0.6925883610363763, "grad_norm": 0.3173982891401866, "learning_rate": 9.489588923610006e-06, "loss": 0.3835, "step": 1566 }, { "epoch": 0.6930306269118786, "grad_norm": 0.3443358326128829, "learning_rate": 9.488456175111594e-06, "loss": 0.3607, "step": 1567 }, { "epoch": 0.6934728927873807, "grad_norm": 0.28532568405253167, "learning_rate": 9.487322238819843e-06, "loss": 0.3823, "step": 1568 }, { "epoch": 0.6939151586628829, "grad_norm": 0.3175322434578156, "learning_rate": 9.48618711503483e-06, "loss": 0.4322, "step": 1569 }, { "epoch": 0.694357424538385, "grad_norm": 0.2947544113451974, "learning_rate": 9.48505080405695e-06, "loss": 0.3904, "step": 1570 }, { "epoch": 0.6947996904138871, "grad_norm": 0.3029515996788336, "learning_rate": 9.483913306186904e-06, "loss": 0.3749, "step": 1571 }, { "epoch": 0.6952419562893893, "grad_norm": 0.29399779008250077, "learning_rate": 9.482774621725714e-06, "loss": 0.4076, "step": 1572 }, { "epoch": 0.6956842221648915, "grad_norm": 0.3515067781081394, "learning_rate": 9.481634750974714e-06, "loss": 0.3849, "step": 1573 }, { "epoch": 0.6961264880403936, "grad_norm": 0.2901071733344124, "learning_rate": 9.480493694235547e-06, "loss": 0.3397, "step": 1574 }, { "epoch": 0.6965687539158958, "grad_norm": 0.46626154994369245, "learning_rate": 9.47935145181018e-06, "loss": 0.3861, "step": 1575 }, { "epoch": 0.6970110197913979, "grad_norm": 0.3184789506932874, "learning_rate": 9.478208024000884e-06, "loss": 0.3999, "step": 1576 }, { "epoch": 0.6974532856669001, "grad_norm": 0.3300167023016541, "learning_rate": 9.477063411110248e-06, "loss": 0.3877, "step": 1577 }, { "epoch": 0.6978955515424022, "grad_norm": 0.2859243391985115, "learning_rate": 9.475917613441174e-06, "loss": 0.3491, "step": 1578 }, { "epoch": 0.6983378174179043, "grad_norm": 0.30375913873783, "learning_rate": 9.474770631296882e-06, "loss": 0.376, "step": 1579 }, { "epoch": 0.6987800832934066, "grad_norm": 0.35832141914508847, "learning_rate": 9.473622464980897e-06, "loss": 0.3765, "step": 1580 }, { "epoch": 0.6992223491689087, "grad_norm": 0.34952120229166117, "learning_rate": 9.472473114797061e-06, "loss": 0.4247, "step": 1581 }, { "epoch": 0.6996646150444109, "grad_norm": 0.32425046002622626, "learning_rate": 9.471322581049534e-06, "loss": 0.3788, "step": 1582 }, { "epoch": 0.700106880919913, "grad_norm": 0.3914718726705865, "learning_rate": 9.47017086404278e-06, "loss": 0.3769, "step": 1583 }, { "epoch": 0.7005491467954151, "grad_norm": 0.29735503025644067, "learning_rate": 9.469017964081588e-06, "loss": 0.3625, "step": 1584 }, { "epoch": 0.7009914126709174, "grad_norm": 0.2717582556490758, "learning_rate": 9.467863881471047e-06, "loss": 0.3722, "step": 1585 }, { "epoch": 0.7014336785464195, "grad_norm": 0.28660798502635243, "learning_rate": 9.466708616516567e-06, "loss": 0.362, "step": 1586 }, { "epoch": 0.7018759444219217, "grad_norm": 0.312476379928846, "learning_rate": 9.465552169523871e-06, "loss": 0.408, "step": 1587 }, { "epoch": 0.7023182102974238, "grad_norm": 0.3647292535701867, "learning_rate": 9.46439454079899e-06, "loss": 0.4191, "step": 1588 }, { "epoch": 0.7027604761729259, "grad_norm": 0.3098486037426103, "learning_rate": 9.463235730648274e-06, "loss": 0.3577, "step": 1589 }, { "epoch": 0.7032027420484281, "grad_norm": 0.32518466789728834, "learning_rate": 9.46207573937838e-06, "loss": 0.4082, "step": 1590 }, { "epoch": 0.7036450079239303, "grad_norm": 0.27302621653206066, "learning_rate": 9.46091456729628e-06, "loss": 0.3635, "step": 1591 }, { "epoch": 0.7040872737994325, "grad_norm": 0.2690781139415216, "learning_rate": 9.45975221470926e-06, "loss": 0.3842, "step": 1592 }, { "epoch": 0.7045295396749346, "grad_norm": 0.3623102892399632, "learning_rate": 9.458588681924916e-06, "loss": 0.3894, "step": 1593 }, { "epoch": 0.7049718055504367, "grad_norm": 0.3334581377979194, "learning_rate": 9.457423969251157e-06, "loss": 0.3939, "step": 1594 }, { "epoch": 0.7054140714259389, "grad_norm": 0.3067538800311085, "learning_rate": 9.456258076996203e-06, "loss": 0.4216, "step": 1595 }, { "epoch": 0.705856337301441, "grad_norm": 0.27297082817647306, "learning_rate": 9.45509100546859e-06, "loss": 0.3761, "step": 1596 }, { "epoch": 0.7062986031769432, "grad_norm": 0.3778247281405022, "learning_rate": 9.45392275497716e-06, "loss": 0.4201, "step": 1597 }, { "epoch": 0.7067408690524454, "grad_norm": 0.33553846239462526, "learning_rate": 9.452753325831076e-06, "loss": 0.4657, "step": 1598 }, { "epoch": 0.7071831349279475, "grad_norm": 0.3048315279276765, "learning_rate": 9.451582718339803e-06, "loss": 0.3853, "step": 1599 }, { "epoch": 0.7076254008034497, "grad_norm": 0.2708687436570743, "learning_rate": 9.450410932813123e-06, "loss": 0.39, "step": 1600 }, { "epoch": 0.7080676666789518, "grad_norm": 0.3633698813575066, "learning_rate": 9.449237969561129e-06, "loss": 0.4555, "step": 1601 }, { "epoch": 0.7085099325544539, "grad_norm": 0.32958496490955486, "learning_rate": 9.448063828894229e-06, "loss": 0.3769, "step": 1602 }, { "epoch": 0.7089521984299562, "grad_norm": 0.2869006881448628, "learning_rate": 9.446888511123133e-06, "loss": 0.3734, "step": 1603 }, { "epoch": 0.7093944643054583, "grad_norm": 0.44903507921751923, "learning_rate": 9.445712016558875e-06, "loss": 0.4198, "step": 1604 }, { "epoch": 0.7098367301809605, "grad_norm": 0.3170807166568624, "learning_rate": 9.444534345512793e-06, "loss": 0.364, "step": 1605 }, { "epoch": 0.7102789960564626, "grad_norm": 0.33080107386095325, "learning_rate": 9.443355498296534e-06, "loss": 0.3988, "step": 1606 }, { "epoch": 0.7107212619319647, "grad_norm": 0.2882214331196935, "learning_rate": 9.442175475222062e-06, "loss": 0.395, "step": 1607 }, { "epoch": 0.7111635278074669, "grad_norm": 0.2971817847629645, "learning_rate": 9.44099427660165e-06, "loss": 0.3968, "step": 1608 }, { "epoch": 0.7116057936829691, "grad_norm": 0.34305892039138525, "learning_rate": 9.439811902747882e-06, "loss": 0.4196, "step": 1609 }, { "epoch": 0.7120480595584713, "grad_norm": 0.3308224335523372, "learning_rate": 9.438628353973654e-06, "loss": 0.3903, "step": 1610 }, { "epoch": 0.7124903254339734, "grad_norm": 0.30903283309336327, "learning_rate": 9.43744363059217e-06, "loss": 0.3637, "step": 1611 }, { "epoch": 0.7129325913094755, "grad_norm": 0.3401148189520625, "learning_rate": 9.436257732916947e-06, "loss": 0.4365, "step": 1612 }, { "epoch": 0.7133748571849777, "grad_norm": 0.35562523162284865, "learning_rate": 9.435070661261815e-06, "loss": 0.3979, "step": 1613 }, { "epoch": 0.7138171230604798, "grad_norm": 0.32646315771650125, "learning_rate": 9.433882415940911e-06, "loss": 0.3555, "step": 1614 }, { "epoch": 0.7142593889359821, "grad_norm": 0.30637321060263273, "learning_rate": 9.432692997268682e-06, "loss": 0.3916, "step": 1615 }, { "epoch": 0.7147016548114842, "grad_norm": 0.4359995668075346, "learning_rate": 9.431502405559892e-06, "loss": 0.3869, "step": 1616 }, { "epoch": 0.7151439206869863, "grad_norm": 0.3602940386126426, "learning_rate": 9.430310641129605e-06, "loss": 0.4065, "step": 1617 }, { "epoch": 0.7155861865624885, "grad_norm": 0.35765664075524767, "learning_rate": 9.429117704293205e-06, "loss": 0.4147, "step": 1618 }, { "epoch": 0.7160284524379906, "grad_norm": 0.2993584016167907, "learning_rate": 9.42792359536638e-06, "loss": 0.4144, "step": 1619 }, { "epoch": 0.7164707183134928, "grad_norm": 0.3162268033409356, "learning_rate": 9.426728314665134e-06, "loss": 0.3891, "step": 1620 }, { "epoch": 0.716912984188995, "grad_norm": 0.4272502239198801, "learning_rate": 9.425531862505774e-06, "loss": 0.4644, "step": 1621 }, { "epoch": 0.7173552500644971, "grad_norm": 0.3053645143625031, "learning_rate": 9.424334239204922e-06, "loss": 0.3864, "step": 1622 }, { "epoch": 0.7177975159399993, "grad_norm": 0.31785042120432916, "learning_rate": 9.423135445079507e-06, "loss": 0.4253, "step": 1623 }, { "epoch": 0.7182397818155014, "grad_norm": 0.30748977956140566, "learning_rate": 9.421935480446774e-06, "loss": 0.3421, "step": 1624 }, { "epoch": 0.7186820476910035, "grad_norm": 0.2837046054645895, "learning_rate": 9.42073434562427e-06, "loss": 0.3789, "step": 1625 }, { "epoch": 0.7191243135665057, "grad_norm": 0.28901152569032973, "learning_rate": 9.419532040929853e-06, "loss": 0.3616, "step": 1626 }, { "epoch": 0.7195665794420079, "grad_norm": 0.2935132823265459, "learning_rate": 9.418328566681695e-06, "loss": 0.3878, "step": 1627 }, { "epoch": 0.7200088453175101, "grad_norm": 0.31725284789238034, "learning_rate": 9.417123923198277e-06, "loss": 0.376, "step": 1628 }, { "epoch": 0.7204511111930122, "grad_norm": 0.3592543594493535, "learning_rate": 9.41591811079838e-06, "loss": 0.4276, "step": 1629 }, { "epoch": 0.7208933770685143, "grad_norm": 0.33790143915537446, "learning_rate": 9.414711129801111e-06, "loss": 0.3962, "step": 1630 }, { "epoch": 0.7213356429440165, "grad_norm": 0.3635922732904285, "learning_rate": 9.413502980525871e-06, "loss": 0.4027, "step": 1631 }, { "epoch": 0.7217779088195186, "grad_norm": 0.3396239504442322, "learning_rate": 9.412293663292376e-06, "loss": 0.3925, "step": 1632 }, { "epoch": 0.7222201746950209, "grad_norm": 0.37347436677108414, "learning_rate": 9.411083178420654e-06, "loss": 0.4081, "step": 1633 }, { "epoch": 0.722662440570523, "grad_norm": 0.28840788596036887, "learning_rate": 9.409871526231036e-06, "loss": 0.3895, "step": 1634 }, { "epoch": 0.7231047064460251, "grad_norm": 0.2799434346376729, "learning_rate": 9.40865870704417e-06, "loss": 0.3972, "step": 1635 }, { "epoch": 0.7235469723215273, "grad_norm": 0.38842964914390704, "learning_rate": 9.407444721181001e-06, "loss": 0.4226, "step": 1636 }, { "epoch": 0.7239892381970294, "grad_norm": 0.301116167267562, "learning_rate": 9.406229568962796e-06, "loss": 0.3761, "step": 1637 }, { "epoch": 0.7244315040725317, "grad_norm": 0.3190175788677148, "learning_rate": 9.40501325071112e-06, "loss": 0.38, "step": 1638 }, { "epoch": 0.7248737699480338, "grad_norm": 0.3564511735293726, "learning_rate": 9.403795766747854e-06, "loss": 0.3733, "step": 1639 }, { "epoch": 0.7253160358235359, "grad_norm": 0.3063885931271243, "learning_rate": 9.402577117395183e-06, "loss": 0.3923, "step": 1640 }, { "epoch": 0.7257583016990381, "grad_norm": 0.3684061669321572, "learning_rate": 9.4013573029756e-06, "loss": 0.4582, "step": 1641 }, { "epoch": 0.7262005675745402, "grad_norm": 0.2821673093001152, "learning_rate": 9.400136323811912e-06, "loss": 0.3885, "step": 1642 }, { "epoch": 0.7266428334500424, "grad_norm": 0.3127336363568668, "learning_rate": 9.398914180227226e-06, "loss": 0.3882, "step": 1643 }, { "epoch": 0.7270850993255445, "grad_norm": 0.28223826931053964, "learning_rate": 9.397690872544964e-06, "loss": 0.3644, "step": 1644 }, { "epoch": 0.7275273652010467, "grad_norm": 0.33647845592096715, "learning_rate": 9.396466401088855e-06, "loss": 0.4314, "step": 1645 }, { "epoch": 0.7279696310765489, "grad_norm": 0.3886599055622626, "learning_rate": 9.39524076618293e-06, "loss": 0.4165, "step": 1646 }, { "epoch": 0.728411896952051, "grad_norm": 0.3312855571217781, "learning_rate": 9.394013968151536e-06, "loss": 0.4021, "step": 1647 }, { "epoch": 0.7288541628275531, "grad_norm": 0.35875211545654156, "learning_rate": 9.392786007319321e-06, "loss": 0.4225, "step": 1648 }, { "epoch": 0.7292964287030553, "grad_norm": 0.40902448372619005, "learning_rate": 9.391556884011248e-06, "loss": 0.4226, "step": 1649 }, { "epoch": 0.7297386945785574, "grad_norm": 0.35203191571868453, "learning_rate": 9.390326598552581e-06, "loss": 0.3691, "step": 1650 }, { "epoch": 0.7301809604540597, "grad_norm": 0.28402822290377666, "learning_rate": 9.389095151268894e-06, "loss": 0.3633, "step": 1651 }, { "epoch": 0.7306232263295618, "grad_norm": 0.299781877054248, "learning_rate": 9.387862542486071e-06, "loss": 0.3729, "step": 1652 }, { "epoch": 0.7310654922050639, "grad_norm": 0.40733357071168397, "learning_rate": 9.386628772530295e-06, "loss": 0.384, "step": 1653 }, { "epoch": 0.7315077580805661, "grad_norm": 0.3255059439029067, "learning_rate": 9.38539384172807e-06, "loss": 0.3827, "step": 1654 }, { "epoch": 0.7319500239560682, "grad_norm": 0.3174799891423009, "learning_rate": 9.384157750406193e-06, "loss": 0.4243, "step": 1655 }, { "epoch": 0.7323922898315705, "grad_norm": 0.3283715761601307, "learning_rate": 9.382920498891777e-06, "loss": 0.3573, "step": 1656 }, { "epoch": 0.7328345557070726, "grad_norm": 0.30360928714572466, "learning_rate": 9.38168208751224e-06, "loss": 0.3695, "step": 1657 }, { "epoch": 0.7332768215825747, "grad_norm": 0.3021292781702085, "learning_rate": 9.380442516595305e-06, "loss": 0.3758, "step": 1658 }, { "epoch": 0.7337190874580769, "grad_norm": 0.33065371834441487, "learning_rate": 9.379201786469005e-06, "loss": 0.3744, "step": 1659 }, { "epoch": 0.734161353333579, "grad_norm": 0.41566959946936954, "learning_rate": 9.377959897461677e-06, "loss": 0.4534, "step": 1660 }, { "epoch": 0.7346036192090812, "grad_norm": 0.30142181680225505, "learning_rate": 9.376716849901966e-06, "loss": 0.3695, "step": 1661 }, { "epoch": 0.7350458850845834, "grad_norm": 0.3335304521244606, "learning_rate": 9.375472644118822e-06, "loss": 0.4104, "step": 1662 }, { "epoch": 0.7354881509600855, "grad_norm": 0.3301049738892589, "learning_rate": 9.374227280441504e-06, "loss": 0.3742, "step": 1663 }, { "epoch": 0.7359304168355877, "grad_norm": 0.3027939204780596, "learning_rate": 9.372980759199576e-06, "loss": 0.3918, "step": 1664 }, { "epoch": 0.7363726827110898, "grad_norm": 0.3117665680438885, "learning_rate": 9.371733080722911e-06, "loss": 0.3902, "step": 1665 }, { "epoch": 0.736814948586592, "grad_norm": 0.4260278535415941, "learning_rate": 9.370484245341683e-06, "loss": 0.3987, "step": 1666 }, { "epoch": 0.7372572144620941, "grad_norm": 0.335739935485861, "learning_rate": 9.369234253386374e-06, "loss": 0.3858, "step": 1667 }, { "epoch": 0.7376994803375962, "grad_norm": 0.3205378895417316, "learning_rate": 9.367983105187776e-06, "loss": 0.3902, "step": 1668 }, { "epoch": 0.7381417462130985, "grad_norm": 0.3036043534034749, "learning_rate": 9.366730801076985e-06, "loss": 0.3513, "step": 1669 }, { "epoch": 0.7385840120886006, "grad_norm": 0.4430125822550183, "learning_rate": 9.365477341385397e-06, "loss": 0.4275, "step": 1670 }, { "epoch": 0.7390262779641028, "grad_norm": 0.36105428712889487, "learning_rate": 9.36422272644472e-06, "loss": 0.3951, "step": 1671 }, { "epoch": 0.7394685438396049, "grad_norm": 0.3824982573220564, "learning_rate": 9.36296695658697e-06, "loss": 0.4255, "step": 1672 }, { "epoch": 0.739910809715107, "grad_norm": 0.29218414569038237, "learning_rate": 9.361710032144463e-06, "loss": 0.4001, "step": 1673 }, { "epoch": 0.7403530755906093, "grad_norm": 0.34766229934840404, "learning_rate": 9.36045195344982e-06, "loss": 0.3937, "step": 1674 }, { "epoch": 0.7407953414661114, "grad_norm": 0.3424613781848369, "learning_rate": 9.359192720835975e-06, "loss": 0.3991, "step": 1675 }, { "epoch": 0.7412376073416135, "grad_norm": 0.3047354229746975, "learning_rate": 9.35793233463616e-06, "loss": 0.396, "step": 1676 }, { "epoch": 0.7416798732171157, "grad_norm": 0.3781615979190017, "learning_rate": 9.356670795183913e-06, "loss": 0.3774, "step": 1677 }, { "epoch": 0.7421221390926178, "grad_norm": 0.39464374264848634, "learning_rate": 9.35540810281308e-06, "loss": 0.4145, "step": 1678 }, { "epoch": 0.74256440496812, "grad_norm": 0.32181964025713783, "learning_rate": 9.354144257857813e-06, "loss": 0.4098, "step": 1679 }, { "epoch": 0.7430066708436222, "grad_norm": 0.28751310927109963, "learning_rate": 9.352879260652566e-06, "loss": 0.3845, "step": 1680 }, { "epoch": 0.7434489367191243, "grad_norm": 0.3533554000092139, "learning_rate": 9.351613111532097e-06, "loss": 0.4078, "step": 1681 }, { "epoch": 0.7438912025946265, "grad_norm": 0.2657889361693365, "learning_rate": 9.350345810831473e-06, "loss": 0.3558, "step": 1682 }, { "epoch": 0.7443334684701286, "grad_norm": 0.3265163146702926, "learning_rate": 9.349077358886063e-06, "loss": 0.3863, "step": 1683 }, { "epoch": 0.7447757343456308, "grad_norm": 0.30527068605950713, "learning_rate": 9.347807756031539e-06, "loss": 0.3933, "step": 1684 }, { "epoch": 0.7452180002211329, "grad_norm": 0.2980169875707563, "learning_rate": 9.346537002603883e-06, "loss": 0.3548, "step": 1685 }, { "epoch": 0.745660266096635, "grad_norm": 0.32452623987979506, "learning_rate": 9.345265098939379e-06, "loss": 0.3586, "step": 1686 }, { "epoch": 0.7461025319721373, "grad_norm": 0.36657478872150023, "learning_rate": 9.343992045374608e-06, "loss": 0.3888, "step": 1687 }, { "epoch": 0.7465447978476394, "grad_norm": 0.44905951264520483, "learning_rate": 9.342717842246468e-06, "loss": 0.3733, "step": 1688 }, { "epoch": 0.7469870637231416, "grad_norm": 0.28724133197492807, "learning_rate": 9.341442489892156e-06, "loss": 0.3726, "step": 1689 }, { "epoch": 0.7474293295986437, "grad_norm": 0.2943843548853431, "learning_rate": 9.340165988649165e-06, "loss": 0.392, "step": 1690 }, { "epoch": 0.7478715954741458, "grad_norm": 0.27486487639476154, "learning_rate": 9.338888338855307e-06, "loss": 0.3759, "step": 1691 }, { "epoch": 0.7483138613496481, "grad_norm": 0.37777091504942734, "learning_rate": 9.337609540848689e-06, "loss": 0.3965, "step": 1692 }, { "epoch": 0.7487561272251502, "grad_norm": 0.29562138394025766, "learning_rate": 9.336329594967718e-06, "loss": 0.3616, "step": 1693 }, { "epoch": 0.7491983931006524, "grad_norm": 0.2863589085626817, "learning_rate": 9.335048501551114e-06, "loss": 0.3772, "step": 1694 }, { "epoch": 0.7496406589761545, "grad_norm": 0.3276279912507926, "learning_rate": 9.333766260937895e-06, "loss": 0.3794, "step": 1695 }, { "epoch": 0.7500829248516566, "grad_norm": 0.32844413274270134, "learning_rate": 9.332482873467385e-06, "loss": 0.4018, "step": 1696 }, { "epoch": 0.7505251907271588, "grad_norm": 0.3703238017945596, "learning_rate": 9.33119833947921e-06, "loss": 0.4004, "step": 1697 }, { "epoch": 0.750967456602661, "grad_norm": 0.2989127724124875, "learning_rate": 9.329912659313302e-06, "loss": 0.429, "step": 1698 }, { "epoch": 0.7514097224781632, "grad_norm": 0.40062063906930684, "learning_rate": 9.328625833309891e-06, "loss": 0.4338, "step": 1699 }, { "epoch": 0.7518519883536653, "grad_norm": 0.37257846509525405, "learning_rate": 9.327337861809516e-06, "loss": 0.3822, "step": 1700 }, { "epoch": 0.7522942542291674, "grad_norm": 0.5473192367571187, "learning_rate": 9.326048745153015e-06, "loss": 0.4434, "step": 1701 }, { "epoch": 0.7527365201046696, "grad_norm": 0.28419599034899184, "learning_rate": 9.324758483681532e-06, "loss": 0.3804, "step": 1702 }, { "epoch": 0.7531787859801717, "grad_norm": 0.3989871915864089, "learning_rate": 9.323467077736513e-06, "loss": 0.394, "step": 1703 }, { "epoch": 0.7536210518556739, "grad_norm": 0.3607273405930398, "learning_rate": 9.322174527659704e-06, "loss": 0.385, "step": 1704 }, { "epoch": 0.7540633177311761, "grad_norm": 0.2999388275235852, "learning_rate": 9.320880833793158e-06, "loss": 0.3607, "step": 1705 }, { "epoch": 0.7545055836066782, "grad_norm": 0.29422883398675415, "learning_rate": 9.31958599647923e-06, "loss": 0.3937, "step": 1706 }, { "epoch": 0.7549478494821804, "grad_norm": 0.29893324558731404, "learning_rate": 9.318290016060576e-06, "loss": 0.4016, "step": 1707 }, { "epoch": 0.7553901153576825, "grad_norm": 0.3032968396124369, "learning_rate": 9.316992892880154e-06, "loss": 0.3911, "step": 1708 }, { "epoch": 0.7558323812331846, "grad_norm": 0.38704674857993593, "learning_rate": 9.315694627281226e-06, "loss": 0.4255, "step": 1709 }, { "epoch": 0.7562746471086869, "grad_norm": 0.33241799643610115, "learning_rate": 9.314395219607359e-06, "loss": 0.4193, "step": 1710 }, { "epoch": 0.756716912984189, "grad_norm": 0.37068831287741755, "learning_rate": 9.313094670202413e-06, "loss": 0.4063, "step": 1711 }, { "epoch": 0.7571591788596912, "grad_norm": 0.3185403708331871, "learning_rate": 9.31179297941056e-06, "loss": 0.3843, "step": 1712 }, { "epoch": 0.7576014447351933, "grad_norm": 0.2829641908671659, "learning_rate": 9.310490147576271e-06, "loss": 0.3636, "step": 1713 }, { "epoch": 0.7580437106106954, "grad_norm": 0.26980083609458205, "learning_rate": 9.309186175044316e-06, "loss": 0.3693, "step": 1714 }, { "epoch": 0.7584859764861976, "grad_norm": 0.3070022722984844, "learning_rate": 9.307881062159771e-06, "loss": 0.3788, "step": 1715 }, { "epoch": 0.7589282423616998, "grad_norm": 0.2975981808537097, "learning_rate": 9.30657480926801e-06, "loss": 0.3987, "step": 1716 }, { "epoch": 0.759370508237202, "grad_norm": 0.2864625360279755, "learning_rate": 9.305267416714711e-06, "loss": 0.3709, "step": 1717 }, { "epoch": 0.7598127741127041, "grad_norm": 0.33424161189526164, "learning_rate": 9.303958884845854e-06, "loss": 0.3707, "step": 1718 }, { "epoch": 0.7602550399882062, "grad_norm": 0.335008483187304, "learning_rate": 9.302649214007718e-06, "loss": 0.4145, "step": 1719 }, { "epoch": 0.7606973058637084, "grad_norm": 0.3021251565535318, "learning_rate": 9.301338404546887e-06, "loss": 0.3542, "step": 1720 }, { "epoch": 0.7611395717392105, "grad_norm": 0.3655071026492945, "learning_rate": 9.300026456810244e-06, "loss": 0.3666, "step": 1721 }, { "epoch": 0.7615818376147128, "grad_norm": 0.30288357310539005, "learning_rate": 9.298713371144974e-06, "loss": 0.3681, "step": 1722 }, { "epoch": 0.7620241034902149, "grad_norm": 0.3060254934555248, "learning_rate": 9.29739914789856e-06, "loss": 0.4308, "step": 1723 }, { "epoch": 0.762466369365717, "grad_norm": 0.3021775707033964, "learning_rate": 9.29608378741879e-06, "loss": 0.3978, "step": 1724 }, { "epoch": 0.7629086352412192, "grad_norm": 0.2992292781529154, "learning_rate": 9.294767290053754e-06, "loss": 0.3751, "step": 1725 }, { "epoch": 0.7633509011167213, "grad_norm": 0.35170235379503056, "learning_rate": 9.293449656151839e-06, "loss": 0.4113, "step": 1726 }, { "epoch": 0.7637931669922234, "grad_norm": 0.3052097394072818, "learning_rate": 9.292130886061733e-06, "loss": 0.3871, "step": 1727 }, { "epoch": 0.7642354328677257, "grad_norm": 0.2892280187996816, "learning_rate": 9.290810980132428e-06, "loss": 0.3607, "step": 1728 }, { "epoch": 0.7646776987432278, "grad_norm": 0.3804799769377048, "learning_rate": 9.289489938713213e-06, "loss": 0.4305, "step": 1729 }, { "epoch": 0.76511996461873, "grad_norm": 0.27498464685212903, "learning_rate": 9.288167762153681e-06, "loss": 0.3924, "step": 1730 }, { "epoch": 0.7655622304942321, "grad_norm": 0.321187998230285, "learning_rate": 9.28684445080372e-06, "loss": 0.3949, "step": 1731 }, { "epoch": 0.7660044963697342, "grad_norm": 0.3336306488236722, "learning_rate": 9.285520005013528e-06, "loss": 0.3567, "step": 1732 }, { "epoch": 0.7664467622452364, "grad_norm": 0.2876230981836362, "learning_rate": 9.28419442513359e-06, "loss": 0.3647, "step": 1733 }, { "epoch": 0.7668890281207386, "grad_norm": 0.3634879148396299, "learning_rate": 9.282867711514703e-06, "loss": 0.3878, "step": 1734 }, { "epoch": 0.7673312939962408, "grad_norm": 0.2980622070066765, "learning_rate": 9.281539864507956e-06, "loss": 0.3655, "step": 1735 }, { "epoch": 0.7677735598717429, "grad_norm": 0.34538784023106617, "learning_rate": 9.280210884464741e-06, "loss": 0.4033, "step": 1736 }, { "epoch": 0.768215825747245, "grad_norm": 0.2850031135837628, "learning_rate": 9.278880771736754e-06, "loss": 0.3687, "step": 1737 }, { "epoch": 0.7686580916227472, "grad_norm": 0.3106935937933193, "learning_rate": 9.277549526675982e-06, "loss": 0.3812, "step": 1738 }, { "epoch": 0.7691003574982493, "grad_norm": 0.33493977812122927, "learning_rate": 9.276217149634719e-06, "loss": 0.3953, "step": 1739 }, { "epoch": 0.7695426233737516, "grad_norm": 0.33549930990032484, "learning_rate": 9.274883640965554e-06, "loss": 0.3742, "step": 1740 }, { "epoch": 0.7699848892492537, "grad_norm": 0.26092438211320546, "learning_rate": 9.273549001021379e-06, "loss": 0.3516, "step": 1741 }, { "epoch": 0.7704271551247558, "grad_norm": 0.32680796984240007, "learning_rate": 9.272213230155382e-06, "loss": 0.4037, "step": 1742 }, { "epoch": 0.770869421000258, "grad_norm": 0.29355952289314874, "learning_rate": 9.270876328721055e-06, "loss": 0.3721, "step": 1743 }, { "epoch": 0.7713116868757601, "grad_norm": 0.3860996637036687, "learning_rate": 9.269538297072182e-06, "loss": 0.4293, "step": 1744 }, { "epoch": 0.7717539527512624, "grad_norm": 0.31690939829355397, "learning_rate": 9.268199135562854e-06, "loss": 0.3741, "step": 1745 }, { "epoch": 0.7721962186267645, "grad_norm": 0.43508923350647727, "learning_rate": 9.266858844547456e-06, "loss": 0.4713, "step": 1746 }, { "epoch": 0.7726384845022666, "grad_norm": 0.40719552458027664, "learning_rate": 9.26551742438067e-06, "loss": 0.4003, "step": 1747 }, { "epoch": 0.7730807503777688, "grad_norm": 0.3297468263699808, "learning_rate": 9.264174875417485e-06, "loss": 0.4087, "step": 1748 }, { "epoch": 0.7735230162532709, "grad_norm": 0.30051783993200387, "learning_rate": 9.262831198013181e-06, "loss": 0.3847, "step": 1749 }, { "epoch": 0.7739652821287731, "grad_norm": 0.3417722651265026, "learning_rate": 9.26148639252334e-06, "loss": 0.3884, "step": 1750 }, { "epoch": 0.7744075480042752, "grad_norm": 0.3246628265041788, "learning_rate": 9.26014045930384e-06, "loss": 0.3621, "step": 1751 }, { "epoch": 0.7748498138797774, "grad_norm": 0.28701103795020877, "learning_rate": 9.25879339871086e-06, "loss": 0.3817, "step": 1752 }, { "epoch": 0.7752920797552796, "grad_norm": 0.30668723233797307, "learning_rate": 9.25744521110088e-06, "loss": 0.405, "step": 1753 }, { "epoch": 0.7757343456307817, "grad_norm": 0.3484570166105187, "learning_rate": 9.256095896830672e-06, "loss": 0.3623, "step": 1754 }, { "epoch": 0.7761766115062838, "grad_norm": 0.3278648680225415, "learning_rate": 9.254745456257309e-06, "loss": 0.3866, "step": 1755 }, { "epoch": 0.776618877381786, "grad_norm": 0.31429692178361274, "learning_rate": 9.25339388973816e-06, "loss": 0.3537, "step": 1756 }, { "epoch": 0.7770611432572881, "grad_norm": 0.3344934167461727, "learning_rate": 9.252041197630899e-06, "loss": 0.4294, "step": 1757 }, { "epoch": 0.7775034091327904, "grad_norm": 0.33707540293983146, "learning_rate": 9.250687380293489e-06, "loss": 0.3812, "step": 1758 }, { "epoch": 0.7779456750082925, "grad_norm": 0.34721034833636055, "learning_rate": 9.249332438084196e-06, "loss": 0.3713, "step": 1759 }, { "epoch": 0.7783879408837946, "grad_norm": 0.35981568540103603, "learning_rate": 9.247976371361582e-06, "loss": 0.3884, "step": 1760 }, { "epoch": 0.7788302067592968, "grad_norm": 0.31134214110209724, "learning_rate": 9.246619180484507e-06, "loss": 0.3902, "step": 1761 }, { "epoch": 0.7792724726347989, "grad_norm": 0.33395289616953694, "learning_rate": 9.245260865812128e-06, "loss": 0.4193, "step": 1762 }, { "epoch": 0.7797147385103012, "grad_norm": 0.3160188364295939, "learning_rate": 9.243901427703902e-06, "loss": 0.391, "step": 1763 }, { "epoch": 0.7801570043858033, "grad_norm": 0.2741061648363806, "learning_rate": 9.242540866519578e-06, "loss": 0.3997, "step": 1764 }, { "epoch": 0.7805992702613054, "grad_norm": 0.32375702264581896, "learning_rate": 9.241179182619207e-06, "loss": 0.4132, "step": 1765 }, { "epoch": 0.7810415361368076, "grad_norm": 0.2808666777207927, "learning_rate": 9.239816376363135e-06, "loss": 0.427, "step": 1766 }, { "epoch": 0.7814838020123097, "grad_norm": 0.28509311699737244, "learning_rate": 9.238452448112005e-06, "loss": 0.3715, "step": 1767 }, { "epoch": 0.7819260678878119, "grad_norm": 0.29695115599991057, "learning_rate": 9.237087398226758e-06, "loss": 0.3785, "step": 1768 }, { "epoch": 0.782368333763314, "grad_norm": 0.30659600845104573, "learning_rate": 9.235721227068633e-06, "loss": 0.3727, "step": 1769 }, { "epoch": 0.7828105996388162, "grad_norm": 0.356769282930287, "learning_rate": 9.23435393499916e-06, "loss": 0.4109, "step": 1770 }, { "epoch": 0.7832528655143184, "grad_norm": 0.29766100763145914, "learning_rate": 9.23298552238017e-06, "loss": 0.3717, "step": 1771 }, { "epoch": 0.7836951313898205, "grad_norm": 0.29304983876484786, "learning_rate": 9.231615989573794e-06, "loss": 0.3828, "step": 1772 }, { "epoch": 0.7841373972653227, "grad_norm": 0.26360680113492135, "learning_rate": 9.230245336942452e-06, "loss": 0.379, "step": 1773 }, { "epoch": 0.7845796631408248, "grad_norm": 0.3459342269609132, "learning_rate": 9.228873564848864e-06, "loss": 0.3958, "step": 1774 }, { "epoch": 0.785021929016327, "grad_norm": 0.2913912856022195, "learning_rate": 9.227500673656048e-06, "loss": 0.3897, "step": 1775 }, { "epoch": 0.7854641948918292, "grad_norm": 0.3760158175081828, "learning_rate": 9.226126663727315e-06, "loss": 0.4128, "step": 1776 }, { "epoch": 0.7859064607673313, "grad_norm": 0.3476804777550466, "learning_rate": 9.224751535426273e-06, "loss": 0.391, "step": 1777 }, { "epoch": 0.7863487266428334, "grad_norm": 0.424764171193664, "learning_rate": 9.223375289116826e-06, "loss": 0.3664, "step": 1778 }, { "epoch": 0.7867909925183356, "grad_norm": 0.29470927316712453, "learning_rate": 9.221997925163176e-06, "loss": 0.387, "step": 1779 }, { "epoch": 0.7872332583938377, "grad_norm": 0.3162985697681358, "learning_rate": 9.220619443929817e-06, "loss": 0.3863, "step": 1780 }, { "epoch": 0.78767552426934, "grad_norm": 0.26029054708047067, "learning_rate": 9.21923984578154e-06, "loss": 0.3723, "step": 1781 }, { "epoch": 0.7881177901448421, "grad_norm": 0.30053883665813297, "learning_rate": 9.217859131083436e-06, "loss": 0.3768, "step": 1782 }, { "epoch": 0.7885600560203442, "grad_norm": 0.3455031529027387, "learning_rate": 9.216477300200881e-06, "loss": 0.4381, "step": 1783 }, { "epoch": 0.7890023218958464, "grad_norm": 0.33897675802579647, "learning_rate": 9.215094353499559e-06, "loss": 0.4164, "step": 1784 }, { "epoch": 0.7894445877713485, "grad_norm": 0.3277982823369914, "learning_rate": 9.213710291345441e-06, "loss": 0.3613, "step": 1785 }, { "epoch": 0.7898868536468507, "grad_norm": 0.3092214366150326, "learning_rate": 9.212325114104796e-06, "loss": 0.383, "step": 1786 }, { "epoch": 0.7903291195223529, "grad_norm": 0.3760502114837525, "learning_rate": 9.210938822144186e-06, "loss": 0.4072, "step": 1787 }, { "epoch": 0.790771385397855, "grad_norm": 0.32962557761451744, "learning_rate": 9.20955141583047e-06, "loss": 0.3701, "step": 1788 }, { "epoch": 0.7912136512733572, "grad_norm": 0.3685763759136502, "learning_rate": 9.208162895530803e-06, "loss": 0.4151, "step": 1789 }, { "epoch": 0.7916559171488593, "grad_norm": 0.32034219159957705, "learning_rate": 9.206773261612632e-06, "loss": 0.3705, "step": 1790 }, { "epoch": 0.7920981830243615, "grad_norm": 0.3489382790394267, "learning_rate": 9.2053825144437e-06, "loss": 0.3943, "step": 1791 }, { "epoch": 0.7925404488998636, "grad_norm": 0.3434694566768477, "learning_rate": 9.203990654392044e-06, "loss": 0.4257, "step": 1792 }, { "epoch": 0.7929827147753657, "grad_norm": 0.3009397692674527, "learning_rate": 9.202597681825995e-06, "loss": 0.3503, "step": 1793 }, { "epoch": 0.793424980650868, "grad_norm": 0.3338520695121851, "learning_rate": 9.201203597114183e-06, "loss": 0.3561, "step": 1794 }, { "epoch": 0.7938672465263701, "grad_norm": 0.3245997795004515, "learning_rate": 9.199808400625525e-06, "loss": 0.3896, "step": 1795 }, { "epoch": 0.7943095124018723, "grad_norm": 0.6197314932053445, "learning_rate": 9.19841209272924e-06, "loss": 0.4677, "step": 1796 }, { "epoch": 0.7947517782773744, "grad_norm": 0.2759719624033556, "learning_rate": 9.197014673794833e-06, "loss": 0.3876, "step": 1797 }, { "epoch": 0.7951940441528765, "grad_norm": 0.2926208941760109, "learning_rate": 9.19561614419211e-06, "loss": 0.3801, "step": 1798 }, { "epoch": 0.7956363100283788, "grad_norm": 0.33066955875900256, "learning_rate": 9.194216504291165e-06, "loss": 0.3786, "step": 1799 }, { "epoch": 0.7960785759038809, "grad_norm": 0.31868475973938587, "learning_rate": 9.192815754462392e-06, "loss": 0.4378, "step": 1800 }, { "epoch": 0.7965208417793831, "grad_norm": 0.29627631731911, "learning_rate": 9.191413895076473e-06, "loss": 0.3652, "step": 1801 }, { "epoch": 0.7969631076548852, "grad_norm": 0.3084754230974719, "learning_rate": 9.190010926504388e-06, "loss": 0.3906, "step": 1802 }, { "epoch": 0.7974053735303873, "grad_norm": 0.5407042005213563, "learning_rate": 9.188606849117407e-06, "loss": 0.361, "step": 1803 }, { "epoch": 0.7978476394058895, "grad_norm": 0.26539703366203754, "learning_rate": 9.187201663287098e-06, "loss": 0.3643, "step": 1804 }, { "epoch": 0.7982899052813917, "grad_norm": 0.2738966548601204, "learning_rate": 9.185795369385316e-06, "loss": 0.3888, "step": 1805 }, { "epoch": 0.7987321711568938, "grad_norm": 0.3042784144190533, "learning_rate": 9.184387967784213e-06, "loss": 0.4156, "step": 1806 }, { "epoch": 0.799174437032396, "grad_norm": 0.38145431863446905, "learning_rate": 9.182979458856237e-06, "loss": 0.3732, "step": 1807 }, { "epoch": 0.7996167029078981, "grad_norm": 0.2901606019807389, "learning_rate": 9.181569842974125e-06, "loss": 0.4047, "step": 1808 }, { "epoch": 0.8000589687834003, "grad_norm": 0.33339911528371613, "learning_rate": 9.180159120510905e-06, "loss": 0.3664, "step": 1809 }, { "epoch": 0.8005012346589024, "grad_norm": 0.30280800432429883, "learning_rate": 9.178747291839902e-06, "loss": 0.4057, "step": 1810 }, { "epoch": 0.8009435005344046, "grad_norm": 0.5820581409441385, "learning_rate": 9.177334357334732e-06, "loss": 0.4071, "step": 1811 }, { "epoch": 0.8013857664099068, "grad_norm": 0.3754652051795092, "learning_rate": 9.175920317369306e-06, "loss": 0.4341, "step": 1812 }, { "epoch": 0.8018280322854089, "grad_norm": 0.38493134095199655, "learning_rate": 9.174505172317825e-06, "loss": 0.3723, "step": 1813 }, { "epoch": 0.8022702981609111, "grad_norm": 0.39530207124020983, "learning_rate": 9.17308892255478e-06, "loss": 0.4199, "step": 1814 }, { "epoch": 0.8027125640364132, "grad_norm": 0.2729549682871308, "learning_rate": 9.171671568454961e-06, "loss": 0.3553, "step": 1815 }, { "epoch": 0.8031548299119153, "grad_norm": 0.4028590052112572, "learning_rate": 9.170253110393446e-06, "loss": 0.4474, "step": 1816 }, { "epoch": 0.8035970957874176, "grad_norm": 0.30273914387316825, "learning_rate": 9.168833548745602e-06, "loss": 0.3701, "step": 1817 }, { "epoch": 0.8040393616629197, "grad_norm": 0.9195001795565988, "learning_rate": 9.167412883887098e-06, "loss": 0.4071, "step": 1818 }, { "epoch": 0.8044816275384219, "grad_norm": 0.29368249903977806, "learning_rate": 9.165991116193881e-06, "loss": 0.3759, "step": 1819 }, { "epoch": 0.804923893413924, "grad_norm": 0.3226874989113338, "learning_rate": 9.164568246042206e-06, "loss": 0.4413, "step": 1820 }, { "epoch": 0.8053661592894261, "grad_norm": 0.3184285423352739, "learning_rate": 9.163144273808607e-06, "loss": 0.3859, "step": 1821 }, { "epoch": 0.8058084251649283, "grad_norm": 0.3357018907498215, "learning_rate": 9.161719199869912e-06, "loss": 0.3632, "step": 1822 }, { "epoch": 0.8062506910404305, "grad_norm": 0.5995292415298172, "learning_rate": 9.160293024603244e-06, "loss": 0.3723, "step": 1823 }, { "epoch": 0.8066929569159327, "grad_norm": 0.3375214563386007, "learning_rate": 9.158865748386019e-06, "loss": 0.3713, "step": 1824 }, { "epoch": 0.8071352227914348, "grad_norm": 0.3186783219800366, "learning_rate": 9.157437371595938e-06, "loss": 0.3837, "step": 1825 }, { "epoch": 0.8075774886669369, "grad_norm": 0.42753633111393563, "learning_rate": 9.156007894610997e-06, "loss": 0.3841, "step": 1826 }, { "epoch": 0.8080197545424391, "grad_norm": 0.3410802462745644, "learning_rate": 9.154577317809483e-06, "loss": 0.3946, "step": 1827 }, { "epoch": 0.8084620204179412, "grad_norm": 0.35775877996752214, "learning_rate": 9.153145641569976e-06, "loss": 0.4022, "step": 1828 }, { "epoch": 0.8089042862934434, "grad_norm": 0.3011954168357783, "learning_rate": 9.151712866271341e-06, "loss": 0.3593, "step": 1829 }, { "epoch": 0.8093465521689456, "grad_norm": 0.391175047199955, "learning_rate": 9.150278992292737e-06, "loss": 0.3898, "step": 1830 }, { "epoch": 0.8097888180444477, "grad_norm": 0.38555221523522615, "learning_rate": 9.14884402001362e-06, "loss": 0.3467, "step": 1831 }, { "epoch": 0.8102310839199499, "grad_norm": 0.29020107185838206, "learning_rate": 9.147407949813725e-06, "loss": 0.3746, "step": 1832 }, { "epoch": 0.810673349795452, "grad_norm": 0.2864530877152059, "learning_rate": 9.145970782073087e-06, "loss": 0.3998, "step": 1833 }, { "epoch": 0.8111156156709541, "grad_norm": 0.3385407105838061, "learning_rate": 9.144532517172025e-06, "loss": 0.3511, "step": 1834 }, { "epoch": 0.8115578815464564, "grad_norm": 0.34368465498505996, "learning_rate": 9.143093155491156e-06, "loss": 0.3937, "step": 1835 }, { "epoch": 0.8120001474219585, "grad_norm": 0.4359311735162348, "learning_rate": 9.141652697411377e-06, "loss": 0.4165, "step": 1836 }, { "epoch": 0.8124424132974607, "grad_norm": 0.34347176189583833, "learning_rate": 9.140211143313883e-06, "loss": 0.3749, "step": 1837 }, { "epoch": 0.8128846791729628, "grad_norm": 0.40939328809922715, "learning_rate": 9.138768493580156e-06, "loss": 0.387, "step": 1838 }, { "epoch": 0.8133269450484649, "grad_norm": 0.37543232118521314, "learning_rate": 9.137324748591973e-06, "loss": 0.4173, "step": 1839 }, { "epoch": 0.8137692109239671, "grad_norm": 0.2822308114012478, "learning_rate": 9.13587990873139e-06, "loss": 0.3564, "step": 1840 }, { "epoch": 0.8142114767994693, "grad_norm": 0.41554916454849233, "learning_rate": 9.134433974380763e-06, "loss": 0.376, "step": 1841 }, { "epoch": 0.8146537426749715, "grad_norm": 0.2798235778018555, "learning_rate": 9.132986945922733e-06, "loss": 0.3652, "step": 1842 }, { "epoch": 0.8150960085504736, "grad_norm": 0.28995136169871777, "learning_rate": 9.131538823740232e-06, "loss": 0.3595, "step": 1843 }, { "epoch": 0.8155382744259757, "grad_norm": 0.29888021918553576, "learning_rate": 9.130089608216478e-06, "loss": 0.3374, "step": 1844 }, { "epoch": 0.8159805403014779, "grad_norm": 0.287291773621424, "learning_rate": 9.128639299734985e-06, "loss": 0.3734, "step": 1845 }, { "epoch": 0.81642280617698, "grad_norm": 0.31114971480795756, "learning_rate": 9.12718789867955e-06, "loss": 0.3925, "step": 1846 }, { "epoch": 0.8168650720524823, "grad_norm": 0.4251250155532863, "learning_rate": 9.125735405434264e-06, "loss": 0.4595, "step": 1847 }, { "epoch": 0.8173073379279844, "grad_norm": 0.33474343142572366, "learning_rate": 9.124281820383502e-06, "loss": 0.4192, "step": 1848 }, { "epoch": 0.8177496038034865, "grad_norm": 0.35081470575754947, "learning_rate": 9.12282714391193e-06, "loss": 0.425, "step": 1849 }, { "epoch": 0.8181918696789887, "grad_norm": 0.6243136062713002, "learning_rate": 9.121371376404506e-06, "loss": 0.4514, "step": 1850 }, { "epoch": 0.8186341355544908, "grad_norm": 0.3056901403377512, "learning_rate": 9.119914518246472e-06, "loss": 0.3717, "step": 1851 }, { "epoch": 0.819076401429993, "grad_norm": 0.29644151072425806, "learning_rate": 9.118456569823362e-06, "loss": 0.3683, "step": 1852 }, { "epoch": 0.8195186673054952, "grad_norm": 0.3061637985851727, "learning_rate": 9.116997531520997e-06, "loss": 0.346, "step": 1853 }, { "epoch": 0.8199609331809973, "grad_norm": 0.4091989536337477, "learning_rate": 9.115537403725483e-06, "loss": 0.4034, "step": 1854 }, { "epoch": 0.8204031990564995, "grad_norm": 0.278276023279047, "learning_rate": 9.114076186823224e-06, "loss": 0.3734, "step": 1855 }, { "epoch": 0.8208454649320016, "grad_norm": 0.6334951962757178, "learning_rate": 9.112613881200903e-06, "loss": 0.3776, "step": 1856 }, { "epoch": 0.8212877308075037, "grad_norm": 0.43659906554624517, "learning_rate": 9.111150487245493e-06, "loss": 0.4376, "step": 1857 }, { "epoch": 0.821729996683006, "grad_norm": 0.35830690055524367, "learning_rate": 9.109686005344258e-06, "loss": 0.3803, "step": 1858 }, { "epoch": 0.8221722625585081, "grad_norm": 0.3588989853196605, "learning_rate": 9.108220435884747e-06, "loss": 0.3868, "step": 1859 }, { "epoch": 0.8226145284340103, "grad_norm": 0.318663363445079, "learning_rate": 9.1067537792548e-06, "loss": 0.3845, "step": 1860 }, { "epoch": 0.8230567943095124, "grad_norm": 0.4387015780596408, "learning_rate": 9.10528603584254e-06, "loss": 0.4487, "step": 1861 }, { "epoch": 0.8234990601850145, "grad_norm": 0.39419252075518396, "learning_rate": 9.103817206036383e-06, "loss": 0.4442, "step": 1862 }, { "epoch": 0.8239413260605167, "grad_norm": 0.37836456337805024, "learning_rate": 9.102347290225027e-06, "loss": 0.4156, "step": 1863 }, { "epoch": 0.8243835919360188, "grad_norm": 0.30834621007693624, "learning_rate": 9.100876288797463e-06, "loss": 0.3723, "step": 1864 }, { "epoch": 0.8248258578115211, "grad_norm": 0.30824810400112357, "learning_rate": 9.099404202142963e-06, "loss": 0.4092, "step": 1865 }, { "epoch": 0.8252681236870232, "grad_norm": 0.2690046580270507, "learning_rate": 9.097931030651093e-06, "loss": 0.3657, "step": 1866 }, { "epoch": 0.8257103895625253, "grad_norm": 0.3237616037827796, "learning_rate": 9.0964567747117e-06, "loss": 0.395, "step": 1867 }, { "epoch": 0.8261526554380275, "grad_norm": 0.3002826927195491, "learning_rate": 9.094981434714922e-06, "loss": 0.3785, "step": 1868 }, { "epoch": 0.8265949213135296, "grad_norm": 0.28936583800959226, "learning_rate": 9.093505011051182e-06, "loss": 0.3701, "step": 1869 }, { "epoch": 0.8270371871890319, "grad_norm": 0.3061757098259707, "learning_rate": 9.092027504111192e-06, "loss": 0.3973, "step": 1870 }, { "epoch": 0.827479453064534, "grad_norm": 0.31221131441386574, "learning_rate": 9.090548914285945e-06, "loss": 0.3817, "step": 1871 }, { "epoch": 0.8279217189400361, "grad_norm": 0.3075699918243119, "learning_rate": 9.089069241966731e-06, "loss": 0.3818, "step": 1872 }, { "epoch": 0.8283639848155383, "grad_norm": 0.2539721080957755, "learning_rate": 9.087588487545115e-06, "loss": 0.3409, "step": 1873 }, { "epoch": 0.8288062506910404, "grad_norm": 0.296652460351408, "learning_rate": 9.086106651412955e-06, "loss": 0.3247, "step": 1874 }, { "epoch": 0.8292485165665426, "grad_norm": 0.4381876657120618, "learning_rate": 9.084623733962393e-06, "loss": 0.4475, "step": 1875 }, { "epoch": 0.8296907824420447, "grad_norm": 0.32296101983977354, "learning_rate": 9.083139735585857e-06, "loss": 0.3575, "step": 1876 }, { "epoch": 0.8301330483175469, "grad_norm": 0.37326693500132807, "learning_rate": 9.081654656676064e-06, "loss": 0.4354, "step": 1877 }, { "epoch": 0.8305753141930491, "grad_norm": 0.2949643506633233, "learning_rate": 9.080168497626014e-06, "loss": 0.3458, "step": 1878 }, { "epoch": 0.8310175800685512, "grad_norm": 0.3303007839560102, "learning_rate": 9.078681258828994e-06, "loss": 0.3793, "step": 1879 }, { "epoch": 0.8314598459440534, "grad_norm": 0.34244281655112935, "learning_rate": 9.077192940678574e-06, "loss": 0.3736, "step": 1880 }, { "epoch": 0.8319021118195555, "grad_norm": 0.30792302057976934, "learning_rate": 9.075703543568617e-06, "loss": 0.3849, "step": 1881 }, { "epoch": 0.8323443776950576, "grad_norm": 0.3197419205360491, "learning_rate": 9.074213067893261e-06, "loss": 0.3724, "step": 1882 }, { "epoch": 0.8327866435705599, "grad_norm": 0.3154420719597125, "learning_rate": 9.072721514046936e-06, "loss": 0.3769, "step": 1883 }, { "epoch": 0.833228909446062, "grad_norm": 0.2944861548568868, "learning_rate": 9.07122888242436e-06, "loss": 0.4045, "step": 1884 }, { "epoch": 0.8336711753215641, "grad_norm": 0.32931569793176785, "learning_rate": 9.06973517342053e-06, "loss": 0.3807, "step": 1885 }, { "epoch": 0.8341134411970663, "grad_norm": 0.37402462152807964, "learning_rate": 9.068240387430729e-06, "loss": 0.4388, "step": 1886 }, { "epoch": 0.8345557070725684, "grad_norm": 0.3062255570487479, "learning_rate": 9.06674452485053e-06, "loss": 0.4222, "step": 1887 }, { "epoch": 0.8349979729480707, "grad_norm": 0.2737950402477365, "learning_rate": 9.065247586075785e-06, "loss": 0.3875, "step": 1888 }, { "epoch": 0.8354402388235728, "grad_norm": 0.3196060856078125, "learning_rate": 9.063749571502633e-06, "loss": 0.3881, "step": 1889 }, { "epoch": 0.8358825046990749, "grad_norm": 0.31030847757959823, "learning_rate": 9.0622504815275e-06, "loss": 0.3723, "step": 1890 }, { "epoch": 0.8363247705745771, "grad_norm": 0.36090045264346043, "learning_rate": 9.060750316547093e-06, "loss": 0.3977, "step": 1891 }, { "epoch": 0.8367670364500792, "grad_norm": 0.27726841300378297, "learning_rate": 9.059249076958407e-06, "loss": 0.358, "step": 1892 }, { "epoch": 0.8372093023255814, "grad_norm": 0.35524009759526903, "learning_rate": 9.057746763158719e-06, "loss": 0.354, "step": 1893 }, { "epoch": 0.8376515682010836, "grad_norm": 0.33421923521243135, "learning_rate": 9.056243375545588e-06, "loss": 0.4179, "step": 1894 }, { "epoch": 0.8380938340765857, "grad_norm": 0.32857855415409953, "learning_rate": 9.054738914516862e-06, "loss": 0.3971, "step": 1895 }, { "epoch": 0.8385360999520879, "grad_norm": 0.2717695736712288, "learning_rate": 9.053233380470671e-06, "loss": 0.3743, "step": 1896 }, { "epoch": 0.83897836582759, "grad_norm": 0.41791841816826986, "learning_rate": 9.051726773805429e-06, "loss": 0.3908, "step": 1897 }, { "epoch": 0.8394206317030922, "grad_norm": 0.3086444205843454, "learning_rate": 9.050219094919832e-06, "loss": 0.4039, "step": 1898 }, { "epoch": 0.8398628975785943, "grad_norm": 0.3401852511346422, "learning_rate": 9.048710344212862e-06, "loss": 0.4025, "step": 1899 }, { "epoch": 0.8403051634540964, "grad_norm": 0.2918128772796915, "learning_rate": 9.047200522083789e-06, "loss": 0.3414, "step": 1900 }, { "epoch": 0.8407474293295987, "grad_norm": 0.3438632709050144, "learning_rate": 9.045689628932155e-06, "loss": 0.3908, "step": 1901 }, { "epoch": 0.8411896952051008, "grad_norm": 0.33256347471360304, "learning_rate": 9.044177665157795e-06, "loss": 0.3708, "step": 1902 }, { "epoch": 0.841631961080603, "grad_norm": 0.30641666475132284, "learning_rate": 9.042664631160824e-06, "loss": 0.3976, "step": 1903 }, { "epoch": 0.8420742269561051, "grad_norm": 0.3025370098616315, "learning_rate": 9.041150527341641e-06, "loss": 0.3722, "step": 1904 }, { "epoch": 0.8425164928316072, "grad_norm": 0.29745923217367065, "learning_rate": 9.039635354100927e-06, "loss": 0.3978, "step": 1905 }, { "epoch": 0.8429587587071095, "grad_norm": 0.3385465277660138, "learning_rate": 9.038119111839648e-06, "loss": 0.4077, "step": 1906 }, { "epoch": 0.8434010245826116, "grad_norm": 0.32926027019553333, "learning_rate": 9.036601800959051e-06, "loss": 0.4106, "step": 1907 }, { "epoch": 0.8438432904581137, "grad_norm": 0.34989224481638564, "learning_rate": 9.035083421860666e-06, "loss": 0.3629, "step": 1908 }, { "epoch": 0.8442855563336159, "grad_norm": 0.3146606151769585, "learning_rate": 9.033563974946306e-06, "loss": 0.3647, "step": 1909 }, { "epoch": 0.844727822209118, "grad_norm": 0.29968865624185576, "learning_rate": 9.032043460618067e-06, "loss": 0.3827, "step": 1910 }, { "epoch": 0.8451700880846202, "grad_norm": 0.2615410014210391, "learning_rate": 9.030521879278326e-06, "loss": 0.3449, "step": 1911 }, { "epoch": 0.8456123539601224, "grad_norm": 0.2884564290845975, "learning_rate": 9.028999231329746e-06, "loss": 0.3776, "step": 1912 }, { "epoch": 0.8460546198356245, "grad_norm": 0.3093242274940163, "learning_rate": 9.027475517175271e-06, "loss": 0.3909, "step": 1913 }, { "epoch": 0.8464968857111267, "grad_norm": 0.31508029406448346, "learning_rate": 9.02595073721812e-06, "loss": 0.3855, "step": 1914 }, { "epoch": 0.8469391515866288, "grad_norm": 0.35267464473661886, "learning_rate": 9.024424891861807e-06, "loss": 0.4159, "step": 1915 }, { "epoch": 0.847381417462131, "grad_norm": 0.5009505925542289, "learning_rate": 9.022897981510115e-06, "loss": 0.3911, "step": 1916 }, { "epoch": 0.8478236833376331, "grad_norm": 0.3244950017096235, "learning_rate": 9.021370006567119e-06, "loss": 0.3903, "step": 1917 }, { "epoch": 0.8482659492131353, "grad_norm": 0.32303448175089444, "learning_rate": 9.019840967437169e-06, "loss": 0.4039, "step": 1918 }, { "epoch": 0.8487082150886375, "grad_norm": 0.30518110533474396, "learning_rate": 9.0183108645249e-06, "loss": 0.3691, "step": 1919 }, { "epoch": 0.8491504809641396, "grad_norm": 0.2913417627740007, "learning_rate": 9.016779698235227e-06, "loss": 0.4011, "step": 1920 }, { "epoch": 0.8495927468396418, "grad_norm": 0.275867140041412, "learning_rate": 9.015247468973347e-06, "loss": 0.3567, "step": 1921 }, { "epoch": 0.8500350127151439, "grad_norm": 0.2764016292971081, "learning_rate": 9.013714177144742e-06, "loss": 0.3663, "step": 1922 }, { "epoch": 0.850477278590646, "grad_norm": 0.32062601652110984, "learning_rate": 9.012179823155165e-06, "loss": 0.3731, "step": 1923 }, { "epoch": 0.8509195444661483, "grad_norm": 0.2942047108811238, "learning_rate": 9.01064440741066e-06, "loss": 0.3991, "step": 1924 }, { "epoch": 0.8513618103416504, "grad_norm": 0.2778818876699875, "learning_rate": 9.00910793031755e-06, "loss": 0.4059, "step": 1925 }, { "epoch": 0.8518040762171526, "grad_norm": 0.28464913319006074, "learning_rate": 9.007570392282436e-06, "loss": 0.3364, "step": 1926 }, { "epoch": 0.8522463420926547, "grad_norm": 0.32939834476951274, "learning_rate": 9.006031793712203e-06, "loss": 0.3924, "step": 1927 }, { "epoch": 0.8526886079681568, "grad_norm": 0.3406803868133715, "learning_rate": 9.004492135014009e-06, "loss": 0.4013, "step": 1928 }, { "epoch": 0.853130873843659, "grad_norm": 0.3377630807757793, "learning_rate": 9.002951416595306e-06, "loss": 0.3984, "step": 1929 }, { "epoch": 0.8535731397191612, "grad_norm": 0.29305034621565074, "learning_rate": 9.001409638863814e-06, "loss": 0.4153, "step": 1930 }, { "epoch": 0.8540154055946634, "grad_norm": 0.35959265081880765, "learning_rate": 8.999866802227539e-06, "loss": 0.3831, "step": 1931 }, { "epoch": 0.8544576714701655, "grad_norm": 0.32326696280490735, "learning_rate": 8.998322907094768e-06, "loss": 0.3971, "step": 1932 }, { "epoch": 0.8548999373456676, "grad_norm": 0.2889545652828919, "learning_rate": 8.996777953874063e-06, "loss": 0.3496, "step": 1933 }, { "epoch": 0.8553422032211698, "grad_norm": 0.3616163293986738, "learning_rate": 8.995231942974273e-06, "loss": 0.3967, "step": 1934 }, { "epoch": 0.8557844690966719, "grad_norm": 0.3404546788881744, "learning_rate": 8.993684874804525e-06, "loss": 0.37, "step": 1935 }, { "epoch": 0.856226734972174, "grad_norm": 0.26299526838240095, "learning_rate": 8.99213674977422e-06, "loss": 0.3423, "step": 1936 }, { "epoch": 0.8566690008476763, "grad_norm": 0.30649465462047043, "learning_rate": 8.990587568293043e-06, "loss": 0.4019, "step": 1937 }, { "epoch": 0.8571112667231784, "grad_norm": 0.3422316100344087, "learning_rate": 8.989037330770961e-06, "loss": 0.3826, "step": 1938 }, { "epoch": 0.8575535325986806, "grad_norm": 0.38091413060398516, "learning_rate": 8.987486037618215e-06, "loss": 0.393, "step": 1939 }, { "epoch": 0.8579957984741827, "grad_norm": 0.2679777156476888, "learning_rate": 8.985933689245332e-06, "loss": 0.4013, "step": 1940 }, { "epoch": 0.8584380643496848, "grad_norm": 0.34076245694396334, "learning_rate": 8.984380286063113e-06, "loss": 0.3865, "step": 1941 }, { "epoch": 0.8588803302251871, "grad_norm": 0.3489635067998301, "learning_rate": 8.982825828482638e-06, "loss": 0.4214, "step": 1942 }, { "epoch": 0.8593225961006892, "grad_norm": 0.3164787127249309, "learning_rate": 8.98127031691527e-06, "loss": 0.3836, "step": 1943 }, { "epoch": 0.8597648619761914, "grad_norm": 0.29542669198427096, "learning_rate": 8.979713751772646e-06, "loss": 0.3737, "step": 1944 }, { "epoch": 0.8602071278516935, "grad_norm": 0.33771993402289613, "learning_rate": 8.978156133466685e-06, "loss": 0.3994, "step": 1945 }, { "epoch": 0.8606493937271956, "grad_norm": 0.30182138717317275, "learning_rate": 8.976597462409587e-06, "loss": 0.3366, "step": 1946 }, { "epoch": 0.8610916596026978, "grad_norm": 0.2844502666719186, "learning_rate": 8.975037739013823e-06, "loss": 0.3963, "step": 1947 }, { "epoch": 0.8615339254782, "grad_norm": 0.30722095301283925, "learning_rate": 8.973476963692151e-06, "loss": 0.4133, "step": 1948 }, { "epoch": 0.8619761913537022, "grad_norm": 0.31037827900040527, "learning_rate": 8.971915136857604e-06, "loss": 0.3879, "step": 1949 }, { "epoch": 0.8624184572292043, "grad_norm": 0.2849614612375569, "learning_rate": 8.970352258923491e-06, "loss": 0.3842, "step": 1950 }, { "epoch": 0.8628607231047064, "grad_norm": 0.3532389624262884, "learning_rate": 8.9687883303034e-06, "loss": 0.4036, "step": 1951 }, { "epoch": 0.8633029889802086, "grad_norm": 0.31056270334317526, "learning_rate": 8.967223351411199e-06, "loss": 0.3692, "step": 1952 }, { "epoch": 0.8637452548557107, "grad_norm": 0.31463981526107726, "learning_rate": 8.965657322661032e-06, "loss": 0.3238, "step": 1953 }, { "epoch": 0.864187520731213, "grad_norm": 0.340964089041336, "learning_rate": 8.964090244467324e-06, "loss": 0.3976, "step": 1954 }, { "epoch": 0.8646297866067151, "grad_norm": 0.26245260393025394, "learning_rate": 8.962522117244777e-06, "loss": 0.3745, "step": 1955 }, { "epoch": 0.8650720524822172, "grad_norm": 0.29006740449135465, "learning_rate": 8.960952941408364e-06, "loss": 0.3457, "step": 1956 }, { "epoch": 0.8655143183577194, "grad_norm": 0.33601794709482624, "learning_rate": 8.959382717373343e-06, "loss": 0.4184, "step": 1957 }, { "epoch": 0.8659565842332215, "grad_norm": 0.379727145415358, "learning_rate": 8.957811445555249e-06, "loss": 0.38, "step": 1958 }, { "epoch": 0.8663988501087236, "grad_norm": 0.30611855557000955, "learning_rate": 8.956239126369888e-06, "loss": 0.3836, "step": 1959 }, { "epoch": 0.8668411159842259, "grad_norm": 0.29348525298269557, "learning_rate": 8.954665760233351e-06, "loss": 0.3735, "step": 1960 }, { "epoch": 0.867283381859728, "grad_norm": 0.609645458926908, "learning_rate": 8.953091347562e-06, "loss": 0.4476, "step": 1961 }, { "epoch": 0.8677256477352302, "grad_norm": 0.29954176580593034, "learning_rate": 8.95151588877248e-06, "loss": 0.3885, "step": 1962 }, { "epoch": 0.8681679136107323, "grad_norm": 0.2757364272552284, "learning_rate": 8.949939384281706e-06, "loss": 0.3787, "step": 1963 }, { "epoch": 0.8686101794862344, "grad_norm": 0.3046504669832418, "learning_rate": 8.948361834506874e-06, "loss": 0.403, "step": 1964 }, { "epoch": 0.8690524453617366, "grad_norm": 0.28548681080497856, "learning_rate": 8.946783239865457e-06, "loss": 0.3977, "step": 1965 }, { "epoch": 0.8694947112372388, "grad_norm": 0.2808395773477441, "learning_rate": 8.945203600775201e-06, "loss": 0.3662, "step": 1966 }, { "epoch": 0.869936977112741, "grad_norm": 0.30971348946062505, "learning_rate": 8.943622917654132e-06, "loss": 0.3731, "step": 1967 }, { "epoch": 0.8703792429882431, "grad_norm": 0.32627146876758334, "learning_rate": 8.94204119092055e-06, "loss": 0.3835, "step": 1968 }, { "epoch": 0.8708215088637452, "grad_norm": 0.3435302534650587, "learning_rate": 8.94045842099303e-06, "loss": 0.4162, "step": 1969 }, { "epoch": 0.8712637747392474, "grad_norm": 0.331607421614189, "learning_rate": 8.938874608290428e-06, "loss": 0.3744, "step": 1970 }, { "epoch": 0.8717060406147495, "grad_norm": 0.2966497209786984, "learning_rate": 8.937289753231875e-06, "loss": 0.3995, "step": 1971 }, { "epoch": 0.8721483064902518, "grad_norm": 0.27848864554763947, "learning_rate": 8.935703856236771e-06, "loss": 0.3695, "step": 1972 }, { "epoch": 0.8725905723657539, "grad_norm": 0.32665873301337395, "learning_rate": 8.9341169177248e-06, "loss": 0.4062, "step": 1973 }, { "epoch": 0.873032838241256, "grad_norm": 0.3692556959750827, "learning_rate": 8.932528938115916e-06, "loss": 0.3816, "step": 1974 }, { "epoch": 0.8734751041167582, "grad_norm": 0.3453614499297661, "learning_rate": 8.93093991783035e-06, "loss": 0.3788, "step": 1975 }, { "epoch": 0.8739173699922603, "grad_norm": 0.2886224145245129, "learning_rate": 8.929349857288614e-06, "loss": 0.3707, "step": 1976 }, { "epoch": 0.8743596358677626, "grad_norm": 0.31234380491574365, "learning_rate": 8.927758756911484e-06, "loss": 0.3809, "step": 1977 }, { "epoch": 0.8748019017432647, "grad_norm": 0.44793263222465746, "learning_rate": 8.926166617120024e-06, "loss": 0.4468, "step": 1978 }, { "epoch": 0.8752441676187668, "grad_norm": 0.31807929549379943, "learning_rate": 8.92457343833556e-06, "loss": 0.4117, "step": 1979 }, { "epoch": 0.875686433494269, "grad_norm": 0.35755392666940344, "learning_rate": 8.922979220979705e-06, "loss": 0.3425, "step": 1980 }, { "epoch": 0.8761286993697711, "grad_norm": 0.28817918020851163, "learning_rate": 8.921383965474338e-06, "loss": 0.3877, "step": 1981 }, { "epoch": 0.8765709652452733, "grad_norm": 0.33751684064044846, "learning_rate": 8.919787672241619e-06, "loss": 0.401, "step": 1982 }, { "epoch": 0.8770132311207754, "grad_norm": 0.2968311722876255, "learning_rate": 8.918190341703977e-06, "loss": 0.3739, "step": 1983 }, { "epoch": 0.8774554969962776, "grad_norm": 0.3016556289615227, "learning_rate": 8.916591974284119e-06, "loss": 0.3849, "step": 1984 }, { "epoch": 0.8778977628717798, "grad_norm": 0.2782014359277583, "learning_rate": 8.914992570405028e-06, "loss": 0.3887, "step": 1985 }, { "epoch": 0.8783400287472819, "grad_norm": 0.30447089472326977, "learning_rate": 8.913392130489955e-06, "loss": 0.3999, "step": 1986 }, { "epoch": 0.878782294622784, "grad_norm": 0.30643133116857096, "learning_rate": 8.911790654962432e-06, "loss": 0.4074, "step": 1987 }, { "epoch": 0.8792245604982862, "grad_norm": 0.32258491955946206, "learning_rate": 8.91018814424626e-06, "loss": 0.3858, "step": 1988 }, { "epoch": 0.8796668263737883, "grad_norm": 0.2611600373506494, "learning_rate": 8.908584598765521e-06, "loss": 0.3955, "step": 1989 }, { "epoch": 0.8801090922492906, "grad_norm": 0.28844112767754304, "learning_rate": 8.90698001894456e-06, "loss": 0.3674, "step": 1990 }, { "epoch": 0.8805513581247927, "grad_norm": 0.3582574081433967, "learning_rate": 8.905374405208006e-06, "loss": 0.3731, "step": 1991 }, { "epoch": 0.8809936240002948, "grad_norm": 0.285784960384249, "learning_rate": 8.903767757980752e-06, "loss": 0.3711, "step": 1992 }, { "epoch": 0.881435889875797, "grad_norm": 0.30944984851353746, "learning_rate": 8.902160077687976e-06, "loss": 0.4006, "step": 1993 }, { "epoch": 0.8818781557512991, "grad_norm": 0.46715548119922123, "learning_rate": 8.900551364755118e-06, "loss": 0.4084, "step": 1994 }, { "epoch": 0.8823204216268014, "grad_norm": 0.31907217541613625, "learning_rate": 8.898941619607902e-06, "loss": 0.3697, "step": 1995 }, { "epoch": 0.8827626875023035, "grad_norm": 0.35016316144222137, "learning_rate": 8.897330842672314e-06, "loss": 0.3826, "step": 1996 }, { "epoch": 0.8832049533778056, "grad_norm": 0.31803898163530187, "learning_rate": 8.895719034374622e-06, "loss": 0.3665, "step": 1997 }, { "epoch": 0.8836472192533078, "grad_norm": 0.2855224847047972, "learning_rate": 8.894106195141363e-06, "loss": 0.3534, "step": 1998 }, { "epoch": 0.8840894851288099, "grad_norm": 0.3075689622680875, "learning_rate": 8.892492325399345e-06, "loss": 0.4004, "step": 1999 }, { "epoch": 0.8845317510043121, "grad_norm": 0.3165736401491783, "learning_rate": 8.890877425575653e-06, "loss": 0.3517, "step": 2000 }, { "epoch": 0.8849740168798143, "grad_norm": 0.32409803036371343, "learning_rate": 8.889261496097644e-06, "loss": 0.3618, "step": 2001 }, { "epoch": 0.8854162827553164, "grad_norm": 0.33554505213507146, "learning_rate": 8.887644537392945e-06, "loss": 0.3881, "step": 2002 }, { "epoch": 0.8858585486308186, "grad_norm": 0.34429567865367733, "learning_rate": 8.886026549889455e-06, "loss": 0.4266, "step": 2003 }, { "epoch": 0.8863008145063207, "grad_norm": 0.2825771161591615, "learning_rate": 8.884407534015349e-06, "loss": 0.3864, "step": 2004 }, { "epoch": 0.8867430803818229, "grad_norm": 0.26012201272884755, "learning_rate": 8.882787490199071e-06, "loss": 0.3569, "step": 2005 }, { "epoch": 0.887185346257325, "grad_norm": 0.30658661564569617, "learning_rate": 8.881166418869338e-06, "loss": 0.3703, "step": 2006 }, { "epoch": 0.8876276121328271, "grad_norm": 0.3503766892664284, "learning_rate": 8.87954432045514e-06, "loss": 0.406, "step": 2007 }, { "epoch": 0.8880698780083294, "grad_norm": 0.3434320199395248, "learning_rate": 8.877921195385735e-06, "loss": 0.3726, "step": 2008 }, { "epoch": 0.8885121438838315, "grad_norm": 0.42683570568596146, "learning_rate": 8.876297044090659e-06, "loss": 0.3987, "step": 2009 }, { "epoch": 0.8889544097593337, "grad_norm": 0.36762183236830576, "learning_rate": 8.874671866999712e-06, "loss": 0.4256, "step": 2010 }, { "epoch": 0.8893966756348358, "grad_norm": 0.3112602896931413, "learning_rate": 8.87304566454297e-06, "loss": 0.3581, "step": 2011 }, { "epoch": 0.8898389415103379, "grad_norm": 0.2932939186766304, "learning_rate": 8.871418437150782e-06, "loss": 0.395, "step": 2012 }, { "epoch": 0.8902812073858402, "grad_norm": 0.3324183533618777, "learning_rate": 8.869790185253766e-06, "loss": 0.4025, "step": 2013 }, { "epoch": 0.8907234732613423, "grad_norm": 0.2954962590067121, "learning_rate": 8.86816090928281e-06, "loss": 0.418, "step": 2014 }, { "epoch": 0.8911657391368444, "grad_norm": 0.298254516468128, "learning_rate": 8.866530609669073e-06, "loss": 0.412, "step": 2015 }, { "epoch": 0.8916080050123466, "grad_norm": 0.2493920181966099, "learning_rate": 8.864899286843987e-06, "loss": 0.3761, "step": 2016 }, { "epoch": 0.8920502708878487, "grad_norm": 0.3335896938946505, "learning_rate": 8.863266941239254e-06, "loss": 0.4386, "step": 2017 }, { "epoch": 0.8924925367633509, "grad_norm": 0.3284901688693498, "learning_rate": 8.861633573286847e-06, "loss": 0.3704, "step": 2018 }, { "epoch": 0.892934802638853, "grad_norm": 0.30686374299053437, "learning_rate": 8.859999183419009e-06, "loss": 0.3934, "step": 2019 }, { "epoch": 0.8933770685143552, "grad_norm": 0.3383123249943022, "learning_rate": 8.85836377206825e-06, "loss": 0.3676, "step": 2020 }, { "epoch": 0.8938193343898574, "grad_norm": 0.30838184117270717, "learning_rate": 8.856727339667357e-06, "loss": 0.4132, "step": 2021 }, { "epoch": 0.8942616002653595, "grad_norm": 0.2887994032430145, "learning_rate": 8.855089886649384e-06, "loss": 0.3762, "step": 2022 }, { "epoch": 0.8947038661408617, "grad_norm": 0.3097049400295465, "learning_rate": 8.853451413447654e-06, "loss": 0.3792, "step": 2023 }, { "epoch": 0.8951461320163638, "grad_norm": 0.3523819293335146, "learning_rate": 8.851811920495761e-06, "loss": 0.4158, "step": 2024 }, { "epoch": 0.895588397891866, "grad_norm": 0.31450091060599156, "learning_rate": 8.850171408227569e-06, "loss": 0.3894, "step": 2025 }, { "epoch": 0.8960306637673682, "grad_norm": 0.30285746196141305, "learning_rate": 8.848529877077212e-06, "loss": 0.3572, "step": 2026 }, { "epoch": 0.8964729296428703, "grad_norm": 0.31109635711579103, "learning_rate": 8.846887327479093e-06, "loss": 0.3426, "step": 2027 }, { "epoch": 0.8969151955183725, "grad_norm": 0.3073239477685359, "learning_rate": 8.845243759867882e-06, "loss": 0.3728, "step": 2028 }, { "epoch": 0.8973574613938746, "grad_norm": 0.29005015841728093, "learning_rate": 8.843599174678526e-06, "loss": 0.3929, "step": 2029 }, { "epoch": 0.8977997272693767, "grad_norm": 0.3265604407565236, "learning_rate": 8.841953572346235e-06, "loss": 0.393, "step": 2030 }, { "epoch": 0.898241993144879, "grad_norm": 0.3501173787328908, "learning_rate": 8.840306953306487e-06, "loss": 0.3841, "step": 2031 }, { "epoch": 0.8986842590203811, "grad_norm": 0.2914337588526636, "learning_rate": 8.838659317995034e-06, "loss": 0.3959, "step": 2032 }, { "epoch": 0.8991265248958833, "grad_norm": 0.2734289686912441, "learning_rate": 8.837010666847892e-06, "loss": 0.3451, "step": 2033 }, { "epoch": 0.8995687907713854, "grad_norm": 0.28964953093074397, "learning_rate": 8.835361000301352e-06, "loss": 0.3599, "step": 2034 }, { "epoch": 0.9000110566468875, "grad_norm": 0.39718695002669213, "learning_rate": 8.833710318791968e-06, "loss": 0.3954, "step": 2035 }, { "epoch": 0.9004533225223897, "grad_norm": 0.3380295105810077, "learning_rate": 8.832058622756564e-06, "loss": 0.3626, "step": 2036 }, { "epoch": 0.9008955883978919, "grad_norm": 0.32713334185541393, "learning_rate": 8.830405912632237e-06, "loss": 0.3347, "step": 2037 }, { "epoch": 0.901337854273394, "grad_norm": 0.31810107542577454, "learning_rate": 8.828752188856343e-06, "loss": 0.4257, "step": 2038 }, { "epoch": 0.9017801201488962, "grad_norm": 0.3577488298335262, "learning_rate": 8.827097451866515e-06, "loss": 0.4136, "step": 2039 }, { "epoch": 0.9022223860243983, "grad_norm": 0.2832458081753316, "learning_rate": 8.82544170210065e-06, "loss": 0.3869, "step": 2040 }, { "epoch": 0.9026646518999005, "grad_norm": 0.3289911963149011, "learning_rate": 8.823784939996915e-06, "loss": 0.4205, "step": 2041 }, { "epoch": 0.9031069177754026, "grad_norm": 0.30760215913774536, "learning_rate": 8.822127165993743e-06, "loss": 0.3401, "step": 2042 }, { "epoch": 0.9035491836509048, "grad_norm": 0.37626135934025107, "learning_rate": 8.820468380529837e-06, "loss": 0.4322, "step": 2043 }, { "epoch": 0.903991449526407, "grad_norm": 0.33068177852711, "learning_rate": 8.818808584044163e-06, "loss": 0.3839, "step": 2044 }, { "epoch": 0.9044337154019091, "grad_norm": 0.2916690190370745, "learning_rate": 8.817147776975961e-06, "loss": 0.3656, "step": 2045 }, { "epoch": 0.9048759812774113, "grad_norm": 0.28049041509589007, "learning_rate": 8.815485959764733e-06, "loss": 0.3906, "step": 2046 }, { "epoch": 0.9053182471529134, "grad_norm": 0.31256794206109967, "learning_rate": 8.813823132850253e-06, "loss": 0.4073, "step": 2047 }, { "epoch": 0.9057605130284155, "grad_norm": 0.27382296644892784, "learning_rate": 8.812159296672558e-06, "loss": 0.4198, "step": 2048 }, { "epoch": 0.9062027789039178, "grad_norm": 0.3407400969932986, "learning_rate": 8.810494451671954e-06, "loss": 0.4274, "step": 2049 }, { "epoch": 0.9066450447794199, "grad_norm": 0.2986860810290525, "learning_rate": 8.808828598289014e-06, "loss": 0.3918, "step": 2050 }, { "epoch": 0.9070873106549221, "grad_norm": 0.3331211609468431, "learning_rate": 8.807161736964578e-06, "loss": 0.3761, "step": 2051 }, { "epoch": 0.9075295765304242, "grad_norm": 0.29193988080787964, "learning_rate": 8.805493868139753e-06, "loss": 0.3889, "step": 2052 }, { "epoch": 0.9079718424059263, "grad_norm": 0.2749071838005886, "learning_rate": 8.80382499225591e-06, "loss": 0.345, "step": 2053 }, { "epoch": 0.9084141082814285, "grad_norm": 0.32831762288781396, "learning_rate": 8.802155109754691e-06, "loss": 0.3829, "step": 2054 }, { "epoch": 0.9088563741569307, "grad_norm": 0.30585284337012114, "learning_rate": 8.800484221077999e-06, "loss": 0.4026, "step": 2055 }, { "epoch": 0.9092986400324329, "grad_norm": 0.31594517807218486, "learning_rate": 8.79881232666801e-06, "loss": 0.4043, "step": 2056 }, { "epoch": 0.909740905907935, "grad_norm": 0.4245025236852918, "learning_rate": 8.797139426967156e-06, "loss": 0.3845, "step": 2057 }, { "epoch": 0.9101831717834371, "grad_norm": 0.3887360885034523, "learning_rate": 8.79546552241815e-06, "loss": 0.3648, "step": 2058 }, { "epoch": 0.9106254376589393, "grad_norm": 0.3157099130971035, "learning_rate": 8.793790613463956e-06, "loss": 0.3747, "step": 2059 }, { "epoch": 0.9110677035344414, "grad_norm": 0.3299069075607508, "learning_rate": 8.792114700547811e-06, "loss": 0.4356, "step": 2060 }, { "epoch": 0.9115099694099437, "grad_norm": 0.323332984517472, "learning_rate": 8.790437784113218e-06, "loss": 0.3934, "step": 2061 }, { "epoch": 0.9119522352854458, "grad_norm": 0.32593447759315275, "learning_rate": 8.788759864603945e-06, "loss": 0.3915, "step": 2062 }, { "epoch": 0.9123945011609479, "grad_norm": 0.2754177746901712, "learning_rate": 8.787080942464024e-06, "loss": 0.362, "step": 2063 }, { "epoch": 0.9128367670364501, "grad_norm": 0.2839258860020492, "learning_rate": 8.785401018137752e-06, "loss": 0.3663, "step": 2064 }, { "epoch": 0.9132790329119522, "grad_norm": 0.2557281736710143, "learning_rate": 8.783720092069691e-06, "loss": 0.3443, "step": 2065 }, { "epoch": 0.9137212987874543, "grad_norm": 0.25966141201025855, "learning_rate": 8.782038164704675e-06, "loss": 0.3801, "step": 2066 }, { "epoch": 0.9141635646629566, "grad_norm": 0.26081903962096664, "learning_rate": 8.780355236487792e-06, "loss": 0.3914, "step": 2067 }, { "epoch": 0.9146058305384587, "grad_norm": 0.2563361579189201, "learning_rate": 8.778671307864401e-06, "loss": 0.3963, "step": 2068 }, { "epoch": 0.9150480964139609, "grad_norm": 0.2568396783261307, "learning_rate": 8.776986379280128e-06, "loss": 0.3455, "step": 2069 }, { "epoch": 0.915490362289463, "grad_norm": 0.6931891239895486, "learning_rate": 8.77530045118086e-06, "loss": 0.3747, "step": 2070 }, { "epoch": 0.9159326281649651, "grad_norm": 0.2847849970538579, "learning_rate": 8.773613524012746e-06, "loss": 0.3978, "step": 2071 }, { "epoch": 0.9163748940404673, "grad_norm": 0.30195770196343164, "learning_rate": 8.771925598222203e-06, "loss": 0.3768, "step": 2072 }, { "epoch": 0.9168171599159695, "grad_norm": 0.3406424972110172, "learning_rate": 8.770236674255913e-06, "loss": 0.3631, "step": 2073 }, { "epoch": 0.9172594257914717, "grad_norm": 0.2633140184697449, "learning_rate": 8.768546752560823e-06, "loss": 0.368, "step": 2074 }, { "epoch": 0.9177016916669738, "grad_norm": 0.2893617682610834, "learning_rate": 8.76685583358414e-06, "loss": 0.3718, "step": 2075 }, { "epoch": 0.9181439575424759, "grad_norm": 0.26042534940435275, "learning_rate": 8.765163917773337e-06, "loss": 0.372, "step": 2076 }, { "epoch": 0.9185862234179781, "grad_norm": 0.37927156940824963, "learning_rate": 8.763471005576147e-06, "loss": 0.3476, "step": 2077 }, { "epoch": 0.9190284892934802, "grad_norm": 0.2758019038978435, "learning_rate": 8.761777097440577e-06, "loss": 0.3962, "step": 2078 }, { "epoch": 0.9194707551689825, "grad_norm": 0.3155152748845655, "learning_rate": 8.760082193814886e-06, "loss": 0.392, "step": 2079 }, { "epoch": 0.9199130210444846, "grad_norm": 0.2781645974779366, "learning_rate": 8.758386295147605e-06, "loss": 0.4239, "step": 2080 }, { "epoch": 0.9203552869199867, "grad_norm": 0.29179105142739287, "learning_rate": 8.75668940188752e-06, "loss": 0.367, "step": 2081 }, { "epoch": 0.9207975527954889, "grad_norm": 0.3004959269728948, "learning_rate": 8.754991514483688e-06, "loss": 0.3442, "step": 2082 }, { "epoch": 0.921239818670991, "grad_norm": 0.31922165324149737, "learning_rate": 8.753292633385427e-06, "loss": 0.3982, "step": 2083 }, { "epoch": 0.9216820845464933, "grad_norm": 0.3049791164297285, "learning_rate": 8.751592759042313e-06, "loss": 0.4168, "step": 2084 }, { "epoch": 0.9221243504219954, "grad_norm": 0.2836402142839463, "learning_rate": 8.749891891904192e-06, "loss": 0.38, "step": 2085 }, { "epoch": 0.9225666162974975, "grad_norm": 0.3356931799065305, "learning_rate": 8.748190032421168e-06, "loss": 0.365, "step": 2086 }, { "epoch": 0.9230088821729997, "grad_norm": 0.311977996252479, "learning_rate": 8.746487181043612e-06, "loss": 0.3728, "step": 2087 }, { "epoch": 0.9234511480485018, "grad_norm": 0.391379342178792, "learning_rate": 8.74478333822215e-06, "loss": 0.39, "step": 2088 }, { "epoch": 0.9238934139240039, "grad_norm": 0.2577861505796084, "learning_rate": 8.743078504407675e-06, "loss": 0.3927, "step": 2089 }, { "epoch": 0.9243356797995061, "grad_norm": 0.28000807785418486, "learning_rate": 8.741372680051348e-06, "loss": 0.3681, "step": 2090 }, { "epoch": 0.9247779456750083, "grad_norm": 0.312903225379198, "learning_rate": 8.739665865604581e-06, "loss": 0.3999, "step": 2091 }, { "epoch": 0.9252202115505105, "grad_norm": 0.33481234392502773, "learning_rate": 8.737958061519054e-06, "loss": 0.3538, "step": 2092 }, { "epoch": 0.9256624774260126, "grad_norm": 0.3152225987891274, "learning_rate": 8.73624926824671e-06, "loss": 0.3621, "step": 2093 }, { "epoch": 0.9261047433015147, "grad_norm": 0.2896189631430369, "learning_rate": 8.73453948623975e-06, "loss": 0.3673, "step": 2094 }, { "epoch": 0.9265470091770169, "grad_norm": 0.31213064790755124, "learning_rate": 8.73282871595064e-06, "loss": 0.3714, "step": 2095 }, { "epoch": 0.926989275052519, "grad_norm": 0.2696152039518701, "learning_rate": 8.731116957832106e-06, "loss": 0.3736, "step": 2096 }, { "epoch": 0.9274315409280213, "grad_norm": 0.3174210446067657, "learning_rate": 8.729404212337135e-06, "loss": 0.3611, "step": 2097 }, { "epoch": 0.9278738068035234, "grad_norm": 0.2678737649839693, "learning_rate": 8.727690479918978e-06, "loss": 0.3679, "step": 2098 }, { "epoch": 0.9283160726790255, "grad_norm": 0.2621284382565619, "learning_rate": 8.725975761031142e-06, "loss": 0.3325, "step": 2099 }, { "epoch": 0.9287583385545277, "grad_norm": 0.3053062928692742, "learning_rate": 8.724260056127398e-06, "loss": 0.3864, "step": 2100 }, { "epoch": 0.9292006044300298, "grad_norm": 0.3720622016003881, "learning_rate": 8.722543365661781e-06, "loss": 0.4095, "step": 2101 }, { "epoch": 0.929642870305532, "grad_norm": 0.3053762353117937, "learning_rate": 8.720825690088582e-06, "loss": 0.3782, "step": 2102 }, { "epoch": 0.9300851361810342, "grad_norm": 0.3142448165804328, "learning_rate": 8.719107029862355e-06, "loss": 0.3973, "step": 2103 }, { "epoch": 0.9305274020565363, "grad_norm": 0.4500136741765898, "learning_rate": 8.717387385437914e-06, "loss": 0.3988, "step": 2104 }, { "epoch": 0.9309696679320385, "grad_norm": 0.2855030570838903, "learning_rate": 8.715666757270334e-06, "loss": 0.4052, "step": 2105 }, { "epoch": 0.9314119338075406, "grad_norm": 0.35239015253824135, "learning_rate": 8.713945145814948e-06, "loss": 0.3672, "step": 2106 }, { "epoch": 0.9318541996830428, "grad_norm": 0.2979922990531793, "learning_rate": 8.712222551527351e-06, "loss": 0.3541, "step": 2107 }, { "epoch": 0.932296465558545, "grad_norm": 0.2781378386771119, "learning_rate": 8.710498974863402e-06, "loss": 0.319, "step": 2108 }, { "epoch": 0.9327387314340471, "grad_norm": 0.2840829150622574, "learning_rate": 8.708774416279213e-06, "loss": 0.381, "step": 2109 }, { "epoch": 0.9331809973095493, "grad_norm": 0.28751091324640166, "learning_rate": 8.70704887623116e-06, "loss": 0.4014, "step": 2110 }, { "epoch": 0.9336232631850514, "grad_norm": 0.3002619626022667, "learning_rate": 8.705322355175878e-06, "loss": 0.4157, "step": 2111 }, { "epoch": 0.9340655290605536, "grad_norm": 0.33712514529667265, "learning_rate": 8.703594853570259e-06, "loss": 0.3963, "step": 2112 }, { "epoch": 0.9345077949360557, "grad_norm": 0.373600774134628, "learning_rate": 8.701866371871461e-06, "loss": 0.4249, "step": 2113 }, { "epoch": 0.9349500608115578, "grad_norm": 0.29202695761639713, "learning_rate": 8.700136910536895e-06, "loss": 0.3879, "step": 2114 }, { "epoch": 0.9353923266870601, "grad_norm": 0.3052863630111616, "learning_rate": 8.69840647002423e-06, "loss": 0.3827, "step": 2115 }, { "epoch": 0.9358345925625622, "grad_norm": 0.28828607485417196, "learning_rate": 8.696675050791403e-06, "loss": 0.3606, "step": 2116 }, { "epoch": 0.9362768584380643, "grad_norm": 0.30964492273784666, "learning_rate": 8.694942653296604e-06, "loss": 0.3402, "step": 2117 }, { "epoch": 0.9367191243135665, "grad_norm": 0.2890704907290182, "learning_rate": 8.693209277998278e-06, "loss": 0.3853, "step": 2118 }, { "epoch": 0.9371613901890686, "grad_norm": 0.2455221881827102, "learning_rate": 8.691474925355136e-06, "loss": 0.345, "step": 2119 }, { "epoch": 0.9376036560645709, "grad_norm": 0.2839111711450535, "learning_rate": 8.689739595826146e-06, "loss": 0.3653, "step": 2120 }, { "epoch": 0.938045921940073, "grad_norm": 0.35802875663097755, "learning_rate": 8.688003289870533e-06, "loss": 0.3862, "step": 2121 }, { "epoch": 0.9384881878155751, "grad_norm": 0.303279275167374, "learning_rate": 8.686266007947778e-06, "loss": 0.4009, "step": 2122 }, { "epoch": 0.9389304536910773, "grad_norm": 0.2771835833986532, "learning_rate": 8.684527750517626e-06, "loss": 0.3805, "step": 2123 }, { "epoch": 0.9393727195665794, "grad_norm": 0.3128220426384494, "learning_rate": 8.682788518040075e-06, "loss": 0.3781, "step": 2124 }, { "epoch": 0.9398149854420816, "grad_norm": 0.2907259291628217, "learning_rate": 8.681048310975387e-06, "loss": 0.3958, "step": 2125 }, { "epoch": 0.9402572513175838, "grad_norm": 0.27036098214615395, "learning_rate": 8.679307129784073e-06, "loss": 0.3984, "step": 2126 }, { "epoch": 0.9406995171930859, "grad_norm": 0.38558543179235155, "learning_rate": 8.677564974926909e-06, "loss": 0.387, "step": 2127 }, { "epoch": 0.9411417830685881, "grad_norm": 0.4392284044553016, "learning_rate": 8.67582184686493e-06, "loss": 0.4463, "step": 2128 }, { "epoch": 0.9415840489440902, "grad_norm": 0.3272129981825824, "learning_rate": 8.674077746059418e-06, "loss": 0.3885, "step": 2129 }, { "epoch": 0.9420263148195924, "grad_norm": 0.3336801734119711, "learning_rate": 8.672332672971926e-06, "loss": 0.3555, "step": 2130 }, { "epoch": 0.9424685806950945, "grad_norm": 0.2973149527531389, "learning_rate": 8.670586628064254e-06, "loss": 0.3988, "step": 2131 }, { "epoch": 0.9429108465705967, "grad_norm": 0.4002241989166015, "learning_rate": 8.668839611798465e-06, "loss": 0.3545, "step": 2132 }, { "epoch": 0.9433531124460989, "grad_norm": 0.29949702832777403, "learning_rate": 8.667091624636877e-06, "loss": 0.3542, "step": 2133 }, { "epoch": 0.943795378321601, "grad_norm": 0.310995691192476, "learning_rate": 8.665342667042063e-06, "loss": 0.405, "step": 2134 }, { "epoch": 0.9442376441971032, "grad_norm": 0.3018867066081581, "learning_rate": 8.663592739476855e-06, "loss": 0.3447, "step": 2135 }, { "epoch": 0.9446799100726053, "grad_norm": 0.27134928308987094, "learning_rate": 8.661841842404345e-06, "loss": 0.4222, "step": 2136 }, { "epoch": 0.9451221759481074, "grad_norm": 0.2779026981882561, "learning_rate": 8.660089976287875e-06, "loss": 0.3634, "step": 2137 }, { "epoch": 0.9455644418236097, "grad_norm": 0.29820906523382534, "learning_rate": 8.658337141591047e-06, "loss": 0.396, "step": 2138 }, { "epoch": 0.9460067076991118, "grad_norm": 0.3163805370381336, "learning_rate": 8.656583338777718e-06, "loss": 0.428, "step": 2139 }, { "epoch": 0.946448973574614, "grad_norm": 0.33025665296769546, "learning_rate": 8.654828568312003e-06, "loss": 0.4036, "step": 2140 }, { "epoch": 0.9468912394501161, "grad_norm": 0.29172032513630286, "learning_rate": 8.653072830658273e-06, "loss": 0.3245, "step": 2141 }, { "epoch": 0.9473335053256182, "grad_norm": 0.28483814843856275, "learning_rate": 8.651316126281152e-06, "loss": 0.3858, "step": 2142 }, { "epoch": 0.9477757712011204, "grad_norm": 0.2567248783541275, "learning_rate": 8.649558455645524e-06, "loss": 0.3855, "step": 2143 }, { "epoch": 0.9482180370766226, "grad_norm": 0.413906490136942, "learning_rate": 8.647799819216524e-06, "loss": 0.3986, "step": 2144 }, { "epoch": 0.9486603029521247, "grad_norm": 0.3106519049952046, "learning_rate": 8.646040217459545e-06, "loss": 0.4077, "step": 2145 }, { "epoch": 0.9491025688276269, "grad_norm": 0.354135887472457, "learning_rate": 8.64427965084024e-06, "loss": 0.4146, "step": 2146 }, { "epoch": 0.949544834703129, "grad_norm": 0.3078221624330456, "learning_rate": 8.64251811982451e-06, "loss": 0.3544, "step": 2147 }, { "epoch": 0.9499871005786312, "grad_norm": 0.3367112203609712, "learning_rate": 8.640755624878512e-06, "loss": 0.3922, "step": 2148 }, { "epoch": 0.9504293664541333, "grad_norm": 0.3466995494725965, "learning_rate": 8.638992166468662e-06, "loss": 0.3948, "step": 2149 }, { "epoch": 0.9508716323296355, "grad_norm": 0.28153930561920454, "learning_rate": 8.637227745061632e-06, "loss": 0.3342, "step": 2150 }, { "epoch": 0.9513138982051377, "grad_norm": 0.3053265504228329, "learning_rate": 8.63546236112434e-06, "loss": 0.3878, "step": 2151 }, { "epoch": 0.9517561640806398, "grad_norm": 0.3238299044686391, "learning_rate": 8.63369601512397e-06, "loss": 0.3905, "step": 2152 }, { "epoch": 0.952198429956142, "grad_norm": 0.32080447545178353, "learning_rate": 8.631928707527955e-06, "loss": 0.4053, "step": 2153 }, { "epoch": 0.9526406958316441, "grad_norm": 0.30880438036526153, "learning_rate": 8.63016043880398e-06, "loss": 0.4363, "step": 2154 }, { "epoch": 0.9530829617071462, "grad_norm": 0.324129041877565, "learning_rate": 8.628391209419988e-06, "loss": 0.3894, "step": 2155 }, { "epoch": 0.9535252275826485, "grad_norm": 0.33756747610350113, "learning_rate": 8.626621019844177e-06, "loss": 0.4143, "step": 2156 }, { "epoch": 0.9539674934581506, "grad_norm": 0.3403190361511969, "learning_rate": 8.624849870544994e-06, "loss": 0.4129, "step": 2157 }, { "epoch": 0.9544097593336528, "grad_norm": 0.3396919170527439, "learning_rate": 8.623077761991147e-06, "loss": 0.3654, "step": 2158 }, { "epoch": 0.9548520252091549, "grad_norm": 0.30308004220738244, "learning_rate": 8.621304694651594e-06, "loss": 0.3909, "step": 2159 }, { "epoch": 0.955294291084657, "grad_norm": 0.31007514515226037, "learning_rate": 8.619530668995546e-06, "loss": 0.4153, "step": 2160 }, { "epoch": 0.9557365569601592, "grad_norm": 0.2857829477743856, "learning_rate": 8.617755685492467e-06, "loss": 0.3546, "step": 2161 }, { "epoch": 0.9561788228356614, "grad_norm": 0.2936010779259119, "learning_rate": 8.615979744612079e-06, "loss": 0.3615, "step": 2162 }, { "epoch": 0.9566210887111636, "grad_norm": 0.31566754773584443, "learning_rate": 8.614202846824354e-06, "loss": 0.3473, "step": 2163 }, { "epoch": 0.9570633545866657, "grad_norm": 0.371859906784558, "learning_rate": 8.612424992599515e-06, "loss": 0.4044, "step": 2164 }, { "epoch": 0.9575056204621678, "grad_norm": 0.31939517511160426, "learning_rate": 8.610646182408043e-06, "loss": 0.394, "step": 2165 }, { "epoch": 0.95794788633767, "grad_norm": 0.3177495139681432, "learning_rate": 8.60886641672067e-06, "loss": 0.3946, "step": 2166 }, { "epoch": 0.9583901522131721, "grad_norm": 0.29746608454496226, "learning_rate": 8.607085696008379e-06, "loss": 0.3723, "step": 2167 }, { "epoch": 0.9588324180886743, "grad_norm": 0.2969848851433556, "learning_rate": 8.60530402074241e-06, "loss": 0.3903, "step": 2168 }, { "epoch": 0.9592746839641765, "grad_norm": 0.3794788495540187, "learning_rate": 8.60352139139425e-06, "loss": 0.4101, "step": 2169 }, { "epoch": 0.9597169498396786, "grad_norm": 0.2995345480947232, "learning_rate": 8.601737808435644e-06, "loss": 0.3779, "step": 2170 }, { "epoch": 0.9601592157151808, "grad_norm": 0.32721097749046885, "learning_rate": 8.599953272338584e-06, "loss": 0.3865, "step": 2171 }, { "epoch": 0.9606014815906829, "grad_norm": 0.2791396940064978, "learning_rate": 8.598167783575319e-06, "loss": 0.4192, "step": 2172 }, { "epoch": 0.961043747466185, "grad_norm": 0.3132002324036676, "learning_rate": 8.59638134261835e-06, "loss": 0.3598, "step": 2173 }, { "epoch": 0.9614860133416873, "grad_norm": 0.3121125419002714, "learning_rate": 8.594593949940423e-06, "loss": 0.3829, "step": 2174 }, { "epoch": 0.9619282792171894, "grad_norm": 0.28048208372930433, "learning_rate": 8.592805606014547e-06, "loss": 0.378, "step": 2175 }, { "epoch": 0.9623705450926916, "grad_norm": 0.3114519639121409, "learning_rate": 8.591016311313971e-06, "loss": 0.3942, "step": 2176 }, { "epoch": 0.9628128109681937, "grad_norm": 0.2901534575243393, "learning_rate": 8.589226066312204e-06, "loss": 0.3715, "step": 2177 }, { "epoch": 0.9632550768436958, "grad_norm": 0.3114403401155004, "learning_rate": 8.587434871483005e-06, "loss": 0.3504, "step": 2178 }, { "epoch": 0.963697342719198, "grad_norm": 0.5869376153633215, "learning_rate": 8.585642727300381e-06, "loss": 0.3606, "step": 2179 }, { "epoch": 0.9641396085947002, "grad_norm": 0.32043883297819714, "learning_rate": 8.583849634238592e-06, "loss": 0.4386, "step": 2180 }, { "epoch": 0.9645818744702024, "grad_norm": 0.30744656115183805, "learning_rate": 8.582055592772151e-06, "loss": 0.398, "step": 2181 }, { "epoch": 0.9650241403457045, "grad_norm": 0.2697581525422787, "learning_rate": 8.580260603375821e-06, "loss": 0.3751, "step": 2182 }, { "epoch": 0.9654664062212066, "grad_norm": 0.27508528647113323, "learning_rate": 8.578464666524615e-06, "loss": 0.378, "step": 2183 }, { "epoch": 0.9659086720967088, "grad_norm": 0.3114012609290744, "learning_rate": 8.576667782693797e-06, "loss": 0.3958, "step": 2184 }, { "epoch": 0.9663509379722109, "grad_norm": 0.335145280200944, "learning_rate": 8.57486995235888e-06, "loss": 0.3792, "step": 2185 }, { "epoch": 0.9667932038477132, "grad_norm": 0.29518401185448845, "learning_rate": 8.573071175995628e-06, "loss": 0.3818, "step": 2186 }, { "epoch": 0.9672354697232153, "grad_norm": 0.30983870253991186, "learning_rate": 8.571271454080061e-06, "loss": 0.3942, "step": 2187 }, { "epoch": 0.9676777355987174, "grad_norm": 0.2710878392468706, "learning_rate": 8.569470787088444e-06, "loss": 0.3912, "step": 2188 }, { "epoch": 0.9681200014742196, "grad_norm": 0.39404075070529054, "learning_rate": 8.567669175497288e-06, "loss": 0.3698, "step": 2189 }, { "epoch": 0.9685622673497217, "grad_norm": 0.3438711386336347, "learning_rate": 8.565866619783365e-06, "loss": 0.3662, "step": 2190 }, { "epoch": 0.969004533225224, "grad_norm": 0.3081681384001625, "learning_rate": 8.564063120423684e-06, "loss": 0.4084, "step": 2191 }, { "epoch": 0.9694467991007261, "grad_norm": 0.2857482570228498, "learning_rate": 8.562258677895516e-06, "loss": 0.3715, "step": 2192 }, { "epoch": 0.9698890649762282, "grad_norm": 0.3111178822103783, "learning_rate": 8.560453292676374e-06, "loss": 0.4118, "step": 2193 }, { "epoch": 0.9703313308517304, "grad_norm": 0.3076168359228794, "learning_rate": 8.558646965244022e-06, "loss": 0.3917, "step": 2194 }, { "epoch": 0.9707735967272325, "grad_norm": 0.26980079401102075, "learning_rate": 8.556839696076474e-06, "loss": 0.3634, "step": 2195 }, { "epoch": 0.9712158626027346, "grad_norm": 0.47193808899355816, "learning_rate": 8.555031485651993e-06, "loss": 0.3996, "step": 2196 }, { "epoch": 0.9716581284782368, "grad_norm": 0.33490200367259115, "learning_rate": 8.553222334449092e-06, "loss": 0.3992, "step": 2197 }, { "epoch": 0.972100394353739, "grad_norm": 0.2991150941441964, "learning_rate": 8.551412242946529e-06, "loss": 0.3328, "step": 2198 }, { "epoch": 0.9725426602292412, "grad_norm": 0.26968750421751037, "learning_rate": 8.549601211623316e-06, "loss": 0.3642, "step": 2199 }, { "epoch": 0.9729849261047433, "grad_norm": 0.3321427572785379, "learning_rate": 8.547789240958713e-06, "loss": 0.3981, "step": 2200 }, { "epoch": 0.9734271919802454, "grad_norm": 0.2926364559245285, "learning_rate": 8.545976331432224e-06, "loss": 0.3205, "step": 2201 }, { "epoch": 0.9738694578557476, "grad_norm": 0.31693803305516366, "learning_rate": 8.544162483523606e-06, "loss": 0.3838, "step": 2202 }, { "epoch": 0.9743117237312497, "grad_norm": 0.6219323148912522, "learning_rate": 8.542347697712863e-06, "loss": 0.4027, "step": 2203 }, { "epoch": 0.974753989606752, "grad_norm": 0.2644390870273134, "learning_rate": 8.540531974480246e-06, "loss": 0.4018, "step": 2204 }, { "epoch": 0.9751962554822541, "grad_norm": 0.3054001248279244, "learning_rate": 8.538715314306257e-06, "loss": 0.3464, "step": 2205 }, { "epoch": 0.9756385213577562, "grad_norm": 0.3642962840580337, "learning_rate": 8.536897717671644e-06, "loss": 0.3735, "step": 2206 }, { "epoch": 0.9760807872332584, "grad_norm": 0.3142324473170217, "learning_rate": 8.535079185057402e-06, "loss": 0.3977, "step": 2207 }, { "epoch": 0.9765230531087605, "grad_norm": 0.29528451566988667, "learning_rate": 8.533259716944775e-06, "loss": 0.3703, "step": 2208 }, { "epoch": 0.9769653189842628, "grad_norm": 0.2776513836424747, "learning_rate": 8.531439313815252e-06, "loss": 0.3765, "step": 2209 }, { "epoch": 0.9774075848597649, "grad_norm": 0.33033914275414233, "learning_rate": 8.529617976150574e-06, "loss": 0.3975, "step": 2210 }, { "epoch": 0.977849850735267, "grad_norm": 0.3141409331811595, "learning_rate": 8.52779570443273e-06, "loss": 0.3703, "step": 2211 }, { "epoch": 0.9782921166107692, "grad_norm": 0.25900413946139994, "learning_rate": 8.525972499143947e-06, "loss": 0.3767, "step": 2212 }, { "epoch": 0.9787343824862713, "grad_norm": 0.2894979496694246, "learning_rate": 8.52414836076671e-06, "loss": 0.3685, "step": 2213 }, { "epoch": 0.9791766483617735, "grad_norm": 0.28555411612788967, "learning_rate": 8.52232328978374e-06, "loss": 0.3701, "step": 2214 }, { "epoch": 0.9796189142372757, "grad_norm": 0.30492853948308524, "learning_rate": 8.520497286678017e-06, "loss": 0.3682, "step": 2215 }, { "epoch": 0.9800611801127778, "grad_norm": 0.3156630502632115, "learning_rate": 8.51867035193276e-06, "loss": 0.413, "step": 2216 }, { "epoch": 0.98050344598828, "grad_norm": 0.27190841644403774, "learning_rate": 8.516842486031436e-06, "loss": 0.4002, "step": 2217 }, { "epoch": 0.9809457118637821, "grad_norm": 0.2704631409978324, "learning_rate": 8.515013689457756e-06, "loss": 0.3448, "step": 2218 }, { "epoch": 0.9813879777392842, "grad_norm": 0.32783820979915357, "learning_rate": 8.513183962695684e-06, "loss": 0.3881, "step": 2219 }, { "epoch": 0.9818302436147864, "grad_norm": 0.27619383763308913, "learning_rate": 8.511353306229422e-06, "loss": 0.3918, "step": 2220 }, { "epoch": 0.9822725094902885, "grad_norm": 0.29402917407370094, "learning_rate": 8.509521720543423e-06, "loss": 0.3687, "step": 2221 }, { "epoch": 0.9827147753657908, "grad_norm": 0.35062458222488935, "learning_rate": 8.50768920612239e-06, "loss": 0.3891, "step": 2222 }, { "epoch": 0.9831570412412929, "grad_norm": 0.35406007051656135, "learning_rate": 8.50585576345126e-06, "loss": 0.3952, "step": 2223 }, { "epoch": 0.983599307116795, "grad_norm": 0.31819298796733525, "learning_rate": 8.504021393015223e-06, "loss": 0.4033, "step": 2224 }, { "epoch": 0.9840415729922972, "grad_norm": 0.2742095905694052, "learning_rate": 8.502186095299716e-06, "loss": 0.3511, "step": 2225 }, { "epoch": 0.9844838388677993, "grad_norm": 0.3890705436970354, "learning_rate": 8.50034987079042e-06, "loss": 0.4345, "step": 2226 }, { "epoch": 0.9849261047433016, "grad_norm": 0.2585204505741853, "learning_rate": 8.49851271997326e-06, "loss": 0.361, "step": 2227 }, { "epoch": 0.9853683706188037, "grad_norm": 0.3162826362449364, "learning_rate": 8.496674643334403e-06, "loss": 0.3809, "step": 2228 }, { "epoch": 0.9858106364943058, "grad_norm": 0.3255636685586391, "learning_rate": 8.49483564136027e-06, "loss": 0.4344, "step": 2229 }, { "epoch": 0.986252902369808, "grad_norm": 0.290733275672023, "learning_rate": 8.492995714537519e-06, "loss": 0.3814, "step": 2230 }, { "epoch": 0.9866951682453101, "grad_norm": 0.29711194852797085, "learning_rate": 8.491154863353053e-06, "loss": 0.3934, "step": 2231 }, { "epoch": 0.9871374341208123, "grad_norm": 0.23097881009339527, "learning_rate": 8.489313088294027e-06, "loss": 0.3572, "step": 2232 }, { "epoch": 0.9875796999963145, "grad_norm": 0.33244019429990523, "learning_rate": 8.487470389847831e-06, "loss": 0.3678, "step": 2233 }, { "epoch": 0.9880219658718166, "grad_norm": 0.3603219731064031, "learning_rate": 8.485626768502104e-06, "loss": 0.3565, "step": 2234 }, { "epoch": 0.9884642317473188, "grad_norm": 0.27650428990994896, "learning_rate": 8.483782224744732e-06, "loss": 0.4027, "step": 2235 }, { "epoch": 0.9889064976228209, "grad_norm": 0.3966985489553557, "learning_rate": 8.481936759063838e-06, "loss": 0.3818, "step": 2236 }, { "epoch": 0.9893487634983231, "grad_norm": 0.31061047161629746, "learning_rate": 8.480090371947796e-06, "loss": 0.3627, "step": 2237 }, { "epoch": 0.9897910293738252, "grad_norm": 0.3477331007485318, "learning_rate": 8.478243063885221e-06, "loss": 0.3896, "step": 2238 }, { "epoch": 0.9902332952493273, "grad_norm": 0.28152357246023313, "learning_rate": 8.47639483536497e-06, "loss": 0.3602, "step": 2239 }, { "epoch": 0.9906755611248296, "grad_norm": 0.3101297664702079, "learning_rate": 8.474545686876144e-06, "loss": 0.3644, "step": 2240 }, { "epoch": 0.9911178270003317, "grad_norm": 0.32819373015120595, "learning_rate": 8.472695618908092e-06, "loss": 0.369, "step": 2241 }, { "epoch": 0.9915600928758339, "grad_norm": 0.2816097624390068, "learning_rate": 8.4708446319504e-06, "loss": 0.3673, "step": 2242 }, { "epoch": 0.992002358751336, "grad_norm": 0.3013053345028628, "learning_rate": 8.468992726492901e-06, "loss": 0.3516, "step": 2243 }, { "epoch": 0.9924446246268381, "grad_norm": 0.2943182010228169, "learning_rate": 8.467139903025671e-06, "loss": 0.3447, "step": 2244 }, { "epoch": 0.9928868905023404, "grad_norm": 0.34008563526792923, "learning_rate": 8.46528616203903e-06, "loss": 0.3809, "step": 2245 }, { "epoch": 0.9933291563778425, "grad_norm": 0.3126214550627396, "learning_rate": 8.463431504023534e-06, "loss": 0.3983, "step": 2246 }, { "epoch": 0.9937714222533446, "grad_norm": 0.35071846148194086, "learning_rate": 8.461575929469989e-06, "loss": 0.4449, "step": 2247 }, { "epoch": 0.9942136881288468, "grad_norm": 0.2877659900462922, "learning_rate": 8.45971943886944e-06, "loss": 0.3896, "step": 2248 }, { "epoch": 0.9946559540043489, "grad_norm": 0.3087373512796715, "learning_rate": 8.45786203271318e-06, "loss": 0.3874, "step": 2249 }, { "epoch": 0.9950982198798511, "grad_norm": 0.3306873757475773, "learning_rate": 8.456003711492732e-06, "loss": 0.367, "step": 2250 }, { "epoch": 0.9955404857553533, "grad_norm": 0.3275031482146109, "learning_rate": 8.45414447569988e-06, "loss": 0.4042, "step": 2251 }, { "epoch": 0.9959827516308554, "grad_norm": 0.3545508551076261, "learning_rate": 8.452284325826627e-06, "loss": 0.3995, "step": 2252 }, { "epoch": 0.9964250175063576, "grad_norm": 0.4606620852425495, "learning_rate": 8.450423262365238e-06, "loss": 0.4266, "step": 2253 }, { "epoch": 0.9968672833818597, "grad_norm": 0.2683115266782698, "learning_rate": 8.448561285808209e-06, "loss": 0.3746, "step": 2254 }, { "epoch": 0.9973095492573619, "grad_norm": 0.3381941211841519, "learning_rate": 8.44669839664828e-06, "loss": 0.3663, "step": 2255 }, { "epoch": 0.997751815132864, "grad_norm": 0.2672256894851146, "learning_rate": 8.444834595378434e-06, "loss": 0.3602, "step": 2256 }, { "epoch": 0.9981940810083662, "grad_norm": 0.29716634085713073, "learning_rate": 8.442969882491893e-06, "loss": 0.415, "step": 2257 }, { "epoch": 0.9986363468838684, "grad_norm": 0.29920063327444146, "learning_rate": 8.441104258482123e-06, "loss": 0.3502, "step": 2258 }, { "epoch": 0.9990786127593705, "grad_norm": 0.33851203936209834, "learning_rate": 8.439237723842828e-06, "loss": 0.3863, "step": 2259 }, { "epoch": 0.9995208786348727, "grad_norm": 0.29832112080365425, "learning_rate": 8.437370279067953e-06, "loss": 0.3205, "step": 2260 }, { "epoch": 0.9999631445103748, "grad_norm": 0.30283172757288956, "learning_rate": 8.43550192465169e-06, "loss": 0.3926, "step": 2261 }, { "epoch": 1.0, "grad_norm": 0.30283172757288956, "learning_rate": 8.433632661088466e-06, "loss": 0.8149, "step": 2262 }, { "epoch": 1.0004422658755021, "grad_norm": 7.153353890324201, "learning_rate": 8.431762488872945e-06, "loss": 0.3549, "step": 2263 }, { "epoch": 1.0008845317510042, "grad_norm": 0.31064713037963226, "learning_rate": 8.429891408500042e-06, "loss": 0.2977, "step": 2264 }, { "epoch": 1.0013267976265066, "grad_norm": 0.2959306137697171, "learning_rate": 8.428019420464904e-06, "loss": 0.3417, "step": 2265 }, { "epoch": 1.0017690635020087, "grad_norm": 0.3610528634528582, "learning_rate": 8.42614652526292e-06, "loss": 0.3804, "step": 2266 }, { "epoch": 1.0022113293775108, "grad_norm": 0.42162748434642067, "learning_rate": 8.42427272338972e-06, "loss": 0.3638, "step": 2267 }, { "epoch": 1.002653595253013, "grad_norm": 0.3069414256207635, "learning_rate": 8.422398015341173e-06, "loss": 0.3213, "step": 2268 }, { "epoch": 1.003095861128515, "grad_norm": 0.3176815495580194, "learning_rate": 8.420522401613393e-06, "loss": 0.3575, "step": 2269 }, { "epoch": 1.0035381270040173, "grad_norm": 0.3315803293719766, "learning_rate": 8.418645882702723e-06, "loss": 0.3187, "step": 2270 }, { "epoch": 1.0039803928795195, "grad_norm": 0.3232606386912816, "learning_rate": 8.416768459105755e-06, "loss": 0.322, "step": 2271 }, { "epoch": 1.0044226587550216, "grad_norm": 0.29128434646266166, "learning_rate": 8.414890131319318e-06, "loss": 0.3213, "step": 2272 }, { "epoch": 1.0048649246305237, "grad_norm": 0.4080195860363376, "learning_rate": 8.413010899840477e-06, "loss": 0.3584, "step": 2273 }, { "epoch": 1.0053071905060258, "grad_norm": 0.3323365275143871, "learning_rate": 8.411130765166539e-06, "loss": 0.302, "step": 2274 }, { "epoch": 1.0057494563815281, "grad_norm": 0.3216963698031913, "learning_rate": 8.409249727795052e-06, "loss": 0.3062, "step": 2275 }, { "epoch": 1.0061917222570302, "grad_norm": 0.3555157962165505, "learning_rate": 8.407367788223797e-06, "loss": 0.3264, "step": 2276 }, { "epoch": 1.0066339881325324, "grad_norm": 0.2961987739352015, "learning_rate": 8.4054849469508e-06, "loss": 0.327, "step": 2277 }, { "epoch": 1.0070762540080345, "grad_norm": 0.2645850715615522, "learning_rate": 8.40360120447432e-06, "loss": 0.3493, "step": 2278 }, { "epoch": 1.0075185198835366, "grad_norm": 0.2859596616494115, "learning_rate": 8.40171656129286e-06, "loss": 0.3038, "step": 2279 }, { "epoch": 1.007960785759039, "grad_norm": 0.31260447023009375, "learning_rate": 8.399831017905159e-06, "loss": 0.345, "step": 2280 }, { "epoch": 1.008403051634541, "grad_norm": 0.2885722697542524, "learning_rate": 8.39794457481019e-06, "loss": 0.3363, "step": 2281 }, { "epoch": 1.0088453175100431, "grad_norm": 0.3091706612046924, "learning_rate": 8.396057232507171e-06, "loss": 0.3145, "step": 2282 }, { "epoch": 1.0092875833855453, "grad_norm": 0.31473666300183717, "learning_rate": 8.394168991495556e-06, "loss": 0.3234, "step": 2283 }, { "epoch": 1.0097298492610474, "grad_norm": 0.2878853469278177, "learning_rate": 8.392279852275031e-06, "loss": 0.3105, "step": 2284 }, { "epoch": 1.0101721151365497, "grad_norm": 0.30748027359459024, "learning_rate": 8.39038981534553e-06, "loss": 0.3487, "step": 2285 }, { "epoch": 1.0106143810120518, "grad_norm": 0.3290171395460303, "learning_rate": 8.388498881207217e-06, "loss": 0.3394, "step": 2286 }, { "epoch": 1.011056646887554, "grad_norm": 0.35472102025572344, "learning_rate": 8.386607050360494e-06, "loss": 0.3185, "step": 2287 }, { "epoch": 1.011498912763056, "grad_norm": 0.2654571428011246, "learning_rate": 8.384714323306002e-06, "loss": 0.3066, "step": 2288 }, { "epoch": 1.0119411786385581, "grad_norm": 0.6777430097844926, "learning_rate": 8.38282070054462e-06, "loss": 0.3656, "step": 2289 }, { "epoch": 1.0123834445140603, "grad_norm": 0.28861006726708865, "learning_rate": 8.380926182577463e-06, "loss": 0.3499, "step": 2290 }, { "epoch": 1.0128257103895626, "grad_norm": 0.3312668746500124, "learning_rate": 8.379030769905884e-06, "loss": 0.3338, "step": 2291 }, { "epoch": 1.0132679762650647, "grad_norm": 0.32863677562945975, "learning_rate": 8.377134463031468e-06, "loss": 0.3501, "step": 2292 }, { "epoch": 1.0137102421405668, "grad_norm": 0.32398705551899804, "learning_rate": 8.375237262456044e-06, "loss": 0.3521, "step": 2293 }, { "epoch": 1.014152508016069, "grad_norm": 0.27407911949530495, "learning_rate": 8.373339168681672e-06, "loss": 0.3028, "step": 2294 }, { "epoch": 1.014594773891571, "grad_norm": 0.319821287305308, "learning_rate": 8.371440182210648e-06, "loss": 0.3119, "step": 2295 }, { "epoch": 1.0150370397670734, "grad_norm": 0.3373902524877863, "learning_rate": 8.369540303545512e-06, "loss": 0.3307, "step": 2296 }, { "epoch": 1.0154793056425755, "grad_norm": 0.29872242992653186, "learning_rate": 8.367639533189028e-06, "loss": 0.3225, "step": 2297 }, { "epoch": 1.0159215715180776, "grad_norm": 1.2211468227724123, "learning_rate": 8.365737871644208e-06, "loss": 0.3786, "step": 2298 }, { "epoch": 1.0163638373935797, "grad_norm": 0.3061872985326688, "learning_rate": 8.36383531941429e-06, "loss": 0.3145, "step": 2299 }, { "epoch": 1.0168061032690818, "grad_norm": 0.4245917806681053, "learning_rate": 8.361931877002755e-06, "loss": 0.3451, "step": 2300 }, { "epoch": 1.0172483691445842, "grad_norm": 0.27955681389423404, "learning_rate": 8.360027544913315e-06, "loss": 0.3344, "step": 2301 }, { "epoch": 1.0176906350200863, "grad_norm": 0.32069265718579404, "learning_rate": 8.35812232364992e-06, "loss": 0.3136, "step": 2302 }, { "epoch": 1.0181329008955884, "grad_norm": 0.4780766542070092, "learning_rate": 8.356216213716755e-06, "loss": 0.3934, "step": 2303 }, { "epoch": 1.0185751667710905, "grad_norm": 0.3490201198150145, "learning_rate": 8.354309215618236e-06, "loss": 0.3223, "step": 2304 }, { "epoch": 1.0190174326465926, "grad_norm": 0.37122544700549476, "learning_rate": 8.352401329859018e-06, "loss": 0.3147, "step": 2305 }, { "epoch": 1.019459698522095, "grad_norm": 0.38234288440623687, "learning_rate": 8.350492556943996e-06, "loss": 0.304, "step": 2306 }, { "epoch": 1.019901964397597, "grad_norm": 0.3075186042526609, "learning_rate": 8.348582897378288e-06, "loss": 0.3203, "step": 2307 }, { "epoch": 1.0203442302730992, "grad_norm": 0.3270101061551584, "learning_rate": 8.346672351667256e-06, "loss": 0.3094, "step": 2308 }, { "epoch": 1.0207864961486013, "grad_norm": 0.39840066967242066, "learning_rate": 8.344760920316492e-06, "loss": 0.35, "step": 2309 }, { "epoch": 1.0212287620241034, "grad_norm": 0.4064177328493202, "learning_rate": 8.342848603831823e-06, "loss": 0.3546, "step": 2310 }, { "epoch": 1.0216710278996057, "grad_norm": 0.3531322078756765, "learning_rate": 8.340935402719315e-06, "loss": 0.3602, "step": 2311 }, { "epoch": 1.0221132937751078, "grad_norm": 0.31546571194003664, "learning_rate": 8.339021317485259e-06, "loss": 0.3129, "step": 2312 }, { "epoch": 1.02255555965061, "grad_norm": 0.3112748085534974, "learning_rate": 8.337106348636186e-06, "loss": 0.3511, "step": 2313 }, { "epoch": 1.022997825526112, "grad_norm": 0.3065744263035341, "learning_rate": 8.335190496678862e-06, "loss": 0.3393, "step": 2314 }, { "epoch": 1.0234400914016142, "grad_norm": 0.3062863949408427, "learning_rate": 8.333273762120284e-06, "loss": 0.3215, "step": 2315 }, { "epoch": 1.0238823572771165, "grad_norm": 0.2839273471194481, "learning_rate": 8.331356145467682e-06, "loss": 0.3112, "step": 2316 }, { "epoch": 1.0243246231526186, "grad_norm": 0.300417422771467, "learning_rate": 8.32943764722852e-06, "loss": 0.3342, "step": 2317 }, { "epoch": 1.0247668890281207, "grad_norm": 0.4691005591068085, "learning_rate": 8.327518267910498e-06, "loss": 0.3312, "step": 2318 }, { "epoch": 1.0252091549036229, "grad_norm": 0.25850530157165286, "learning_rate": 8.325598008021544e-06, "loss": 0.3389, "step": 2319 }, { "epoch": 1.025651420779125, "grad_norm": 0.3236063050832242, "learning_rate": 8.323676868069826e-06, "loss": 0.3351, "step": 2320 }, { "epoch": 1.0260936866546273, "grad_norm": 0.3931208021848622, "learning_rate": 8.321754848563739e-06, "loss": 0.3442, "step": 2321 }, { "epoch": 1.0265359525301294, "grad_norm": 0.323633217857614, "learning_rate": 8.319831950011912e-06, "loss": 0.3297, "step": 2322 }, { "epoch": 1.0269782184056315, "grad_norm": 0.4145838354560546, "learning_rate": 8.317908172923207e-06, "loss": 0.3037, "step": 2323 }, { "epoch": 1.0274204842811336, "grad_norm": 0.2882954431866344, "learning_rate": 8.315983517806717e-06, "loss": 0.3284, "step": 2324 }, { "epoch": 1.0278627501566358, "grad_norm": 0.27919289765496397, "learning_rate": 8.314057985171776e-06, "loss": 0.3371, "step": 2325 }, { "epoch": 1.028305016032138, "grad_norm": 0.27866063902177424, "learning_rate": 8.312131575527938e-06, "loss": 0.3466, "step": 2326 }, { "epoch": 1.0287472819076402, "grad_norm": 0.2545342886760366, "learning_rate": 8.310204289384995e-06, "loss": 0.3414, "step": 2327 }, { "epoch": 1.0291895477831423, "grad_norm": 0.34836110965655986, "learning_rate": 8.30827612725297e-06, "loss": 0.3541, "step": 2328 }, { "epoch": 1.0296318136586444, "grad_norm": 0.32561123685608806, "learning_rate": 8.30634708964212e-06, "loss": 0.2932, "step": 2329 }, { "epoch": 1.0300740795341465, "grad_norm": 0.4377770101104962, "learning_rate": 8.304417177062928e-06, "loss": 0.3322, "step": 2330 }, { "epoch": 1.0305163454096489, "grad_norm": 0.32386434384599133, "learning_rate": 8.302486390026118e-06, "loss": 0.3432, "step": 2331 }, { "epoch": 1.030958611285151, "grad_norm": 0.2860987268389987, "learning_rate": 8.300554729042637e-06, "loss": 0.349, "step": 2332 }, { "epoch": 1.031400877160653, "grad_norm": 0.38562126551438236, "learning_rate": 8.298622194623667e-06, "loss": 0.3067, "step": 2333 }, { "epoch": 1.0318431430361552, "grad_norm": 0.3197411594982416, "learning_rate": 8.296688787280618e-06, "loss": 0.3377, "step": 2334 }, { "epoch": 1.0322854089116573, "grad_norm": 0.3379960281936614, "learning_rate": 8.294754507525135e-06, "loss": 0.3397, "step": 2335 }, { "epoch": 1.0327276747871594, "grad_norm": 0.3095558670637717, "learning_rate": 8.292819355869094e-06, "loss": 0.3379, "step": 2336 }, { "epoch": 1.0331699406626618, "grad_norm": 0.30125271054027897, "learning_rate": 8.290883332824596e-06, "loss": 0.334, "step": 2337 }, { "epoch": 1.0336122065381639, "grad_norm": 0.3242109905024203, "learning_rate": 8.288946438903978e-06, "loss": 0.3097, "step": 2338 }, { "epoch": 1.034054472413666, "grad_norm": 0.29424051828532277, "learning_rate": 8.287008674619805e-06, "loss": 0.3247, "step": 2339 }, { "epoch": 1.034496738289168, "grad_norm": 0.32948786946480024, "learning_rate": 8.285070040484877e-06, "loss": 0.3567, "step": 2340 }, { "epoch": 1.0349390041646704, "grad_norm": 0.3127817350538908, "learning_rate": 8.283130537012218e-06, "loss": 0.3188, "step": 2341 }, { "epoch": 1.0353812700401726, "grad_norm": 0.39587328879001904, "learning_rate": 8.281190164715083e-06, "loss": 0.3522, "step": 2342 }, { "epoch": 1.0358235359156747, "grad_norm": 0.2786164135877559, "learning_rate": 8.27924892410696e-06, "loss": 0.3227, "step": 2343 }, { "epoch": 1.0362658017911768, "grad_norm": 0.292063861751596, "learning_rate": 8.277306815701565e-06, "loss": 0.283, "step": 2344 }, { "epoch": 1.036708067666679, "grad_norm": 0.3730319954330297, "learning_rate": 8.275363840012844e-06, "loss": 0.35, "step": 2345 }, { "epoch": 1.037150333542181, "grad_norm": 0.3426616432428177, "learning_rate": 8.273419997554972e-06, "loss": 0.3397, "step": 2346 }, { "epoch": 1.0375925994176833, "grad_norm": 0.31153007684761963, "learning_rate": 8.271475288842354e-06, "loss": 0.3287, "step": 2347 }, { "epoch": 1.0380348652931854, "grad_norm": 0.3384489832187917, "learning_rate": 8.269529714389624e-06, "loss": 0.3263, "step": 2348 }, { "epoch": 1.0384771311686876, "grad_norm": 0.26935681695700875, "learning_rate": 8.267583274711645e-06, "loss": 0.3444, "step": 2349 }, { "epoch": 1.0389193970441897, "grad_norm": 0.35373466397401954, "learning_rate": 8.26563597032351e-06, "loss": 0.3354, "step": 2350 }, { "epoch": 1.0393616629196918, "grad_norm": 0.3792091572238985, "learning_rate": 8.263687801740538e-06, "loss": 0.3378, "step": 2351 }, { "epoch": 1.0398039287951941, "grad_norm": 0.2938967361125827, "learning_rate": 8.26173876947828e-06, "loss": 0.3253, "step": 2352 }, { "epoch": 1.0402461946706962, "grad_norm": 0.32257507167245975, "learning_rate": 8.259788874052513e-06, "loss": 0.3167, "step": 2353 }, { "epoch": 1.0406884605461983, "grad_norm": 0.3998107277331622, "learning_rate": 8.257838115979244e-06, "loss": 0.38, "step": 2354 }, { "epoch": 1.0411307264217005, "grad_norm": 0.3180470214713486, "learning_rate": 8.25588649577471e-06, "loss": 0.3305, "step": 2355 }, { "epoch": 1.0415729922972026, "grad_norm": 0.32517568422859944, "learning_rate": 8.25393401395537e-06, "loss": 0.3432, "step": 2356 }, { "epoch": 1.042015258172705, "grad_norm": 0.29241861739419484, "learning_rate": 8.251980671037917e-06, "loss": 0.3341, "step": 2357 }, { "epoch": 1.042457524048207, "grad_norm": 0.31136993439103683, "learning_rate": 8.250026467539271e-06, "loss": 0.316, "step": 2358 }, { "epoch": 1.0428997899237091, "grad_norm": 0.273128119526128, "learning_rate": 8.248071403976576e-06, "loss": 0.3191, "step": 2359 }, { "epoch": 1.0433420557992112, "grad_norm": 0.3257812418850747, "learning_rate": 8.24611548086721e-06, "loss": 0.3444, "step": 2360 }, { "epoch": 1.0437843216747134, "grad_norm": 0.3254024176127342, "learning_rate": 8.24415869872877e-06, "loss": 0.3334, "step": 2361 }, { "epoch": 1.0442265875502157, "grad_norm": 0.3382067776530988, "learning_rate": 8.242201058079089e-06, "loss": 0.333, "step": 2362 }, { "epoch": 1.0446688534257178, "grad_norm": 0.3078219632005735, "learning_rate": 8.24024255943622e-06, "loss": 0.303, "step": 2363 }, { "epoch": 1.04511111930122, "grad_norm": 0.3045532889951893, "learning_rate": 8.23828320331845e-06, "loss": 0.3133, "step": 2364 }, { "epoch": 1.045553385176722, "grad_norm": 0.3448450371279707, "learning_rate": 8.236322990244287e-06, "loss": 0.3119, "step": 2365 }, { "epoch": 1.0459956510522241, "grad_norm": 0.31328034179337727, "learning_rate": 8.234361920732466e-06, "loss": 0.3648, "step": 2366 }, { "epoch": 1.0464379169277265, "grad_norm": 0.32085718792576073, "learning_rate": 8.232399995301955e-06, "loss": 0.3229, "step": 2367 }, { "epoch": 1.0468801828032286, "grad_norm": 0.24419387802159462, "learning_rate": 8.230437214471938e-06, "loss": 0.3109, "step": 2368 }, { "epoch": 1.0473224486787307, "grad_norm": 0.4034167545097128, "learning_rate": 8.228473578761838e-06, "loss": 0.3297, "step": 2369 }, { "epoch": 1.0477647145542328, "grad_norm": 0.31793703373348614, "learning_rate": 8.226509088691293e-06, "loss": 0.3458, "step": 2370 }, { "epoch": 1.048206980429735, "grad_norm": 0.38399470262307955, "learning_rate": 8.224543744780173e-06, "loss": 0.2986, "step": 2371 }, { "epoch": 1.0486492463052373, "grad_norm": 0.3715330483908964, "learning_rate": 8.222577547548576e-06, "loss": 0.3464, "step": 2372 }, { "epoch": 1.0490915121807394, "grad_norm": 0.32232252943120077, "learning_rate": 8.220610497516819e-06, "loss": 0.3408, "step": 2373 }, { "epoch": 1.0495337780562415, "grad_norm": 0.3019818785872827, "learning_rate": 8.218642595205447e-06, "loss": 0.3152, "step": 2374 }, { "epoch": 1.0499760439317436, "grad_norm": 0.27252733801530266, "learning_rate": 8.216673841135235e-06, "loss": 0.3353, "step": 2375 }, { "epoch": 1.0504183098072457, "grad_norm": 0.3719304913595202, "learning_rate": 8.214704235827178e-06, "loss": 0.3245, "step": 2376 }, { "epoch": 1.050860575682748, "grad_norm": 0.31372118023027334, "learning_rate": 8.212733779802503e-06, "loss": 0.3328, "step": 2377 }, { "epoch": 1.0513028415582502, "grad_norm": 0.6033565537719314, "learning_rate": 8.21076247358265e-06, "loss": 0.3345, "step": 2378 }, { "epoch": 1.0517451074337523, "grad_norm": 0.2961511688034379, "learning_rate": 8.208790317689296e-06, "loss": 0.3089, "step": 2379 }, { "epoch": 1.0521873733092544, "grad_norm": 0.31546088789452675, "learning_rate": 8.206817312644339e-06, "loss": 0.3219, "step": 2380 }, { "epoch": 1.0526296391847565, "grad_norm": 0.280958824862648, "learning_rate": 8.2048434589699e-06, "loss": 0.309, "step": 2381 }, { "epoch": 1.0530719050602588, "grad_norm": 0.2934619311302312, "learning_rate": 8.202868757188327e-06, "loss": 0.3428, "step": 2382 }, { "epoch": 1.053514170935761, "grad_norm": 0.3719772388305975, "learning_rate": 8.200893207822189e-06, "loss": 0.3245, "step": 2383 }, { "epoch": 1.053956436811263, "grad_norm": 0.25905491676331394, "learning_rate": 8.19891681139428e-06, "loss": 0.321, "step": 2384 }, { "epoch": 1.0543987026867652, "grad_norm": 0.2992915615617365, "learning_rate": 8.196939568427624e-06, "loss": 0.3298, "step": 2385 }, { "epoch": 1.0548409685622673, "grad_norm": 0.2952235509816342, "learning_rate": 8.194961479445463e-06, "loss": 0.3191, "step": 2386 }, { "epoch": 1.0552832344377696, "grad_norm": 0.3573812483124472, "learning_rate": 8.192982544971263e-06, "loss": 0.3487, "step": 2387 }, { "epoch": 1.0557255003132717, "grad_norm": 0.33653332571363936, "learning_rate": 8.191002765528717e-06, "loss": 0.3308, "step": 2388 }, { "epoch": 1.0561677661887738, "grad_norm": 0.43119856234999726, "learning_rate": 8.189022141641738e-06, "loss": 0.316, "step": 2389 }, { "epoch": 1.056610032064276, "grad_norm": 0.29649211672883896, "learning_rate": 8.187040673834465e-06, "loss": 0.3292, "step": 2390 }, { "epoch": 1.057052297939778, "grad_norm": 0.38610768284916047, "learning_rate": 8.185058362631262e-06, "loss": 0.3667, "step": 2391 }, { "epoch": 1.0574945638152804, "grad_norm": 0.30897497833944104, "learning_rate": 8.18307520855671e-06, "loss": 0.3345, "step": 2392 }, { "epoch": 1.0579368296907825, "grad_norm": 0.2961891995499439, "learning_rate": 8.18109121213562e-06, "loss": 0.3483, "step": 2393 }, { "epoch": 1.0583790955662846, "grad_norm": 0.33643232343446516, "learning_rate": 8.179106373893022e-06, "loss": 0.3108, "step": 2394 }, { "epoch": 1.0588213614417867, "grad_norm": 0.3532784427208817, "learning_rate": 8.177120694354167e-06, "loss": 0.34, "step": 2395 }, { "epoch": 1.0592636273172888, "grad_norm": 0.30042889343072965, "learning_rate": 8.175134174044534e-06, "loss": 0.3264, "step": 2396 }, { "epoch": 1.059705893192791, "grad_norm": 0.30039362651073814, "learning_rate": 8.173146813489823e-06, "loss": 0.2923, "step": 2397 }, { "epoch": 1.0601481590682933, "grad_norm": 0.28769072616328123, "learning_rate": 8.17115861321595e-06, "loss": 0.3222, "step": 2398 }, { "epoch": 1.0605904249437954, "grad_norm": 0.356832332134811, "learning_rate": 8.169169573749064e-06, "loss": 0.3523, "step": 2399 }, { "epoch": 1.0610326908192975, "grad_norm": 0.30393617959797314, "learning_rate": 8.167179695615526e-06, "loss": 0.3, "step": 2400 }, { "epoch": 1.0614749566947996, "grad_norm": 0.3000488898598037, "learning_rate": 8.165188979341928e-06, "loss": 0.3066, "step": 2401 }, { "epoch": 1.0619172225703017, "grad_norm": 0.35047562983531655, "learning_rate": 8.163197425455075e-06, "loss": 0.3181, "step": 2402 }, { "epoch": 1.062359488445804, "grad_norm": 0.28328331681513436, "learning_rate": 8.161205034481996e-06, "loss": 0.3515, "step": 2403 }, { "epoch": 1.0628017543213062, "grad_norm": 0.2876359948227042, "learning_rate": 8.159211806949947e-06, "loss": 0.3058, "step": 2404 }, { "epoch": 1.0632440201968083, "grad_norm": 0.340694493483621, "learning_rate": 8.157217743386403e-06, "loss": 0.3261, "step": 2405 }, { "epoch": 1.0636862860723104, "grad_norm": 0.4037949251615118, "learning_rate": 8.155222844319055e-06, "loss": 0.3699, "step": 2406 }, { "epoch": 1.0641285519478125, "grad_norm": 0.34232368104945404, "learning_rate": 8.153227110275821e-06, "loss": 0.3278, "step": 2407 }, { "epoch": 1.0645708178233149, "grad_norm": 0.5219275824283625, "learning_rate": 8.151230541784838e-06, "loss": 0.354, "step": 2408 }, { "epoch": 1.065013083698817, "grad_norm": 0.3542714141885101, "learning_rate": 8.149233139374464e-06, "loss": 0.3813, "step": 2409 }, { "epoch": 1.065455349574319, "grad_norm": 0.3359236598207323, "learning_rate": 8.147234903573276e-06, "loss": 0.355, "step": 2410 }, { "epoch": 1.0658976154498212, "grad_norm": 0.383312018472175, "learning_rate": 8.145235834910073e-06, "loss": 0.3484, "step": 2411 }, { "epoch": 1.0663398813253233, "grad_norm": 0.2993005514247704, "learning_rate": 8.143235933913877e-06, "loss": 0.3207, "step": 2412 }, { "epoch": 1.0667821472008256, "grad_norm": 0.254848502458691, "learning_rate": 8.141235201113924e-06, "loss": 0.2969, "step": 2413 }, { "epoch": 1.0672244130763278, "grad_norm": 0.3302029029179795, "learning_rate": 8.139233637039678e-06, "loss": 0.3399, "step": 2414 }, { "epoch": 1.0676666789518299, "grad_norm": 0.3045875370042859, "learning_rate": 8.137231242220817e-06, "loss": 0.3542, "step": 2415 }, { "epoch": 1.068108944827332, "grad_norm": 0.30008740272434287, "learning_rate": 8.135228017187238e-06, "loss": 0.3205, "step": 2416 }, { "epoch": 1.068551210702834, "grad_norm": 0.25135547760391225, "learning_rate": 8.133223962469064e-06, "loss": 0.3422, "step": 2417 }, { "epoch": 1.0689934765783364, "grad_norm": 0.2787911701408232, "learning_rate": 8.131219078596632e-06, "loss": 0.3107, "step": 2418 }, { "epoch": 1.0694357424538385, "grad_norm": 0.3682139435638506, "learning_rate": 8.129213366100498e-06, "loss": 0.3652, "step": 2419 }, { "epoch": 1.0698780083293407, "grad_norm": 0.3345859946095334, "learning_rate": 8.127206825511447e-06, "loss": 0.3264, "step": 2420 }, { "epoch": 1.0703202742048428, "grad_norm": 0.31628703812906633, "learning_rate": 8.125199457360468e-06, "loss": 0.342, "step": 2421 }, { "epoch": 1.0707625400803449, "grad_norm": 0.2939280018075023, "learning_rate": 8.12319126217878e-06, "loss": 0.2988, "step": 2422 }, { "epoch": 1.0712048059558472, "grad_norm": 0.42395797209437036, "learning_rate": 8.12118224049782e-06, "loss": 0.3702, "step": 2423 }, { "epoch": 1.0716470718313493, "grad_norm": 0.34690455423690336, "learning_rate": 8.119172392849234e-06, "loss": 0.3449, "step": 2424 }, { "epoch": 1.0720893377068514, "grad_norm": 0.3295056785383796, "learning_rate": 8.117161719764903e-06, "loss": 0.3458, "step": 2425 }, { "epoch": 1.0725316035823536, "grad_norm": 0.29578446649148454, "learning_rate": 8.11515022177691e-06, "loss": 0.2879, "step": 2426 }, { "epoch": 1.0729738694578557, "grad_norm": 0.30821875319495107, "learning_rate": 8.113137899417568e-06, "loss": 0.3466, "step": 2427 }, { "epoch": 1.073416135333358, "grad_norm": 0.29899344892505775, "learning_rate": 8.111124753219401e-06, "loss": 0.3426, "step": 2428 }, { "epoch": 1.0738584012088601, "grad_norm": 0.3531632621889402, "learning_rate": 8.109110783715157e-06, "loss": 0.3484, "step": 2429 }, { "epoch": 1.0743006670843622, "grad_norm": 0.2730505084029503, "learning_rate": 8.107095991437794e-06, "loss": 0.3212, "step": 2430 }, { "epoch": 1.0747429329598643, "grad_norm": 0.36964335762623163, "learning_rate": 8.105080376920498e-06, "loss": 0.3262, "step": 2431 }, { "epoch": 1.0751851988353665, "grad_norm": 0.3074477654203298, "learning_rate": 8.103063940696662e-06, "loss": 0.3391, "step": 2432 }, { "epoch": 1.0756274647108688, "grad_norm": 0.3313163861385352, "learning_rate": 8.101046683299903e-06, "loss": 0.3543, "step": 2433 }, { "epoch": 1.076069730586371, "grad_norm": 0.4372432285253392, "learning_rate": 8.099028605264057e-06, "loss": 0.3549, "step": 2434 }, { "epoch": 1.076511996461873, "grad_norm": 0.2806892918882414, "learning_rate": 8.097009707123169e-06, "loss": 0.3214, "step": 2435 }, { "epoch": 1.0769542623373751, "grad_norm": 0.31642493985769965, "learning_rate": 8.094989989411508e-06, "loss": 0.3515, "step": 2436 }, { "epoch": 1.0773965282128772, "grad_norm": 0.27164871571245247, "learning_rate": 8.092969452663559e-06, "loss": 0.3347, "step": 2437 }, { "epoch": 1.0778387940883793, "grad_norm": 0.2583977288164514, "learning_rate": 8.09094809741402e-06, "loss": 0.3098, "step": 2438 }, { "epoch": 1.0782810599638817, "grad_norm": 0.30804126106571467, "learning_rate": 8.088925924197813e-06, "loss": 0.2967, "step": 2439 }, { "epoch": 1.0787233258393838, "grad_norm": 0.3280505122289885, "learning_rate": 8.086902933550064e-06, "loss": 0.35, "step": 2440 }, { "epoch": 1.079165591714886, "grad_norm": 0.30770468405132423, "learning_rate": 8.084879126006131e-06, "loss": 0.3232, "step": 2441 }, { "epoch": 1.079607857590388, "grad_norm": 0.2787411570052703, "learning_rate": 8.082854502101574e-06, "loss": 0.3136, "step": 2442 }, { "epoch": 1.0800501234658904, "grad_norm": 0.3670516767007484, "learning_rate": 8.080829062372179e-06, "loss": 0.3419, "step": 2443 }, { "epoch": 1.0804923893413925, "grad_norm": 0.3649179468685062, "learning_rate": 8.078802807353941e-06, "loss": 0.3369, "step": 2444 }, { "epoch": 1.0809346552168946, "grad_norm": 0.32693822121198796, "learning_rate": 8.076775737583077e-06, "loss": 0.3152, "step": 2445 }, { "epoch": 1.0813769210923967, "grad_norm": 0.3574114564998175, "learning_rate": 8.074747853596014e-06, "loss": 0.3446, "step": 2446 }, { "epoch": 1.0818191869678988, "grad_norm": 0.3461827487857562, "learning_rate": 8.0727191559294e-06, "loss": 0.3486, "step": 2447 }, { "epoch": 1.082261452843401, "grad_norm": 0.3219291533779336, "learning_rate": 8.070689645120092e-06, "loss": 0.3193, "step": 2448 }, { "epoch": 1.0827037187189033, "grad_norm": 0.26506627963908314, "learning_rate": 8.068659321705164e-06, "loss": 0.2839, "step": 2449 }, { "epoch": 1.0831459845944054, "grad_norm": 0.3649164093199239, "learning_rate": 8.06662818622191e-06, "loss": 0.3296, "step": 2450 }, { "epoch": 1.0835882504699075, "grad_norm": 0.2993749218674451, "learning_rate": 8.064596239207835e-06, "loss": 0.3075, "step": 2451 }, { "epoch": 1.0840305163454096, "grad_norm": 0.2999666198787671, "learning_rate": 8.06256348120066e-06, "loss": 0.3362, "step": 2452 }, { "epoch": 1.0844727822209117, "grad_norm": 0.36600503654548144, "learning_rate": 8.060529912738316e-06, "loss": 0.3435, "step": 2453 }, { "epoch": 1.084915048096414, "grad_norm": 0.30947620517190455, "learning_rate": 8.058495534358955e-06, "loss": 0.3041, "step": 2454 }, { "epoch": 1.0853573139719161, "grad_norm": 0.34563154865853246, "learning_rate": 8.05646034660094e-06, "loss": 0.3358, "step": 2455 }, { "epoch": 1.0857995798474183, "grad_norm": 0.2633614474384424, "learning_rate": 8.054424350002846e-06, "loss": 0.3205, "step": 2456 }, { "epoch": 1.0862418457229204, "grad_norm": 0.369237515220262, "learning_rate": 8.052387545103468e-06, "loss": 0.3288, "step": 2457 }, { "epoch": 1.0866841115984225, "grad_norm": 0.37211457895946065, "learning_rate": 8.050349932441812e-06, "loss": 0.3455, "step": 2458 }, { "epoch": 1.0871263774739248, "grad_norm": 0.3196063007892165, "learning_rate": 8.048311512557096e-06, "loss": 0.3471, "step": 2459 }, { "epoch": 1.087568643349427, "grad_norm": 0.306069373360966, "learning_rate": 8.046272285988755e-06, "loss": 0.318, "step": 2460 }, { "epoch": 1.088010909224929, "grad_norm": 0.33196197866061256, "learning_rate": 8.04423225327643e-06, "loss": 0.3665, "step": 2461 }, { "epoch": 1.0884531751004312, "grad_norm": 0.3352454316448335, "learning_rate": 8.042191414959988e-06, "loss": 0.3163, "step": 2462 }, { "epoch": 1.0888954409759333, "grad_norm": 0.30747151239268744, "learning_rate": 8.040149771579498e-06, "loss": 0.3312, "step": 2463 }, { "epoch": 1.0893377068514356, "grad_norm": 0.3334488352705068, "learning_rate": 8.038107323675246e-06, "loss": 0.3432, "step": 2464 }, { "epoch": 1.0897799727269377, "grad_norm": 0.32513667341370656, "learning_rate": 8.036064071787732e-06, "loss": 0.3001, "step": 2465 }, { "epoch": 1.0902222386024398, "grad_norm": 0.35720740997351863, "learning_rate": 8.034020016457668e-06, "loss": 0.3258, "step": 2466 }, { "epoch": 1.090664504477942, "grad_norm": 0.2880714889324895, "learning_rate": 8.031975158225978e-06, "loss": 0.2913, "step": 2467 }, { "epoch": 1.091106770353444, "grad_norm": 0.32975299416556486, "learning_rate": 8.0299294976338e-06, "loss": 0.3536, "step": 2468 }, { "epoch": 1.0915490362289464, "grad_norm": 0.3205818616203226, "learning_rate": 8.02788303522248e-06, "loss": 0.3569, "step": 2469 }, { "epoch": 1.0919913021044485, "grad_norm": 0.27330069691588027, "learning_rate": 8.025835771533581e-06, "loss": 0.3098, "step": 2470 }, { "epoch": 1.0924335679799506, "grad_norm": 0.33901928059780073, "learning_rate": 8.023787707108878e-06, "loss": 0.3553, "step": 2471 }, { "epoch": 1.0928758338554527, "grad_norm": 0.41062892112937244, "learning_rate": 8.021738842490356e-06, "loss": 0.3405, "step": 2472 }, { "epoch": 1.0933180997309548, "grad_norm": 0.8340712937380467, "learning_rate": 8.01968917822021e-06, "loss": 0.3554, "step": 2473 }, { "epoch": 1.0937603656064572, "grad_norm": 0.289042971071702, "learning_rate": 8.017638714840847e-06, "loss": 0.3019, "step": 2474 }, { "epoch": 1.0942026314819593, "grad_norm": 0.39183166205126796, "learning_rate": 8.015587452894893e-06, "loss": 0.3793, "step": 2475 }, { "epoch": 1.0946448973574614, "grad_norm": 0.36090258764700023, "learning_rate": 8.013535392925177e-06, "loss": 0.3498, "step": 2476 }, { "epoch": 1.0950871632329635, "grad_norm": 0.41244187177388575, "learning_rate": 8.01148253547474e-06, "loss": 0.3366, "step": 2477 }, { "epoch": 1.0955294291084656, "grad_norm": 0.30099218834003366, "learning_rate": 8.009428881086836e-06, "loss": 0.2961, "step": 2478 }, { "epoch": 1.095971694983968, "grad_norm": 0.34491456341052473, "learning_rate": 8.00737443030493e-06, "loss": 0.366, "step": 2479 }, { "epoch": 1.09641396085947, "grad_norm": 0.31102806241417924, "learning_rate": 8.0053191836727e-06, "loss": 0.3346, "step": 2480 }, { "epoch": 1.0968562267349722, "grad_norm": 0.3523196280863003, "learning_rate": 8.003263141734025e-06, "loss": 0.3309, "step": 2481 }, { "epoch": 1.0972984926104743, "grad_norm": 0.26790590108727014, "learning_rate": 8.001206305033009e-06, "loss": 0.3268, "step": 2482 }, { "epoch": 1.0977407584859764, "grad_norm": 0.29775690007359246, "learning_rate": 7.999148674113957e-06, "loss": 0.3157, "step": 2483 }, { "epoch": 1.0981830243614787, "grad_norm": 0.3398172601849922, "learning_rate": 7.997090249521384e-06, "loss": 0.296, "step": 2484 }, { "epoch": 1.0986252902369809, "grad_norm": 0.2842907738281299, "learning_rate": 7.995031031800016e-06, "loss": 0.3536, "step": 2485 }, { "epoch": 1.099067556112483, "grad_norm": 0.35505291469647865, "learning_rate": 7.992971021494793e-06, "loss": 0.3336, "step": 2486 }, { "epoch": 1.099509821987985, "grad_norm": 0.3751187174177176, "learning_rate": 7.990910219150858e-06, "loss": 0.3514, "step": 2487 }, { "epoch": 1.0999520878634872, "grad_norm": 0.27637487333060506, "learning_rate": 7.988848625313571e-06, "loss": 0.3314, "step": 2488 }, { "epoch": 1.1003943537389893, "grad_norm": 0.29491096927381816, "learning_rate": 7.986786240528495e-06, "loss": 0.3261, "step": 2489 }, { "epoch": 1.1008366196144916, "grad_norm": 0.2741035300779732, "learning_rate": 7.984723065341407e-06, "loss": 0.3241, "step": 2490 }, { "epoch": 1.1012788854899938, "grad_norm": 0.2970979781510689, "learning_rate": 7.98265910029829e-06, "loss": 0.3219, "step": 2491 }, { "epoch": 1.1017211513654959, "grad_norm": 0.28509051347186226, "learning_rate": 7.980594345945333e-06, "loss": 0.2938, "step": 2492 }, { "epoch": 1.102163417240998, "grad_norm": 0.27245946765808104, "learning_rate": 7.978528802828947e-06, "loss": 0.3129, "step": 2493 }, { "epoch": 1.1026056831165003, "grad_norm": 0.30586911475543194, "learning_rate": 7.976462471495735e-06, "loss": 0.3313, "step": 2494 }, { "epoch": 1.1030479489920024, "grad_norm": 0.2602406836626206, "learning_rate": 7.97439535249252e-06, "loss": 0.3119, "step": 2495 }, { "epoch": 1.1034902148675045, "grad_norm": 0.3840812072981307, "learning_rate": 7.972327446366328e-06, "loss": 0.3479, "step": 2496 }, { "epoch": 1.1039324807430066, "grad_norm": 0.3423685903657019, "learning_rate": 7.970258753664398e-06, "loss": 0.3596, "step": 2497 }, { "epoch": 1.1043747466185088, "grad_norm": 0.312544883109534, "learning_rate": 7.968189274934168e-06, "loss": 0.3297, "step": 2498 }, { "epoch": 1.1048170124940109, "grad_norm": 0.33951413764020233, "learning_rate": 7.966119010723298e-06, "loss": 0.3881, "step": 2499 }, { "epoch": 1.1052592783695132, "grad_norm": 0.27562135342659266, "learning_rate": 7.964047961579644e-06, "loss": 0.3413, "step": 2500 }, { "epoch": 1.1057015442450153, "grad_norm": 0.3342176555723036, "learning_rate": 7.961976128051273e-06, "loss": 0.3102, "step": 2501 }, { "epoch": 1.1061438101205174, "grad_norm": 0.3642132411769321, "learning_rate": 7.959903510686463e-06, "loss": 0.3474, "step": 2502 }, { "epoch": 1.1065860759960195, "grad_norm": 0.29389342981459715, "learning_rate": 7.957830110033695e-06, "loss": 0.3475, "step": 2503 }, { "epoch": 1.1070283418715217, "grad_norm": 0.32275937764294127, "learning_rate": 7.95575592664166e-06, "loss": 0.3065, "step": 2504 }, { "epoch": 1.107470607747024, "grad_norm": 0.3068716134365917, "learning_rate": 7.953680961059254e-06, "loss": 0.3455, "step": 2505 }, { "epoch": 1.107912873622526, "grad_norm": 0.3208867885104607, "learning_rate": 7.951605213835582e-06, "loss": 0.3745, "step": 2506 }, { "epoch": 1.1083551394980282, "grad_norm": 0.3123695758705067, "learning_rate": 7.949528685519957e-06, "loss": 0.3417, "step": 2507 }, { "epoch": 1.1087974053735303, "grad_norm": 0.24947273028045935, "learning_rate": 7.947451376661891e-06, "loss": 0.308, "step": 2508 }, { "epoch": 1.1092396712490324, "grad_norm": 0.30620400244899865, "learning_rate": 7.945373287811116e-06, "loss": 0.3371, "step": 2509 }, { "epoch": 1.1096819371245348, "grad_norm": 0.3363765482974151, "learning_rate": 7.943294419517557e-06, "loss": 0.3541, "step": 2510 }, { "epoch": 1.110124203000037, "grad_norm": 0.29190471714035265, "learning_rate": 7.941214772331354e-06, "loss": 0.3352, "step": 2511 }, { "epoch": 1.110566468875539, "grad_norm": 0.40224462522638654, "learning_rate": 7.939134346802849e-06, "loss": 0.3362, "step": 2512 }, { "epoch": 1.1110087347510411, "grad_norm": 0.3927063420143256, "learning_rate": 7.93705314348259e-06, "loss": 0.3143, "step": 2513 }, { "epoch": 1.1114510006265432, "grad_norm": 0.36874692489399086, "learning_rate": 7.934971162921336e-06, "loss": 0.3332, "step": 2514 }, { "epoch": 1.1118932665020456, "grad_norm": 0.33043108678040306, "learning_rate": 7.932888405670044e-06, "loss": 0.3473, "step": 2515 }, { "epoch": 1.1123355323775477, "grad_norm": 0.3749220313652345, "learning_rate": 7.93080487227988e-06, "loss": 0.3306, "step": 2516 }, { "epoch": 1.1127777982530498, "grad_norm": 0.2972900910976631, "learning_rate": 7.928720563302216e-06, "loss": 0.3605, "step": 2517 }, { "epoch": 1.113220064128552, "grad_norm": 0.3442618548827303, "learning_rate": 7.926635479288632e-06, "loss": 0.3281, "step": 2518 }, { "epoch": 1.113662330004054, "grad_norm": 0.3051719799513858, "learning_rate": 7.924549620790905e-06, "loss": 0.3329, "step": 2519 }, { "epoch": 1.1141045958795563, "grad_norm": 0.31370834712119094, "learning_rate": 7.922462988361027e-06, "loss": 0.3945, "step": 2520 }, { "epoch": 1.1145468617550585, "grad_norm": 0.32440371639290894, "learning_rate": 7.920375582551185e-06, "loss": 0.2969, "step": 2521 }, { "epoch": 1.1149891276305606, "grad_norm": 0.3443171231622827, "learning_rate": 7.918287403913776e-06, "loss": 0.358, "step": 2522 }, { "epoch": 1.1154313935060627, "grad_norm": 0.30764033460914036, "learning_rate": 7.916198453001404e-06, "loss": 0.3045, "step": 2523 }, { "epoch": 1.1158736593815648, "grad_norm": 0.2842852974601881, "learning_rate": 7.91410873036687e-06, "loss": 0.3142, "step": 2524 }, { "epoch": 1.1163159252570671, "grad_norm": 0.3658513878352526, "learning_rate": 7.912018236563187e-06, "loss": 0.3695, "step": 2525 }, { "epoch": 1.1167581911325692, "grad_norm": 0.3468763704955441, "learning_rate": 7.909926972143567e-06, "loss": 0.3351, "step": 2526 }, { "epoch": 1.1172004570080714, "grad_norm": 0.3931005131881645, "learning_rate": 7.90783493766143e-06, "loss": 0.3502, "step": 2527 }, { "epoch": 1.1176427228835735, "grad_norm": 0.3534153028365865, "learning_rate": 7.905742133670393e-06, "loss": 0.3216, "step": 2528 }, { "epoch": 1.1180849887590756, "grad_norm": 0.3771272062949214, "learning_rate": 7.903648560724283e-06, "loss": 0.3488, "step": 2529 }, { "epoch": 1.118527254634578, "grad_norm": 0.3683785830857635, "learning_rate": 7.901554219377125e-06, "loss": 0.3136, "step": 2530 }, { "epoch": 1.11896952051008, "grad_norm": 0.2878882578482892, "learning_rate": 7.899459110183158e-06, "loss": 0.3391, "step": 2531 }, { "epoch": 1.1194117863855821, "grad_norm": 0.3498161711566392, "learning_rate": 7.897363233696809e-06, "loss": 0.3554, "step": 2532 }, { "epoch": 1.1198540522610843, "grad_norm": 0.3417770300474191, "learning_rate": 7.895266590472723e-06, "loss": 0.3291, "step": 2533 }, { "epoch": 1.1202963181365864, "grad_norm": 0.28586969895893366, "learning_rate": 7.893169181065737e-06, "loss": 0.2999, "step": 2534 }, { "epoch": 1.1207385840120887, "grad_norm": 0.2664458664372235, "learning_rate": 7.891071006030892e-06, "loss": 0.3114, "step": 2535 }, { "epoch": 1.1211808498875908, "grad_norm": 0.33125018308792953, "learning_rate": 7.888972065923441e-06, "loss": 0.3155, "step": 2536 }, { "epoch": 1.121623115763093, "grad_norm": 0.31771810421843727, "learning_rate": 7.886872361298828e-06, "loss": 0.33, "step": 2537 }, { "epoch": 1.122065381638595, "grad_norm": 0.34102041688432544, "learning_rate": 7.884771892712704e-06, "loss": 0.3282, "step": 2538 }, { "epoch": 1.1225076475140972, "grad_norm": 0.3358559931605298, "learning_rate": 7.882670660720923e-06, "loss": 0.3351, "step": 2539 }, { "epoch": 1.1229499133895993, "grad_norm": 0.24110022221955935, "learning_rate": 7.880568665879542e-06, "loss": 0.3034, "step": 2540 }, { "epoch": 1.1233921792651016, "grad_norm": 0.36000295371278673, "learning_rate": 7.878465908744815e-06, "loss": 0.3135, "step": 2541 }, { "epoch": 1.1238344451406037, "grad_norm": 0.32875016100711724, "learning_rate": 7.876362389873205e-06, "loss": 0.342, "step": 2542 }, { "epoch": 1.1242767110161058, "grad_norm": 0.32210145428005893, "learning_rate": 7.874258109821366e-06, "loss": 0.3564, "step": 2543 }, { "epoch": 1.124718976891608, "grad_norm": 0.2767159654666495, "learning_rate": 7.872153069146165e-06, "loss": 0.2839, "step": 2544 }, { "epoch": 1.1251612427671103, "grad_norm": 0.3397213800736937, "learning_rate": 7.870047268404664e-06, "loss": 0.3008, "step": 2545 }, { "epoch": 1.1256035086426124, "grad_norm": 0.32059571608803245, "learning_rate": 7.867940708154126e-06, "loss": 0.3213, "step": 2546 }, { "epoch": 1.1260457745181145, "grad_norm": 0.32119800583572233, "learning_rate": 7.865833388952016e-06, "loss": 0.326, "step": 2547 }, { "epoch": 1.1264880403936166, "grad_norm": 0.32293770955409307, "learning_rate": 7.863725311356002e-06, "loss": 0.3615, "step": 2548 }, { "epoch": 1.1269303062691187, "grad_norm": 0.32673208694604966, "learning_rate": 7.86161647592395e-06, "loss": 0.3297, "step": 2549 }, { "epoch": 1.1273725721446208, "grad_norm": 0.43582956698571085, "learning_rate": 7.859506883213927e-06, "loss": 0.3396, "step": 2550 }, { "epoch": 1.1278148380201232, "grad_norm": 0.321215873600389, "learning_rate": 7.857396533784201e-06, "loss": 0.3448, "step": 2551 }, { "epoch": 1.1282571038956253, "grad_norm": 0.35150975873218754, "learning_rate": 7.85528542819324e-06, "loss": 0.3574, "step": 2552 }, { "epoch": 1.1286993697711274, "grad_norm": 0.276724399979837, "learning_rate": 7.85317356699971e-06, "loss": 0.334, "step": 2553 }, { "epoch": 1.1291416356466295, "grad_norm": 0.3454440103040142, "learning_rate": 7.851060950762483e-06, "loss": 0.3373, "step": 2554 }, { "epoch": 1.1295839015221318, "grad_norm": 0.511476878635314, "learning_rate": 7.848947580040624e-06, "loss": 0.2981, "step": 2555 }, { "epoch": 1.130026167397634, "grad_norm": 0.2842102642141102, "learning_rate": 7.846833455393401e-06, "loss": 0.3076, "step": 2556 }, { "epoch": 1.130468433273136, "grad_norm": 0.3142910707369942, "learning_rate": 7.844718577380282e-06, "loss": 0.3304, "step": 2557 }, { "epoch": 1.1309106991486382, "grad_norm": 0.35084111136261054, "learning_rate": 7.842602946560932e-06, "loss": 0.3264, "step": 2558 }, { "epoch": 1.1313529650241403, "grad_norm": 0.3116147993632232, "learning_rate": 7.840486563495218e-06, "loss": 0.3536, "step": 2559 }, { "epoch": 1.1317952308996424, "grad_norm": 0.31287109921855627, "learning_rate": 7.838369428743202e-06, "loss": 0.326, "step": 2560 }, { "epoch": 1.1322374967751447, "grad_norm": 0.29821979421142303, "learning_rate": 7.836251542865151e-06, "loss": 0.3069, "step": 2561 }, { "epoch": 1.1326797626506468, "grad_norm": 0.30342213317968536, "learning_rate": 7.834132906421525e-06, "loss": 0.3517, "step": 2562 }, { "epoch": 1.133122028526149, "grad_norm": 0.30857182122620136, "learning_rate": 7.832013519972987e-06, "loss": 0.3138, "step": 2563 }, { "epoch": 1.133564294401651, "grad_norm": 0.30642862404014104, "learning_rate": 7.829893384080394e-06, "loss": 0.3497, "step": 2564 }, { "epoch": 1.1340065602771532, "grad_norm": 0.3492094742165556, "learning_rate": 7.827772499304806e-06, "loss": 0.3425, "step": 2565 }, { "epoch": 1.1344488261526555, "grad_norm": 0.2856467992720953, "learning_rate": 7.825650866207478e-06, "loss": 0.3136, "step": 2566 }, { "epoch": 1.1348910920281576, "grad_norm": 0.33434863031552453, "learning_rate": 7.823528485349862e-06, "loss": 0.3704, "step": 2567 }, { "epoch": 1.1353333579036597, "grad_norm": 0.25273732686777833, "learning_rate": 7.821405357293615e-06, "loss": 0.3271, "step": 2568 }, { "epoch": 1.1357756237791619, "grad_norm": 0.31924804810291585, "learning_rate": 7.819281482600582e-06, "loss": 0.3497, "step": 2569 }, { "epoch": 1.136217889654664, "grad_norm": 0.3108425267832084, "learning_rate": 7.817156861832812e-06, "loss": 0.3713, "step": 2570 }, { "epoch": 1.1366601555301663, "grad_norm": 0.3166323479712193, "learning_rate": 7.81503149555255e-06, "loss": 0.3099, "step": 2571 }, { "epoch": 1.1371024214056684, "grad_norm": 0.36236440500250566, "learning_rate": 7.812905384322236e-06, "loss": 0.3527, "step": 2572 }, { "epoch": 1.1375446872811705, "grad_norm": 0.3436931846405585, "learning_rate": 7.81077852870451e-06, "loss": 0.3235, "step": 2573 }, { "epoch": 1.1379869531566726, "grad_norm": 0.4137679931902678, "learning_rate": 7.80865092926221e-06, "loss": 0.3322, "step": 2574 }, { "epoch": 1.1384292190321748, "grad_norm": 0.3410183562004175, "learning_rate": 7.806522586558365e-06, "loss": 0.3558, "step": 2575 }, { "epoch": 1.138871484907677, "grad_norm": 0.38247833746370785, "learning_rate": 7.804393501156208e-06, "loss": 0.3678, "step": 2576 }, { "epoch": 1.1393137507831792, "grad_norm": 0.27449217154830696, "learning_rate": 7.802263673619165e-06, "loss": 0.3093, "step": 2577 }, { "epoch": 1.1397560166586813, "grad_norm": 0.2765855921133011, "learning_rate": 7.800133104510856e-06, "loss": 0.3194, "step": 2578 }, { "epoch": 1.1401982825341834, "grad_norm": 0.3641392289642196, "learning_rate": 7.798001794395103e-06, "loss": 0.3354, "step": 2579 }, { "epoch": 1.1406405484096855, "grad_norm": 0.2893625575692501, "learning_rate": 7.795869743835916e-06, "loss": 0.3208, "step": 2580 }, { "epoch": 1.1410828142851879, "grad_norm": 0.28937219082543414, "learning_rate": 7.793736953397512e-06, "loss": 0.334, "step": 2581 }, { "epoch": 1.14152508016069, "grad_norm": 0.32510502796491864, "learning_rate": 7.791603423644294e-06, "loss": 0.3125, "step": 2582 }, { "epoch": 1.141967346036192, "grad_norm": 0.34572022733962376, "learning_rate": 7.789469155140863e-06, "loss": 0.362, "step": 2583 }, { "epoch": 1.1424096119116942, "grad_norm": 0.32963555698001035, "learning_rate": 7.78733414845202e-06, "loss": 0.3726, "step": 2584 }, { "epoch": 1.1428518777871963, "grad_norm": 0.28897256415493566, "learning_rate": 7.785198404142756e-06, "loss": 0.3423, "step": 2585 }, { "epoch": 1.1432941436626987, "grad_norm": 0.27601241109665337, "learning_rate": 7.783061922778259e-06, "loss": 0.3069, "step": 2586 }, { "epoch": 1.1437364095382008, "grad_norm": 0.27628977769780666, "learning_rate": 7.780924704923915e-06, "loss": 0.331, "step": 2587 }, { "epoch": 1.1441786754137029, "grad_norm": 0.34371152842388936, "learning_rate": 7.778786751145298e-06, "loss": 0.3566, "step": 2588 }, { "epoch": 1.144620941289205, "grad_norm": 0.3083835343862542, "learning_rate": 7.776648062008186e-06, "loss": 0.3129, "step": 2589 }, { "epoch": 1.145063207164707, "grad_norm": 0.34339452658416425, "learning_rate": 7.774508638078542e-06, "loss": 0.3723, "step": 2590 }, { "epoch": 1.1455054730402092, "grad_norm": 0.26393959395628, "learning_rate": 7.772368479922532e-06, "loss": 0.3111, "step": 2591 }, { "epoch": 1.1459477389157116, "grad_norm": 0.3021751309929743, "learning_rate": 7.770227588106508e-06, "loss": 0.366, "step": 2592 }, { "epoch": 1.1463900047912137, "grad_norm": 0.34390726276194683, "learning_rate": 7.768085963197024e-06, "loss": 0.3436, "step": 2593 }, { "epoch": 1.1468322706667158, "grad_norm": 0.37145925815862696, "learning_rate": 7.765943605760824e-06, "loss": 0.3311, "step": 2594 }, { "epoch": 1.147274536542218, "grad_norm": 0.23539810746930853, "learning_rate": 7.763800516364846e-06, "loss": 0.336, "step": 2595 }, { "epoch": 1.1477168024177202, "grad_norm": 0.32172260991111734, "learning_rate": 7.761656695576218e-06, "loss": 0.341, "step": 2596 }, { "epoch": 1.1481590682932223, "grad_norm": 0.34439050632308543, "learning_rate": 7.759512143962274e-06, "loss": 0.3236, "step": 2597 }, { "epoch": 1.1486013341687245, "grad_norm": 0.31169913904491303, "learning_rate": 7.757366862090527e-06, "loss": 0.3183, "step": 2598 }, { "epoch": 1.1490436000442266, "grad_norm": 0.3198412087973043, "learning_rate": 7.755220850528689e-06, "loss": 0.3336, "step": 2599 }, { "epoch": 1.1494858659197287, "grad_norm": 0.29254415684806223, "learning_rate": 7.753074109844667e-06, "loss": 0.3708, "step": 2600 }, { "epoch": 1.1499281317952308, "grad_norm": 0.31654832758330836, "learning_rate": 7.75092664060656e-06, "loss": 0.3455, "step": 2601 }, { "epoch": 1.1503703976707331, "grad_norm": 0.36573101299648864, "learning_rate": 7.748778443382658e-06, "loss": 0.3689, "step": 2602 }, { "epoch": 1.1508126635462352, "grad_norm": 0.26170040402841477, "learning_rate": 7.746629518741446e-06, "loss": 0.3248, "step": 2603 }, { "epoch": 1.1512549294217373, "grad_norm": 0.2652066832376624, "learning_rate": 7.744479867251598e-06, "loss": 0.2955, "step": 2604 }, { "epoch": 1.1516971952972395, "grad_norm": 0.3004026572462606, "learning_rate": 7.742329489481985e-06, "loss": 0.3022, "step": 2605 }, { "epoch": 1.1521394611727418, "grad_norm": 0.2768247876037381, "learning_rate": 7.740178386001665e-06, "loss": 0.3437, "step": 2606 }, { "epoch": 1.152581727048244, "grad_norm": 0.34653059236272565, "learning_rate": 7.738026557379894e-06, "loss": 0.331, "step": 2607 }, { "epoch": 1.153023992923746, "grad_norm": 0.29194039250018566, "learning_rate": 7.735874004186115e-06, "loss": 0.3195, "step": 2608 }, { "epoch": 1.1534662587992481, "grad_norm": 0.41428649317177746, "learning_rate": 7.733720726989962e-06, "loss": 0.3283, "step": 2609 }, { "epoch": 1.1539085246747502, "grad_norm": 0.33713918363023687, "learning_rate": 7.731566726361268e-06, "loss": 0.3819, "step": 2610 }, { "epoch": 1.1543507905502524, "grad_norm": 0.32086480947747803, "learning_rate": 7.729412002870051e-06, "loss": 0.3573, "step": 2611 }, { "epoch": 1.1547930564257547, "grad_norm": 0.28301431036014063, "learning_rate": 7.72725655708652e-06, "loss": 0.3053, "step": 2612 }, { "epoch": 1.1552353223012568, "grad_norm": 0.4261934104592435, "learning_rate": 7.725100389581078e-06, "loss": 0.3618, "step": 2613 }, { "epoch": 1.155677588176759, "grad_norm": 0.30180981400008466, "learning_rate": 7.722943500924319e-06, "loss": 0.3489, "step": 2614 }, { "epoch": 1.156119854052261, "grad_norm": 0.30826137592529, "learning_rate": 7.720785891687023e-06, "loss": 0.3119, "step": 2615 }, { "epoch": 1.1565621199277631, "grad_norm": 0.3579288135344878, "learning_rate": 7.718627562440168e-06, "loss": 0.3337, "step": 2616 }, { "epoch": 1.1570043858032655, "grad_norm": 0.3509162660055347, "learning_rate": 7.71646851375492e-06, "loss": 0.3037, "step": 2617 }, { "epoch": 1.1574466516787676, "grad_norm": 0.2820534056228366, "learning_rate": 7.714308746202632e-06, "loss": 0.3091, "step": 2618 }, { "epoch": 1.1578889175542697, "grad_norm": 0.3950518760530127, "learning_rate": 7.71214826035485e-06, "loss": 0.3594, "step": 2619 }, { "epoch": 1.1583311834297718, "grad_norm": 0.3163431501631582, "learning_rate": 7.709987056783308e-06, "loss": 0.335, "step": 2620 }, { "epoch": 1.158773449305274, "grad_norm": 0.3328261332049425, "learning_rate": 7.707825136059937e-06, "loss": 0.3347, "step": 2621 }, { "epoch": 1.1592157151807763, "grad_norm": 0.29782767732064186, "learning_rate": 7.705662498756846e-06, "loss": 0.3364, "step": 2622 }, { "epoch": 1.1596579810562784, "grad_norm": 0.28305275024725707, "learning_rate": 7.703499145446344e-06, "loss": 0.3342, "step": 2623 }, { "epoch": 1.1601002469317805, "grad_norm": 0.3106083413688921, "learning_rate": 7.701335076700925e-06, "loss": 0.3223, "step": 2624 }, { "epoch": 1.1605425128072826, "grad_norm": 0.34820011866716094, "learning_rate": 7.699170293093271e-06, "loss": 0.3218, "step": 2625 }, { "epoch": 1.1609847786827847, "grad_norm": 0.3128720128213096, "learning_rate": 7.697004795196257e-06, "loss": 0.2894, "step": 2626 }, { "epoch": 1.161427044558287, "grad_norm": 0.527484217295893, "learning_rate": 7.694838583582946e-06, "loss": 0.35, "step": 2627 }, { "epoch": 1.1618693104337892, "grad_norm": 0.316018849402418, "learning_rate": 7.692671658826588e-06, "loss": 0.3299, "step": 2628 }, { "epoch": 1.1623115763092913, "grad_norm": 0.33566909568758696, "learning_rate": 7.69050402150062e-06, "loss": 0.365, "step": 2629 }, { "epoch": 1.1627538421847934, "grad_norm": 0.29431045149404317, "learning_rate": 7.688335672178674e-06, "loss": 0.3318, "step": 2630 }, { "epoch": 1.1631961080602955, "grad_norm": 0.407892384564212, "learning_rate": 7.686166611434563e-06, "loss": 0.3354, "step": 2631 }, { "epoch": 1.1636383739357978, "grad_norm": 0.3160642460194738, "learning_rate": 7.683996839842295e-06, "loss": 0.289, "step": 2632 }, { "epoch": 1.1640806398113, "grad_norm": 0.3250595693042296, "learning_rate": 7.68182635797606e-06, "loss": 0.3552, "step": 2633 }, { "epoch": 1.164522905686802, "grad_norm": 0.316909392242248, "learning_rate": 7.679655166410245e-06, "loss": 0.3083, "step": 2634 }, { "epoch": 1.1649651715623042, "grad_norm": 0.32104251431801223, "learning_rate": 7.677483265719415e-06, "loss": 0.2921, "step": 2635 }, { "epoch": 1.1654074374378063, "grad_norm": 0.3043548380094738, "learning_rate": 7.675310656478322e-06, "loss": 0.3684, "step": 2636 }, { "epoch": 1.1658497033133086, "grad_norm": 0.34805315949140775, "learning_rate": 7.673137339261916e-06, "loss": 0.3488, "step": 2637 }, { "epoch": 1.1662919691888107, "grad_norm": 0.28526201992194594, "learning_rate": 7.670963314645327e-06, "loss": 0.3346, "step": 2638 }, { "epoch": 1.1667342350643128, "grad_norm": 0.2889884333164704, "learning_rate": 7.668788583203872e-06, "loss": 0.3398, "step": 2639 }, { "epoch": 1.167176500939815, "grad_norm": 0.30818349607657075, "learning_rate": 7.666613145513058e-06, "loss": 0.3514, "step": 2640 }, { "epoch": 1.167618766815317, "grad_norm": 0.2861555248134211, "learning_rate": 7.664437002148577e-06, "loss": 0.3205, "step": 2641 }, { "epoch": 1.1680610326908192, "grad_norm": 0.2947507316930295, "learning_rate": 7.66226015368631e-06, "loss": 0.3208, "step": 2642 }, { "epoch": 1.1685032985663215, "grad_norm": 0.3152205681741334, "learning_rate": 7.66008260070232e-06, "loss": 0.3201, "step": 2643 }, { "epoch": 1.1689455644418236, "grad_norm": 0.356864725159752, "learning_rate": 7.65790434377286e-06, "loss": 0.3244, "step": 2644 }, { "epoch": 1.1693878303173257, "grad_norm": 0.31799929704801827, "learning_rate": 7.655725383474369e-06, "loss": 0.3262, "step": 2645 }, { "epoch": 1.1698300961928279, "grad_norm": 0.27818785921441136, "learning_rate": 7.65354572038347e-06, "loss": 0.3086, "step": 2646 }, { "epoch": 1.1702723620683302, "grad_norm": 0.39146623060106855, "learning_rate": 7.651365355076976e-06, "loss": 0.3362, "step": 2647 }, { "epoch": 1.1707146279438323, "grad_norm": 0.3008360546941396, "learning_rate": 7.649184288131883e-06, "loss": 0.3135, "step": 2648 }, { "epoch": 1.1711568938193344, "grad_norm": 0.3050713642382945, "learning_rate": 7.647002520125372e-06, "loss": 0.3049, "step": 2649 }, { "epoch": 1.1715991596948365, "grad_norm": 0.3219201363013307, "learning_rate": 7.644820051634813e-06, "loss": 0.33, "step": 2650 }, { "epoch": 1.1720414255703386, "grad_norm": 0.25933178740407553, "learning_rate": 7.642636883237757e-06, "loss": 0.3047, "step": 2651 }, { "epoch": 1.1724836914458407, "grad_norm": 0.3084039495570679, "learning_rate": 7.640453015511942e-06, "loss": 0.3356, "step": 2652 }, { "epoch": 1.172925957321343, "grad_norm": 0.3213497253507001, "learning_rate": 7.638268449035293e-06, "loss": 0.3489, "step": 2653 }, { "epoch": 1.1733682231968452, "grad_norm": 0.2785125227287828, "learning_rate": 7.636083184385916e-06, "loss": 0.3296, "step": 2654 }, { "epoch": 1.1738104890723473, "grad_norm": 0.34442331160003165, "learning_rate": 7.633897222142105e-06, "loss": 0.3506, "step": 2655 }, { "epoch": 1.1742527549478494, "grad_norm": 0.2540200733896265, "learning_rate": 7.631710562882339e-06, "loss": 0.3163, "step": 2656 }, { "epoch": 1.1746950208233518, "grad_norm": 0.305589378018398, "learning_rate": 7.629523207185279e-06, "loss": 0.3623, "step": 2657 }, { "epoch": 1.1751372866988539, "grad_norm": 0.3693922753179348, "learning_rate": 7.627335155629773e-06, "loss": 0.3394, "step": 2658 }, { "epoch": 1.175579552574356, "grad_norm": 0.34089898679597336, "learning_rate": 7.625146408794847e-06, "loss": 0.3492, "step": 2659 }, { "epoch": 1.176021818449858, "grad_norm": 0.39509018958747205, "learning_rate": 7.622956967259719e-06, "loss": 0.3865, "step": 2660 }, { "epoch": 1.1764640843253602, "grad_norm": 0.36373328723644816, "learning_rate": 7.6207668316037875e-06, "loss": 0.2894, "step": 2661 }, { "epoch": 1.1769063502008623, "grad_norm": 0.3161679896715477, "learning_rate": 7.618576002406634e-06, "loss": 0.2949, "step": 2662 }, { "epoch": 1.1773486160763647, "grad_norm": 0.2889148221537746, "learning_rate": 7.616384480248022e-06, "loss": 0.3182, "step": 2663 }, { "epoch": 1.1777908819518668, "grad_norm": 0.3365620195916639, "learning_rate": 7.6141922657079045e-06, "loss": 0.3144, "step": 2664 }, { "epoch": 1.1782331478273689, "grad_norm": 0.32421555575463235, "learning_rate": 7.611999359366409e-06, "loss": 0.3028, "step": 2665 }, { "epoch": 1.178675413702871, "grad_norm": 0.28053823133506695, "learning_rate": 7.609805761803856e-06, "loss": 0.3067, "step": 2666 }, { "epoch": 1.1791176795783733, "grad_norm": 0.2897992463718617, "learning_rate": 7.607611473600738e-06, "loss": 0.2954, "step": 2667 }, { "epoch": 1.1795599454538754, "grad_norm": 0.3499476839759406, "learning_rate": 7.605416495337741e-06, "loss": 0.346, "step": 2668 }, { "epoch": 1.1800022113293775, "grad_norm": 0.28887445677027157, "learning_rate": 7.603220827595724e-06, "loss": 0.3385, "step": 2669 }, { "epoch": 1.1804444772048797, "grad_norm": 0.2442037241145402, "learning_rate": 7.601024470955737e-06, "loss": 0.3209, "step": 2670 }, { "epoch": 1.1808867430803818, "grad_norm": 0.40366825376478255, "learning_rate": 7.598827425999006e-06, "loss": 0.3356, "step": 2671 }, { "epoch": 1.1813290089558839, "grad_norm": 0.3534364017211103, "learning_rate": 7.596629693306943e-06, "loss": 0.3378, "step": 2672 }, { "epoch": 1.1817712748313862, "grad_norm": 0.34305973808241397, "learning_rate": 7.594431273461137e-06, "loss": 0.3599, "step": 2673 }, { "epoch": 1.1822135407068883, "grad_norm": 0.4178232386383731, "learning_rate": 7.5922321670433675e-06, "loss": 0.3451, "step": 2674 }, { "epoch": 1.1826558065823904, "grad_norm": 0.35252059608676406, "learning_rate": 7.590032374635586e-06, "loss": 0.3383, "step": 2675 }, { "epoch": 1.1830980724578926, "grad_norm": 0.3226742613448063, "learning_rate": 7.587831896819933e-06, "loss": 0.339, "step": 2676 }, { "epoch": 1.1835403383333947, "grad_norm": 0.2971841467721923, "learning_rate": 7.585630734178726e-06, "loss": 0.3597, "step": 2677 }, { "epoch": 1.183982604208897, "grad_norm": 0.39175576768679043, "learning_rate": 7.583428887294464e-06, "loss": 0.3311, "step": 2678 }, { "epoch": 1.1844248700843991, "grad_norm": 0.30071239296433794, "learning_rate": 7.5812263567498315e-06, "loss": 0.3145, "step": 2679 }, { "epoch": 1.1848671359599012, "grad_norm": 0.3674834099607733, "learning_rate": 7.579023143127686e-06, "loss": 0.3465, "step": 2680 }, { "epoch": 1.1853094018354033, "grad_norm": 0.29289270785128213, "learning_rate": 7.576819247011076e-06, "loss": 0.3108, "step": 2681 }, { "epoch": 1.1857516677109055, "grad_norm": 0.33484330673030943, "learning_rate": 7.574614668983222e-06, "loss": 0.3672, "step": 2682 }, { "epoch": 1.1861939335864078, "grad_norm": 0.3043736887347173, "learning_rate": 7.5724094096275255e-06, "loss": 0.3588, "step": 2683 }, { "epoch": 1.18663619946191, "grad_norm": 0.3598299684113478, "learning_rate": 7.570203469527575e-06, "loss": 0.334, "step": 2684 }, { "epoch": 1.187078465337412, "grad_norm": 0.736465613028508, "learning_rate": 7.567996849267134e-06, "loss": 0.4111, "step": 2685 }, { "epoch": 1.1875207312129141, "grad_norm": 0.28171804020101154, "learning_rate": 7.565789549430145e-06, "loss": 0.3047, "step": 2686 }, { "epoch": 1.1879629970884162, "grad_norm": 0.2999565210327094, "learning_rate": 7.563581570600733e-06, "loss": 0.3419, "step": 2687 }, { "epoch": 1.1884052629639186, "grad_norm": 0.3200280462517869, "learning_rate": 7.561372913363203e-06, "loss": 0.3287, "step": 2688 }, { "epoch": 1.1888475288394207, "grad_norm": 0.2892103707645431, "learning_rate": 7.559163578302038e-06, "loss": 0.3307, "step": 2689 }, { "epoch": 1.1892897947149228, "grad_norm": 0.25162041307632543, "learning_rate": 7.5569535660019e-06, "loss": 0.3406, "step": 2690 }, { "epoch": 1.189732060590425, "grad_norm": 0.31250583555542333, "learning_rate": 7.554742877047631e-06, "loss": 0.3401, "step": 2691 }, { "epoch": 1.190174326465927, "grad_norm": 0.3636404518805782, "learning_rate": 7.552531512024252e-06, "loss": 0.3508, "step": 2692 }, { "epoch": 1.1906165923414291, "grad_norm": 0.3239164617884892, "learning_rate": 7.550319471516966e-06, "loss": 0.3331, "step": 2693 }, { "epoch": 1.1910588582169315, "grad_norm": 0.376576285742854, "learning_rate": 7.548106756111149e-06, "loss": 0.3125, "step": 2694 }, { "epoch": 1.1915011240924336, "grad_norm": 0.31204668771246163, "learning_rate": 7.545893366392358e-06, "loss": 0.3319, "step": 2695 }, { "epoch": 1.1919433899679357, "grad_norm": 0.35209973445436576, "learning_rate": 7.54367930294633e-06, "loss": 0.3623, "step": 2696 }, { "epoch": 1.1923856558434378, "grad_norm": 0.27976528433996, "learning_rate": 7.541464566358978e-06, "loss": 0.311, "step": 2697 }, { "epoch": 1.1928279217189401, "grad_norm": 0.290947926062334, "learning_rate": 7.539249157216396e-06, "loss": 0.2831, "step": 2698 }, { "epoch": 1.1932701875944423, "grad_norm": 0.29083070949431444, "learning_rate": 7.537033076104853e-06, "loss": 0.3565, "step": 2699 }, { "epoch": 1.1937124534699444, "grad_norm": 0.34008007655243055, "learning_rate": 7.534816323610797e-06, "loss": 0.3368, "step": 2700 }, { "epoch": 1.1941547193454465, "grad_norm": 0.2576514560862372, "learning_rate": 7.532598900320855e-06, "loss": 0.3302, "step": 2701 }, { "epoch": 1.1945969852209486, "grad_norm": 0.33768404122873674, "learning_rate": 7.530380806821829e-06, "loss": 0.3125, "step": 2702 }, { "epoch": 1.1950392510964507, "grad_norm": 0.2857838512101091, "learning_rate": 7.5281620437006985e-06, "loss": 0.3402, "step": 2703 }, { "epoch": 1.195481516971953, "grad_norm": 0.2763251678908791, "learning_rate": 7.5259426115446235e-06, "loss": 0.308, "step": 2704 }, { "epoch": 1.1959237828474552, "grad_norm": 0.34451220356307954, "learning_rate": 7.523722510940939e-06, "loss": 0.35, "step": 2705 }, { "epoch": 1.1963660487229573, "grad_norm": 0.2646043268985935, "learning_rate": 7.521501742477155e-06, "loss": 0.3486, "step": 2706 }, { "epoch": 1.1968083145984594, "grad_norm": 0.35980563971524715, "learning_rate": 7.5192803067409595e-06, "loss": 0.3518, "step": 2707 }, { "epoch": 1.1972505804739617, "grad_norm": 0.36933809517626126, "learning_rate": 7.51705820432022e-06, "loss": 0.3646, "step": 2708 }, { "epoch": 1.1976928463494638, "grad_norm": 0.25719127029114, "learning_rate": 7.514835435802978e-06, "loss": 0.317, "step": 2709 }, { "epoch": 1.198135112224966, "grad_norm": 0.2802683765276107, "learning_rate": 7.512612001777448e-06, "loss": 0.3372, "step": 2710 }, { "epoch": 1.198577378100468, "grad_norm": 0.321631391744834, "learning_rate": 7.510387902832027e-06, "loss": 0.3333, "step": 2711 }, { "epoch": 1.1990196439759702, "grad_norm": 0.3208129198248076, "learning_rate": 7.508163139555282e-06, "loss": 0.3628, "step": 2712 }, { "epoch": 1.1994619098514723, "grad_norm": 0.2789155734633606, "learning_rate": 7.5059377125359635e-06, "loss": 0.321, "step": 2713 }, { "epoch": 1.1999041757269746, "grad_norm": 0.30588190869381654, "learning_rate": 7.503711622362988e-06, "loss": 0.352, "step": 2714 }, { "epoch": 1.2003464416024767, "grad_norm": 0.3267327014586866, "learning_rate": 7.501484869625453e-06, "loss": 0.3119, "step": 2715 }, { "epoch": 1.2007887074779788, "grad_norm": 0.27558151458785884, "learning_rate": 7.499257454912633e-06, "loss": 0.3331, "step": 2716 }, { "epoch": 1.201230973353481, "grad_norm": 0.32903791879616834, "learning_rate": 7.497029378813973e-06, "loss": 0.3127, "step": 2717 }, { "epoch": 1.2016732392289833, "grad_norm": 0.2831614083739107, "learning_rate": 7.4948006419190976e-06, "loss": 0.3227, "step": 2718 }, { "epoch": 1.2021155051044854, "grad_norm": 0.2986700277361118, "learning_rate": 7.492571244817801e-06, "loss": 0.3074, "step": 2719 }, { "epoch": 1.2025577709799875, "grad_norm": 0.3708981911577816, "learning_rate": 7.4903411881000575e-06, "loss": 0.3551, "step": 2720 }, { "epoch": 1.2030000368554896, "grad_norm": 0.24400763669607714, "learning_rate": 7.488110472356012e-06, "loss": 0.3297, "step": 2721 }, { "epoch": 1.2034423027309917, "grad_norm": 0.336082164562363, "learning_rate": 7.485879098175988e-06, "loss": 0.3068, "step": 2722 }, { "epoch": 1.2038845686064938, "grad_norm": 0.2744264320691764, "learning_rate": 7.483647066150477e-06, "loss": 0.3143, "step": 2723 }, { "epoch": 1.2043268344819962, "grad_norm": 0.2912226240718767, "learning_rate": 7.48141437687015e-06, "loss": 0.2749, "step": 2724 }, { "epoch": 1.2047691003574983, "grad_norm": 0.2906186118718587, "learning_rate": 7.47918103092585e-06, "loss": 0.3041, "step": 2725 }, { "epoch": 1.2052113662330004, "grad_norm": 0.3599293985833776, "learning_rate": 7.476947028908595e-06, "loss": 0.3552, "step": 2726 }, { "epoch": 1.2056536321085025, "grad_norm": 0.3392462393788916, "learning_rate": 7.4747123714095725e-06, "loss": 0.3055, "step": 2727 }, { "epoch": 1.2060958979840046, "grad_norm": 0.3221886607330652, "learning_rate": 7.47247705902015e-06, "loss": 0.3354, "step": 2728 }, { "epoch": 1.206538163859507, "grad_norm": 0.3262697494909593, "learning_rate": 7.470241092331862e-06, "loss": 0.3766, "step": 2729 }, { "epoch": 1.206980429735009, "grad_norm": 0.2955689856327142, "learning_rate": 7.4680044719364205e-06, "loss": 0.3378, "step": 2730 }, { "epoch": 1.2074226956105112, "grad_norm": 0.35003355826215404, "learning_rate": 7.465767198425708e-06, "loss": 0.3261, "step": 2731 }, { "epoch": 1.2078649614860133, "grad_norm": 0.314789226685438, "learning_rate": 7.463529272391783e-06, "loss": 0.3069, "step": 2732 }, { "epoch": 1.2083072273615154, "grad_norm": 0.3808070952021339, "learning_rate": 7.461290694426873e-06, "loss": 0.3235, "step": 2733 }, { "epoch": 1.2087494932370177, "grad_norm": 0.2904754241730779, "learning_rate": 7.459051465123377e-06, "loss": 0.2776, "step": 2734 }, { "epoch": 1.2091917591125199, "grad_norm": 0.3338159980429927, "learning_rate": 7.456811585073872e-06, "loss": 0.3581, "step": 2735 }, { "epoch": 1.209634024988022, "grad_norm": 0.36602532291666245, "learning_rate": 7.454571054871104e-06, "loss": 0.3431, "step": 2736 }, { "epoch": 1.210076290863524, "grad_norm": 0.29352509725395937, "learning_rate": 7.452329875107989e-06, "loss": 0.3373, "step": 2737 }, { "epoch": 1.2105185567390262, "grad_norm": 0.3293296922687775, "learning_rate": 7.450088046377618e-06, "loss": 0.316, "step": 2738 }, { "epoch": 1.2109608226145285, "grad_norm": 0.3402539316052585, "learning_rate": 7.447845569273252e-06, "loss": 0.3396, "step": 2739 }, { "epoch": 1.2114030884900306, "grad_norm": 0.31902447055240446, "learning_rate": 7.445602444388328e-06, "loss": 0.3253, "step": 2740 }, { "epoch": 1.2118453543655328, "grad_norm": 0.2678424308092245, "learning_rate": 7.443358672316445e-06, "loss": 0.2911, "step": 2741 }, { "epoch": 1.2122876202410349, "grad_norm": 0.3282273220618521, "learning_rate": 7.4411142536513825e-06, "loss": 0.3526, "step": 2742 }, { "epoch": 1.212729886116537, "grad_norm": 0.32123128498187736, "learning_rate": 7.438869188987088e-06, "loss": 0.3229, "step": 2743 }, { "epoch": 1.213172151992039, "grad_norm": 0.3570074201230806, "learning_rate": 7.436623478917678e-06, "loss": 0.3314, "step": 2744 }, { "epoch": 1.2136144178675414, "grad_norm": 0.2639151836268743, "learning_rate": 7.434377124037442e-06, "loss": 0.3223, "step": 2745 }, { "epoch": 1.2140566837430435, "grad_norm": 0.3206999501856125, "learning_rate": 7.432130124940839e-06, "loss": 0.3196, "step": 2746 }, { "epoch": 1.2144989496185457, "grad_norm": 0.25679730573279635, "learning_rate": 7.429882482222499e-06, "loss": 0.3467, "step": 2747 }, { "epoch": 1.2149412154940478, "grad_norm": 0.28237748820622627, "learning_rate": 7.427634196477223e-06, "loss": 0.309, "step": 2748 }, { "epoch": 1.21538348136955, "grad_norm": 0.2779454274421375, "learning_rate": 7.425385268299981e-06, "loss": 0.3293, "step": 2749 }, { "epoch": 1.2158257472450522, "grad_norm": 0.30162303342994506, "learning_rate": 7.423135698285914e-06, "loss": 0.337, "step": 2750 }, { "epoch": 1.2162680131205543, "grad_norm": 0.32872439862574643, "learning_rate": 7.420885487030331e-06, "loss": 0.3405, "step": 2751 }, { "epoch": 1.2167102789960564, "grad_norm": 0.25390123406535675, "learning_rate": 7.418634635128714e-06, "loss": 0.3172, "step": 2752 }, { "epoch": 1.2171525448715586, "grad_norm": 0.35149145414493066, "learning_rate": 7.416383143176711e-06, "loss": 0.3426, "step": 2753 }, { "epoch": 1.2175948107470607, "grad_norm": 0.29848845176874794, "learning_rate": 7.414131011770142e-06, "loss": 0.3649, "step": 2754 }, { "epoch": 1.218037076622563, "grad_norm": 0.2970299962865938, "learning_rate": 7.411878241504993e-06, "loss": 0.3673, "step": 2755 }, { "epoch": 1.218479342498065, "grad_norm": 0.3357976594751877, "learning_rate": 7.409624832977423e-06, "loss": 0.3141, "step": 2756 }, { "epoch": 1.2189216083735672, "grad_norm": 0.32373870866983423, "learning_rate": 7.407370786783757e-06, "loss": 0.3435, "step": 2757 }, { "epoch": 1.2193638742490693, "grad_norm": 0.34038814670095663, "learning_rate": 7.405116103520493e-06, "loss": 0.3157, "step": 2758 }, { "epoch": 1.2198061401245717, "grad_norm": 0.40064317687037604, "learning_rate": 7.402860783784291e-06, "loss": 0.3266, "step": 2759 }, { "epoch": 1.2202484060000738, "grad_norm": 0.40371897433571496, "learning_rate": 7.400604828171984e-06, "loss": 0.3576, "step": 2760 }, { "epoch": 1.220690671875576, "grad_norm": 0.24816877053086894, "learning_rate": 7.398348237280572e-06, "loss": 0.2999, "step": 2761 }, { "epoch": 1.221132937751078, "grad_norm": 0.3296566453276002, "learning_rate": 7.3960910117072224e-06, "loss": 0.352, "step": 2762 }, { "epoch": 1.2215752036265801, "grad_norm": 0.3878015761121927, "learning_rate": 7.393833152049275e-06, "loss": 0.3745, "step": 2763 }, { "epoch": 1.2220174695020822, "grad_norm": 0.4040616773376715, "learning_rate": 7.39157465890423e-06, "loss": 0.3567, "step": 2764 }, { "epoch": 1.2224597353775846, "grad_norm": 0.33703454423587886, "learning_rate": 7.389315532869759e-06, "loss": 0.3402, "step": 2765 }, { "epoch": 1.2229020012530867, "grad_norm": 0.320524779140645, "learning_rate": 7.387055774543704e-06, "loss": 0.3165, "step": 2766 }, { "epoch": 1.2233442671285888, "grad_norm": 0.2688404705078385, "learning_rate": 7.3847953845240705e-06, "loss": 0.3379, "step": 2767 }, { "epoch": 1.223786533004091, "grad_norm": 0.3480885543756884, "learning_rate": 7.382534363409031e-06, "loss": 0.3423, "step": 2768 }, { "epoch": 1.2242287988795932, "grad_norm": 0.395859237948336, "learning_rate": 7.380272711796928e-06, "loss": 0.3251, "step": 2769 }, { "epoch": 1.2246710647550954, "grad_norm": 0.274946265696941, "learning_rate": 7.378010430286265e-06, "loss": 0.3548, "step": 2770 }, { "epoch": 1.2251133306305975, "grad_norm": 0.381817311489499, "learning_rate": 7.375747519475722e-06, "loss": 0.3368, "step": 2771 }, { "epoch": 1.2255555965060996, "grad_norm": 0.33072071688411514, "learning_rate": 7.3734839799641355e-06, "loss": 0.3378, "step": 2772 }, { "epoch": 1.2259978623816017, "grad_norm": 0.3532061782960097, "learning_rate": 7.371219812350513e-06, "loss": 0.3078, "step": 2773 }, { "epoch": 1.2264401282571038, "grad_norm": 0.29839853243800496, "learning_rate": 7.36895501723403e-06, "loss": 0.3374, "step": 2774 }, { "epoch": 1.2268823941326061, "grad_norm": 0.29277587424875945, "learning_rate": 7.366689595214024e-06, "loss": 0.3231, "step": 2775 }, { "epoch": 1.2273246600081082, "grad_norm": 0.320450800587198, "learning_rate": 7.364423546890002e-06, "loss": 0.2939, "step": 2776 }, { "epoch": 1.2277669258836104, "grad_norm": 0.28553639145393583, "learning_rate": 7.362156872861632e-06, "loss": 0.2902, "step": 2777 }, { "epoch": 1.2282091917591125, "grad_norm": 0.3088631605324475, "learning_rate": 7.359889573728753e-06, "loss": 0.3173, "step": 2778 }, { "epoch": 1.2286514576346146, "grad_norm": 0.32443730531438675, "learning_rate": 7.357621650091368e-06, "loss": 0.358, "step": 2779 }, { "epoch": 1.229093723510117, "grad_norm": 0.29871190078906823, "learning_rate": 7.355353102549642e-06, "loss": 0.3267, "step": 2780 }, { "epoch": 1.229535989385619, "grad_norm": 0.32315613809249627, "learning_rate": 7.353083931703906e-06, "loss": 0.3146, "step": 2781 }, { "epoch": 1.2299782552611211, "grad_norm": 0.35045059828996983, "learning_rate": 7.3508141381546625e-06, "loss": 0.3341, "step": 2782 }, { "epoch": 1.2304205211366233, "grad_norm": 0.3574220649091052, "learning_rate": 7.348543722502569e-06, "loss": 0.3418, "step": 2783 }, { "epoch": 1.2308627870121254, "grad_norm": 0.31810422977971753, "learning_rate": 7.346272685348454e-06, "loss": 0.3489, "step": 2784 }, { "epoch": 1.2313050528876277, "grad_norm": 0.3629659720860997, "learning_rate": 7.3440010272933105e-06, "loss": 0.3605, "step": 2785 }, { "epoch": 1.2317473187631298, "grad_norm": 0.30239192559104466, "learning_rate": 7.341728748938291e-06, "loss": 0.3615, "step": 2786 }, { "epoch": 1.232189584638632, "grad_norm": 0.30817115813542734, "learning_rate": 7.339455850884716e-06, "loss": 0.3368, "step": 2787 }, { "epoch": 1.232631850514134, "grad_norm": 0.29745632303498276, "learning_rate": 7.33718233373407e-06, "loss": 0.3514, "step": 2788 }, { "epoch": 1.2330741163896362, "grad_norm": 0.3127632151460013, "learning_rate": 7.334908198088e-06, "loss": 0.3403, "step": 2789 }, { "epoch": 1.2335163822651385, "grad_norm": 0.2717628165102778, "learning_rate": 7.332633444548317e-06, "loss": 0.2882, "step": 2790 }, { "epoch": 1.2339586481406406, "grad_norm": 0.31846490154419066, "learning_rate": 7.330358073716997e-06, "loss": 0.3041, "step": 2791 }, { "epoch": 1.2344009140161427, "grad_norm": 0.319187761491587, "learning_rate": 7.328082086196175e-06, "loss": 0.3455, "step": 2792 }, { "epoch": 1.2348431798916448, "grad_norm": 0.3150602972095242, "learning_rate": 7.325805482588156e-06, "loss": 0.362, "step": 2793 }, { "epoch": 1.235285445767147, "grad_norm": 0.34149745649093616, "learning_rate": 7.323528263495402e-06, "loss": 0.3612, "step": 2794 }, { "epoch": 1.235727711642649, "grad_norm": 0.35397673268354773, "learning_rate": 7.321250429520543e-06, "loss": 0.3778, "step": 2795 }, { "epoch": 1.2361699775181514, "grad_norm": 0.40054999093523297, "learning_rate": 7.318971981266365e-06, "loss": 0.3517, "step": 2796 }, { "epoch": 1.2366122433936535, "grad_norm": 0.29058452608241414, "learning_rate": 7.316692919335821e-06, "loss": 0.3506, "step": 2797 }, { "epoch": 1.2370545092691556, "grad_norm": 0.3727398997649597, "learning_rate": 7.3144132443320284e-06, "loss": 0.3291, "step": 2798 }, { "epoch": 1.2374967751446577, "grad_norm": 0.27343163501484424, "learning_rate": 7.3121329568582624e-06, "loss": 0.355, "step": 2799 }, { "epoch": 1.23793904102016, "grad_norm": 0.3416895616950166, "learning_rate": 7.309852057517963e-06, "loss": 0.302, "step": 2800 }, { "epoch": 1.2383813068956622, "grad_norm": 0.2804539878097299, "learning_rate": 7.307570546914731e-06, "loss": 0.3164, "step": 2801 }, { "epoch": 1.2388235727711643, "grad_norm": 0.3030332808327692, "learning_rate": 7.305288425652328e-06, "loss": 0.3429, "step": 2802 }, { "epoch": 1.2392658386466664, "grad_norm": 0.334308517399552, "learning_rate": 7.303005694334681e-06, "loss": 0.361, "step": 2803 }, { "epoch": 1.2397081045221685, "grad_norm": 0.35346705453522137, "learning_rate": 7.300722353565874e-06, "loss": 0.3282, "step": 2804 }, { "epoch": 1.2401503703976706, "grad_norm": 0.2986966307910958, "learning_rate": 7.298438403950154e-06, "loss": 0.328, "step": 2805 }, { "epoch": 1.240592636273173, "grad_norm": 0.3061119474267093, "learning_rate": 7.296153846091931e-06, "loss": 0.3039, "step": 2806 }, { "epoch": 1.241034902148675, "grad_norm": 0.40988722243913056, "learning_rate": 7.293868680595774e-06, "loss": 0.3295, "step": 2807 }, { "epoch": 1.2414771680241772, "grad_norm": 0.31961753902924556, "learning_rate": 7.291582908066412e-06, "loss": 0.3641, "step": 2808 }, { "epoch": 1.2419194338996793, "grad_norm": 0.3434502157586088, "learning_rate": 7.2892965291087356e-06, "loss": 0.3089, "step": 2809 }, { "epoch": 1.2423616997751816, "grad_norm": 0.29853439361590595, "learning_rate": 7.287009544327796e-06, "loss": 0.2953, "step": 2810 }, { "epoch": 1.2428039656506837, "grad_norm": 0.3431613852141804, "learning_rate": 7.284721954328806e-06, "loss": 0.3241, "step": 2811 }, { "epoch": 1.2432462315261859, "grad_norm": 0.2860781821788095, "learning_rate": 7.2824337597171355e-06, "loss": 0.3355, "step": 2812 }, { "epoch": 1.243688497401688, "grad_norm": 0.32355104965565784, "learning_rate": 7.280144961098317e-06, "loss": 0.3666, "step": 2813 }, { "epoch": 1.24413076327719, "grad_norm": 0.3411516302623045, "learning_rate": 7.277855559078042e-06, "loss": 0.29, "step": 2814 }, { "epoch": 1.2445730291526922, "grad_norm": 0.3227563669278219, "learning_rate": 7.275565554262163e-06, "loss": 0.3242, "step": 2815 }, { "epoch": 1.2450152950281945, "grad_norm": 0.3449151197497652, "learning_rate": 7.2732749472566876e-06, "loss": 0.3516, "step": 2816 }, { "epoch": 1.2454575609036966, "grad_norm": 0.3071815245232627, "learning_rate": 7.2709837386677885e-06, "loss": 0.3278, "step": 2817 }, { "epoch": 1.2458998267791987, "grad_norm": 0.2918247048704013, "learning_rate": 7.268691929101794e-06, "loss": 0.3277, "step": 2818 }, { "epoch": 1.2463420926547009, "grad_norm": 0.3435528805675966, "learning_rate": 7.266399519165193e-06, "loss": 0.3262, "step": 2819 }, { "epoch": 1.2467843585302032, "grad_norm": 0.309649239527417, "learning_rate": 7.2641065094646326e-06, "loss": 0.3356, "step": 2820 }, { "epoch": 1.2472266244057053, "grad_norm": 0.32044307354681434, "learning_rate": 7.261812900606916e-06, "loss": 0.3359, "step": 2821 }, { "epoch": 1.2476688902812074, "grad_norm": 0.3153403976652908, "learning_rate": 7.2595186931990126e-06, "loss": 0.3385, "step": 2822 }, { "epoch": 1.2481111561567095, "grad_norm": 0.2831886856223335, "learning_rate": 7.25722388784804e-06, "loss": 0.3177, "step": 2823 }, { "epoch": 1.2485534220322116, "grad_norm": 0.294178473098734, "learning_rate": 7.254928485161283e-06, "loss": 0.3481, "step": 2824 }, { "epoch": 1.2489956879077138, "grad_norm": 0.3833246921876173, "learning_rate": 7.252632485746181e-06, "loss": 0.3483, "step": 2825 }, { "epoch": 1.249437953783216, "grad_norm": 0.24953789541052382, "learning_rate": 7.250335890210327e-06, "loss": 0.3116, "step": 2826 }, { "epoch": 1.2498802196587182, "grad_norm": 0.3092941643925893, "learning_rate": 7.248038699161481e-06, "loss": 0.3311, "step": 2827 }, { "epoch": 1.2503224855342203, "grad_norm": 0.2822412116466464, "learning_rate": 7.245740913207551e-06, "loss": 0.3182, "step": 2828 }, { "epoch": 1.2507647514097224, "grad_norm": 0.3801012554934967, "learning_rate": 7.243442532956608e-06, "loss": 0.3528, "step": 2829 }, { "epoch": 1.2512070172852248, "grad_norm": 0.3492155615796882, "learning_rate": 7.241143559016883e-06, "loss": 0.325, "step": 2830 }, { "epoch": 1.2516492831607269, "grad_norm": 0.35024884937724626, "learning_rate": 7.2388439919967546e-06, "loss": 0.3171, "step": 2831 }, { "epoch": 1.252091549036229, "grad_norm": 0.3184718822396098, "learning_rate": 7.236543832504768e-06, "loss": 0.337, "step": 2832 }, { "epoch": 1.252533814911731, "grad_norm": 0.29666429659553495, "learning_rate": 7.234243081149619e-06, "loss": 0.3477, "step": 2833 }, { "epoch": 1.2529760807872332, "grad_norm": 0.38337134855796573, "learning_rate": 7.231941738540163e-06, "loss": 0.3671, "step": 2834 }, { "epoch": 1.2534183466627353, "grad_norm": 0.2995142629986089, "learning_rate": 7.229639805285409e-06, "loss": 0.338, "step": 2835 }, { "epoch": 1.2538606125382374, "grad_norm": 0.3762908072118133, "learning_rate": 7.227337281994528e-06, "loss": 0.3229, "step": 2836 }, { "epoch": 1.2543028784137398, "grad_norm": 0.30719444287172293, "learning_rate": 7.225034169276841e-06, "loss": 0.3065, "step": 2837 }, { "epoch": 1.2547451442892419, "grad_norm": 0.4590141602851411, "learning_rate": 7.22273046774183e-06, "loss": 0.3121, "step": 2838 }, { "epoch": 1.255187410164744, "grad_norm": 0.32365753872049374, "learning_rate": 7.220426177999126e-06, "loss": 0.371, "step": 2839 }, { "epoch": 1.2556296760402461, "grad_norm": 0.40271450802368786, "learning_rate": 7.2181213006585225e-06, "loss": 0.3489, "step": 2840 }, { "epoch": 1.2560719419157484, "grad_norm": 0.3470506413911257, "learning_rate": 7.215815836329968e-06, "loss": 0.3375, "step": 2841 }, { "epoch": 1.2565142077912506, "grad_norm": 0.3757407563636082, "learning_rate": 7.21350978562356e-06, "loss": 0.3377, "step": 2842 }, { "epoch": 1.2569564736667527, "grad_norm": 0.3009096205379153, "learning_rate": 7.21120314914956e-06, "loss": 0.3396, "step": 2843 }, { "epoch": 1.2573987395422548, "grad_norm": 0.2691040028297743, "learning_rate": 7.2088959275183756e-06, "loss": 0.3294, "step": 2844 }, { "epoch": 1.257841005417757, "grad_norm": 0.4477896952926006, "learning_rate": 7.206588121340576e-06, "loss": 0.3789, "step": 2845 }, { "epoch": 1.258283271293259, "grad_norm": 0.2780622069822066, "learning_rate": 7.204279731226884e-06, "loss": 0.3027, "step": 2846 }, { "epoch": 1.2587255371687613, "grad_norm": 0.3449329202382045, "learning_rate": 7.201970757788172e-06, "loss": 0.3686, "step": 2847 }, { "epoch": 1.2591678030442635, "grad_norm": 0.35615934609155686, "learning_rate": 7.199661201635474e-06, "loss": 0.3508, "step": 2848 }, { "epoch": 1.2596100689197656, "grad_norm": 0.3240584679186129, "learning_rate": 7.197351063379973e-06, "loss": 0.3516, "step": 2849 }, { "epoch": 1.2600523347952677, "grad_norm": 0.3797509244550354, "learning_rate": 7.195040343633006e-06, "loss": 0.3178, "step": 2850 }, { "epoch": 1.26049460067077, "grad_norm": 0.3016262849911208, "learning_rate": 7.192729043006069e-06, "loss": 0.3155, "step": 2851 }, { "epoch": 1.2609368665462721, "grad_norm": 0.3433418653688949, "learning_rate": 7.1904171621108055e-06, "loss": 0.3612, "step": 2852 }, { "epoch": 1.2613791324217742, "grad_norm": 0.27864644382823534, "learning_rate": 7.188104701559019e-06, "loss": 0.3447, "step": 2853 }, { "epoch": 1.2618213982972764, "grad_norm": 0.3194010145576787, "learning_rate": 7.185791661962658e-06, "loss": 0.3633, "step": 2854 }, { "epoch": 1.2622636641727785, "grad_norm": 0.37705372149795996, "learning_rate": 7.18347804393383e-06, "loss": 0.3422, "step": 2855 }, { "epoch": 1.2627059300482806, "grad_norm": 0.3472758829402589, "learning_rate": 7.181163848084797e-06, "loss": 0.324, "step": 2856 }, { "epoch": 1.263148195923783, "grad_norm": 0.33497048000109236, "learning_rate": 7.178849075027972e-06, "loss": 0.3091, "step": 2857 }, { "epoch": 1.263590461799285, "grad_norm": 0.30557516054293704, "learning_rate": 7.176533725375916e-06, "loss": 0.3604, "step": 2858 }, { "epoch": 1.2640327276747871, "grad_norm": 0.32809234461015674, "learning_rate": 7.17421779974135e-06, "loss": 0.3486, "step": 2859 }, { "epoch": 1.2644749935502893, "grad_norm": 0.3103351711748016, "learning_rate": 7.171901298737142e-06, "loss": 0.3437, "step": 2860 }, { "epoch": 1.2649172594257916, "grad_norm": 0.39758007525234007, "learning_rate": 7.169584222976318e-06, "loss": 0.3382, "step": 2861 }, { "epoch": 1.2653595253012937, "grad_norm": 0.33985302336717205, "learning_rate": 7.167266573072048e-06, "loss": 0.3412, "step": 2862 }, { "epoch": 1.2658017911767958, "grad_norm": 0.3233190769284284, "learning_rate": 7.164948349637661e-06, "loss": 0.3094, "step": 2863 }, { "epoch": 1.266244057052298, "grad_norm": 0.34887567620560067, "learning_rate": 7.1626295532866366e-06, "loss": 0.3494, "step": 2864 }, { "epoch": 1.2666863229278, "grad_norm": 0.39903201360592877, "learning_rate": 7.160310184632602e-06, "loss": 0.3414, "step": 2865 }, { "epoch": 1.2671285888033021, "grad_norm": 0.3770696566143477, "learning_rate": 7.15799024428934e-06, "loss": 0.3128, "step": 2866 }, { "epoch": 1.2675708546788045, "grad_norm": 0.32617786487937533, "learning_rate": 7.155669732870783e-06, "loss": 0.3335, "step": 2867 }, { "epoch": 1.2680131205543066, "grad_norm": 0.3043118740832742, "learning_rate": 7.153348650991014e-06, "loss": 0.3432, "step": 2868 }, { "epoch": 1.2684553864298087, "grad_norm": 0.2775832085046592, "learning_rate": 7.15102699926427e-06, "loss": 0.29, "step": 2869 }, { "epoch": 1.2688976523053108, "grad_norm": 0.28871220514285295, "learning_rate": 7.148704778304934e-06, "loss": 0.3141, "step": 2870 }, { "epoch": 1.2693399181808132, "grad_norm": 0.407366853954977, "learning_rate": 7.146381988727542e-06, "loss": 0.3223, "step": 2871 }, { "epoch": 1.2697821840563153, "grad_norm": 0.30560377233544267, "learning_rate": 7.144058631146782e-06, "loss": 0.3296, "step": 2872 }, { "epoch": 1.2702244499318174, "grad_norm": 0.31877580545887924, "learning_rate": 7.1417347061774924e-06, "loss": 0.3263, "step": 2873 }, { "epoch": 1.2706667158073195, "grad_norm": 0.2972881167882199, "learning_rate": 7.1394102144346564e-06, "loss": 0.3462, "step": 2874 }, { "epoch": 1.2711089816828216, "grad_norm": 0.2537842916550827, "learning_rate": 7.1370851565334145e-06, "loss": 0.3308, "step": 2875 }, { "epoch": 1.2715512475583237, "grad_norm": 0.33013395758968744, "learning_rate": 7.134759533089051e-06, "loss": 0.3241, "step": 2876 }, { "epoch": 1.271993513433826, "grad_norm": 0.36941882887813116, "learning_rate": 7.132433344717006e-06, "loss": 0.3308, "step": 2877 }, { "epoch": 1.2724357793093282, "grad_norm": 0.28581869818828926, "learning_rate": 7.130106592032861e-06, "loss": 0.3468, "step": 2878 }, { "epoch": 1.2728780451848303, "grad_norm": 0.29582381784309897, "learning_rate": 7.127779275652354e-06, "loss": 0.3302, "step": 2879 }, { "epoch": 1.2733203110603324, "grad_norm": 0.3184553600015006, "learning_rate": 7.125451396191371e-06, "loss": 0.3215, "step": 2880 }, { "epoch": 1.2737625769358347, "grad_norm": 0.3433794063048956, "learning_rate": 7.123122954265942e-06, "loss": 0.3447, "step": 2881 }, { "epoch": 1.2742048428113368, "grad_norm": 0.28791316115998455, "learning_rate": 7.120793950492253e-06, "loss": 0.3777, "step": 2882 }, { "epoch": 1.274647108686839, "grad_norm": 0.2807458498235287, "learning_rate": 7.118464385486632e-06, "loss": 0.332, "step": 2883 }, { "epoch": 1.275089374562341, "grad_norm": 0.29477925463196253, "learning_rate": 7.11613425986556e-06, "loss": 0.3197, "step": 2884 }, { "epoch": 1.2755316404378432, "grad_norm": 0.31430048782570263, "learning_rate": 7.113803574245666e-06, "loss": 0.3337, "step": 2885 }, { "epoch": 1.2759739063133453, "grad_norm": 0.3017545788821734, "learning_rate": 7.111472329243725e-06, "loss": 0.3159, "step": 2886 }, { "epoch": 1.2764161721888474, "grad_norm": 0.33453724481765496, "learning_rate": 7.109140525476659e-06, "loss": 0.3519, "step": 2887 }, { "epoch": 1.2768584380643497, "grad_norm": 0.3447282302953286, "learning_rate": 7.106808163561545e-06, "loss": 0.3494, "step": 2888 }, { "epoch": 1.2773007039398518, "grad_norm": 0.31808517672099723, "learning_rate": 7.1044752441155996e-06, "loss": 0.291, "step": 2889 }, { "epoch": 1.277742969815354, "grad_norm": 0.27619899369227985, "learning_rate": 7.102141767756191e-06, "loss": 0.3277, "step": 2890 }, { "epoch": 1.2781852356908563, "grad_norm": 0.32459409676905737, "learning_rate": 7.099807735100833e-06, "loss": 0.3097, "step": 2891 }, { "epoch": 1.2786275015663584, "grad_norm": 0.39951831507547636, "learning_rate": 7.097473146767189e-06, "loss": 0.3421, "step": 2892 }, { "epoch": 1.2790697674418605, "grad_norm": 0.4184276282340512, "learning_rate": 7.095138003373066e-06, "loss": 0.3358, "step": 2893 }, { "epoch": 1.2795120333173626, "grad_norm": 0.30794123664312295, "learning_rate": 7.092802305536423e-06, "loss": 0.3376, "step": 2894 }, { "epoch": 1.2799542991928647, "grad_norm": 0.43694067322001, "learning_rate": 7.0904660538753574e-06, "loss": 0.3, "step": 2895 }, { "epoch": 1.2803965650683669, "grad_norm": 0.28950424488220466, "learning_rate": 7.088129249008124e-06, "loss": 0.3499, "step": 2896 }, { "epoch": 1.280838830943869, "grad_norm": 0.4117204414662033, "learning_rate": 7.085791891553115e-06, "loss": 0.3269, "step": 2897 }, { "epoch": 1.2812810968193713, "grad_norm": 0.27943487303789766, "learning_rate": 7.083453982128874e-06, "loss": 0.3316, "step": 2898 }, { "epoch": 1.2817233626948734, "grad_norm": 0.27888385492637235, "learning_rate": 7.081115521354086e-06, "loss": 0.3243, "step": 2899 }, { "epoch": 1.2821656285703755, "grad_norm": 0.31526069333867707, "learning_rate": 7.078776509847588e-06, "loss": 0.3474, "step": 2900 }, { "epoch": 1.2826078944458776, "grad_norm": 0.40625906678477225, "learning_rate": 7.076436948228358e-06, "loss": 0.3804, "step": 2901 }, { "epoch": 1.28305016032138, "grad_norm": 0.45921394723732833, "learning_rate": 7.074096837115521e-06, "loss": 0.3073, "step": 2902 }, { "epoch": 1.283492426196882, "grad_norm": 0.30688690096068316, "learning_rate": 7.071756177128346e-06, "loss": 0.3258, "step": 2903 }, { "epoch": 1.2839346920723842, "grad_norm": 0.2862021385641247, "learning_rate": 7.069414968886254e-06, "loss": 0.2995, "step": 2904 }, { "epoch": 1.2843769579478863, "grad_norm": 0.27304770432662695, "learning_rate": 7.067073213008798e-06, "loss": 0.3151, "step": 2905 }, { "epoch": 1.2848192238233884, "grad_norm": 0.28328820541387373, "learning_rate": 7.06473091011569e-06, "loss": 0.342, "step": 2906 }, { "epoch": 1.2852614896988905, "grad_norm": 0.3156798447721979, "learning_rate": 7.062388060826778e-06, "loss": 0.3243, "step": 2907 }, { "epoch": 1.2857037555743929, "grad_norm": 0.33329914061264937, "learning_rate": 7.060044665762058e-06, "loss": 0.3482, "step": 2908 }, { "epoch": 1.286146021449895, "grad_norm": 0.3244365306922871, "learning_rate": 7.0577007255416684e-06, "loss": 0.3043, "step": 2909 }, { "epoch": 1.286588287325397, "grad_norm": 0.2947143934380308, "learning_rate": 7.055356240785893e-06, "loss": 0.3215, "step": 2910 }, { "epoch": 1.2870305532008992, "grad_norm": 0.2684974337306337, "learning_rate": 7.053011212115161e-06, "loss": 0.3136, "step": 2911 }, { "epoch": 1.2874728190764015, "grad_norm": 0.30114077732320926, "learning_rate": 7.050665640150045e-06, "loss": 0.3192, "step": 2912 }, { "epoch": 1.2879150849519037, "grad_norm": 0.3243056522363587, "learning_rate": 7.048319525511256e-06, "loss": 0.291, "step": 2913 }, { "epoch": 1.2883573508274058, "grad_norm": 0.2995657303786753, "learning_rate": 7.045972868819658e-06, "loss": 0.3272, "step": 2914 }, { "epoch": 1.2887996167029079, "grad_norm": 0.3899880032427385, "learning_rate": 7.0436256706962526e-06, "loss": 0.327, "step": 2915 }, { "epoch": 1.28924188257841, "grad_norm": 0.34069833964396423, "learning_rate": 7.0412779317621835e-06, "loss": 0.3282, "step": 2916 }, { "epoch": 1.289684148453912, "grad_norm": 0.28269516131248995, "learning_rate": 7.038929652638742e-06, "loss": 0.3165, "step": 2917 }, { "epoch": 1.2901264143294144, "grad_norm": 0.33550525532716474, "learning_rate": 7.036580833947359e-06, "loss": 0.2972, "step": 2918 }, { "epoch": 1.2905686802049166, "grad_norm": 0.26384668424467, "learning_rate": 7.03423147630961e-06, "loss": 0.3387, "step": 2919 }, { "epoch": 1.2910109460804187, "grad_norm": 0.2688236811406865, "learning_rate": 7.031881580347213e-06, "loss": 0.2843, "step": 2920 }, { "epoch": 1.2914532119559208, "grad_norm": 0.30775298267683154, "learning_rate": 7.0295311466820245e-06, "loss": 0.3243, "step": 2921 }, { "epoch": 1.2918954778314231, "grad_norm": 0.3068933469720099, "learning_rate": 7.027180175936051e-06, "loss": 0.3538, "step": 2922 }, { "epoch": 1.2923377437069252, "grad_norm": 0.2864976710959646, "learning_rate": 7.024828668731435e-06, "loss": 0.3291, "step": 2923 }, { "epoch": 1.2927800095824273, "grad_norm": 0.3334298336495684, "learning_rate": 7.022476625690462e-06, "loss": 0.3248, "step": 2924 }, { "epoch": 1.2932222754579294, "grad_norm": 0.3116180569716552, "learning_rate": 7.020124047435562e-06, "loss": 0.3418, "step": 2925 }, { "epoch": 1.2936645413334316, "grad_norm": 0.28200634875198305, "learning_rate": 7.017770934589301e-06, "loss": 0.3138, "step": 2926 }, { "epoch": 1.2941068072089337, "grad_norm": 0.31872193647898023, "learning_rate": 7.015417287774394e-06, "loss": 0.3211, "step": 2927 }, { "epoch": 1.294549073084436, "grad_norm": 0.28900255007047626, "learning_rate": 7.013063107613693e-06, "loss": 0.3189, "step": 2928 }, { "epoch": 1.2949913389599381, "grad_norm": 0.2722613349355211, "learning_rate": 7.010708394730189e-06, "loss": 0.3503, "step": 2929 }, { "epoch": 1.2954336048354402, "grad_norm": 0.322039811035656, "learning_rate": 7.008353149747019e-06, "loss": 0.3612, "step": 2930 }, { "epoch": 1.2958758707109423, "grad_norm": 0.36512321402950015, "learning_rate": 7.005997373287458e-06, "loss": 0.3401, "step": 2931 }, { "epoch": 1.2963181365864447, "grad_norm": 0.27349415613662986, "learning_rate": 7.003641065974921e-06, "loss": 0.3252, "step": 2932 }, { "epoch": 1.2967604024619468, "grad_norm": 0.36524150204125316, "learning_rate": 7.001284228432965e-06, "loss": 0.3174, "step": 2933 }, { "epoch": 1.297202668337449, "grad_norm": 0.33320617864786, "learning_rate": 6.998926861285288e-06, "loss": 0.3502, "step": 2934 }, { "epoch": 1.297644934212951, "grad_norm": 0.3345627084887014, "learning_rate": 6.996568965155725e-06, "loss": 0.2928, "step": 2935 }, { "epoch": 1.2980872000884531, "grad_norm": 0.29878534266703793, "learning_rate": 6.994210540668256e-06, "loss": 0.3439, "step": 2936 }, { "epoch": 1.2985294659639552, "grad_norm": 0.26628481267254056, "learning_rate": 6.991851588446994e-06, "loss": 0.3115, "step": 2937 }, { "epoch": 1.2989717318394574, "grad_norm": 0.290692118055727, "learning_rate": 6.9894921091161975e-06, "loss": 0.2884, "step": 2938 }, { "epoch": 1.2994139977149597, "grad_norm": 0.3735108724614331, "learning_rate": 6.987132103300263e-06, "loss": 0.3428, "step": 2939 }, { "epoch": 1.2998562635904618, "grad_norm": 0.3252196116290288, "learning_rate": 6.9847715716237245e-06, "loss": 0.3336, "step": 2940 }, { "epoch": 1.300298529465964, "grad_norm": 0.29434806321844253, "learning_rate": 6.982410514711256e-06, "loss": 0.3635, "step": 2941 }, { "epoch": 1.3007407953414662, "grad_norm": 0.2995799185521418, "learning_rate": 6.980048933187672e-06, "loss": 0.3218, "step": 2942 }, { "epoch": 1.3011830612169684, "grad_norm": 0.4185641546811666, "learning_rate": 6.977686827677926e-06, "loss": 0.3808, "step": 2943 }, { "epoch": 1.3016253270924705, "grad_norm": 0.3119413100069883, "learning_rate": 6.975324198807107e-06, "loss": 0.3361, "step": 2944 }, { "epoch": 1.3020675929679726, "grad_norm": 0.38576228556584274, "learning_rate": 6.972961047200444e-06, "loss": 0.2948, "step": 2945 }, { "epoch": 1.3025098588434747, "grad_norm": 0.2943392702609417, "learning_rate": 6.9705973734833055e-06, "loss": 0.312, "step": 2946 }, { "epoch": 1.3029521247189768, "grad_norm": 0.28443321595336074, "learning_rate": 6.968233178281198e-06, "loss": 0.3196, "step": 2947 }, { "epoch": 1.303394390594479, "grad_norm": 0.2958936887629846, "learning_rate": 6.965868462219767e-06, "loss": 0.3486, "step": 2948 }, { "epoch": 1.3038366564699813, "grad_norm": 0.30929763048956055, "learning_rate": 6.96350322592479e-06, "loss": 0.315, "step": 2949 }, { "epoch": 1.3042789223454834, "grad_norm": 0.28764360122276783, "learning_rate": 6.961137470022189e-06, "loss": 0.3434, "step": 2950 }, { "epoch": 1.3047211882209855, "grad_norm": 0.2667842627432786, "learning_rate": 6.958771195138021e-06, "loss": 0.3067, "step": 2951 }, { "epoch": 1.3051634540964876, "grad_norm": 0.31453965743144974, "learning_rate": 6.95640440189848e-06, "loss": 0.359, "step": 2952 }, { "epoch": 1.30560571997199, "grad_norm": 0.29485178339007834, "learning_rate": 6.954037090929897e-06, "loss": 0.3134, "step": 2953 }, { "epoch": 1.306047985847492, "grad_norm": 0.29046455280755806, "learning_rate": 6.951669262858741e-06, "loss": 0.3357, "step": 2954 }, { "epoch": 1.3064902517229942, "grad_norm": 0.5455756494698172, "learning_rate": 6.9493009183116165e-06, "loss": 0.3913, "step": 2955 }, { "epoch": 1.3069325175984963, "grad_norm": 0.26187307407188204, "learning_rate": 6.9469320579152674e-06, "loss": 0.297, "step": 2956 }, { "epoch": 1.3073747834739984, "grad_norm": 0.34204282158170346, "learning_rate": 6.9445626822965694e-06, "loss": 0.3449, "step": 2957 }, { "epoch": 1.3078170493495005, "grad_norm": 0.2675459290894739, "learning_rate": 6.942192792082539e-06, "loss": 0.3506, "step": 2958 }, { "epoch": 1.3082593152250028, "grad_norm": 0.3362047466734488, "learning_rate": 6.939822387900328e-06, "loss": 0.3393, "step": 2959 }, { "epoch": 1.308701581100505, "grad_norm": 0.40295570805722, "learning_rate": 6.937451470377222e-06, "loss": 0.3643, "step": 2960 }, { "epoch": 1.309143846976007, "grad_norm": 0.48806873126120337, "learning_rate": 6.9350800401406425e-06, "loss": 0.3301, "step": 2961 }, { "epoch": 1.3095861128515092, "grad_norm": 0.3189821060165589, "learning_rate": 6.932708097818151e-06, "loss": 0.3143, "step": 2962 }, { "epoch": 1.3100283787270115, "grad_norm": 0.370530599533979, "learning_rate": 6.930335644037441e-06, "loss": 0.3459, "step": 2963 }, { "epoch": 1.3104706446025136, "grad_norm": 0.2696851889042849, "learning_rate": 6.927962679426341e-06, "loss": 0.2941, "step": 2964 }, { "epoch": 1.3109129104780157, "grad_norm": 0.3180254678292644, "learning_rate": 6.925589204612816e-06, "loss": 0.3151, "step": 2965 }, { "epoch": 1.3113551763535178, "grad_norm": 0.3178689529589227, "learning_rate": 6.923215220224966e-06, "loss": 0.3232, "step": 2966 }, { "epoch": 1.31179744222902, "grad_norm": 0.3226512351829215, "learning_rate": 6.920840726891024e-06, "loss": 0.3454, "step": 2967 }, { "epoch": 1.312239708104522, "grad_norm": 0.2947556868041388, "learning_rate": 6.9184657252393605e-06, "loss": 0.3331, "step": 2968 }, { "epoch": 1.3126819739800244, "grad_norm": 0.30883078043273965, "learning_rate": 6.916090215898477e-06, "loss": 0.316, "step": 2969 }, { "epoch": 1.3131242398555265, "grad_norm": 0.3035298340818668, "learning_rate": 6.913714199497015e-06, "loss": 0.3085, "step": 2970 }, { "epoch": 1.3135665057310286, "grad_norm": 0.3154730758797362, "learning_rate": 6.911337676663745e-06, "loss": 0.3467, "step": 2971 }, { "epoch": 1.3140087716065307, "grad_norm": 0.35594462331671417, "learning_rate": 6.9089606480275724e-06, "loss": 0.3464, "step": 2972 }, { "epoch": 1.314451037482033, "grad_norm": 0.34135173721519535, "learning_rate": 6.906583114217539e-06, "loss": 0.3511, "step": 2973 }, { "epoch": 1.3148933033575352, "grad_norm": 0.2887726574345298, "learning_rate": 6.904205075862816e-06, "loss": 0.3333, "step": 2974 }, { "epoch": 1.3153355692330373, "grad_norm": 0.3096210808478177, "learning_rate": 6.901826533592714e-06, "loss": 0.3247, "step": 2975 }, { "epoch": 1.3157778351085394, "grad_norm": 0.3216258236285459, "learning_rate": 6.899447488036671e-06, "loss": 0.3638, "step": 2976 }, { "epoch": 1.3162201009840415, "grad_norm": 0.33140756997882914, "learning_rate": 6.897067939824262e-06, "loss": 0.3356, "step": 2977 }, { "epoch": 1.3166623668595436, "grad_norm": 0.3124282122478679, "learning_rate": 6.894687889585192e-06, "loss": 0.3267, "step": 2978 }, { "epoch": 1.317104632735046, "grad_norm": 0.3044791395224456, "learning_rate": 6.8923073379493044e-06, "loss": 0.3365, "step": 2979 }, { "epoch": 1.317546898610548, "grad_norm": 0.2881507306053628, "learning_rate": 6.889926285546569e-06, "loss": 0.366, "step": 2980 }, { "epoch": 1.3179891644860502, "grad_norm": 0.35232595985290993, "learning_rate": 6.88754473300709e-06, "loss": 0.3546, "step": 2981 }, { "epoch": 1.3184314303615523, "grad_norm": 0.2954557274047289, "learning_rate": 6.885162680961106e-06, "loss": 0.3062, "step": 2982 }, { "epoch": 1.3188736962370546, "grad_norm": 0.33943772681441453, "learning_rate": 6.882780130038987e-06, "loss": 0.343, "step": 2983 }, { "epoch": 1.3193159621125568, "grad_norm": 0.3246793521004377, "learning_rate": 6.8803970808712325e-06, "loss": 0.3016, "step": 2984 }, { "epoch": 1.3197582279880589, "grad_norm": 0.312667187626352, "learning_rate": 6.878013534088477e-06, "loss": 0.3549, "step": 2985 }, { "epoch": 1.320200493863561, "grad_norm": 0.3144004815272116, "learning_rate": 6.875629490321487e-06, "loss": 0.3131, "step": 2986 }, { "epoch": 1.320642759739063, "grad_norm": 0.3529732616964667, "learning_rate": 6.873244950201155e-06, "loss": 0.3452, "step": 2987 }, { "epoch": 1.3210850256145652, "grad_norm": 0.27206024843315413, "learning_rate": 6.870859914358513e-06, "loss": 0.2931, "step": 2988 }, { "epoch": 1.3215272914900673, "grad_norm": 0.28024183968141575, "learning_rate": 6.868474383424718e-06, "loss": 0.3078, "step": 2989 }, { "epoch": 1.3219695573655696, "grad_norm": 0.273664293050688, "learning_rate": 6.86608835803106e-06, "loss": 0.3375, "step": 2990 }, { "epoch": 1.3224118232410718, "grad_norm": 0.364283371014021, "learning_rate": 6.863701838808961e-06, "loss": 0.3391, "step": 2991 }, { "epoch": 1.3228540891165739, "grad_norm": 0.3613162481170739, "learning_rate": 6.861314826389969e-06, "loss": 0.3368, "step": 2992 }, { "epoch": 1.3232963549920762, "grad_norm": 0.29266267650776434, "learning_rate": 6.85892732140577e-06, "loss": 0.306, "step": 2993 }, { "epoch": 1.3237386208675783, "grad_norm": 0.32951800538176473, "learning_rate": 6.856539324488178e-06, "loss": 0.3347, "step": 2994 }, { "epoch": 1.3241808867430804, "grad_norm": 0.30814791189574087, "learning_rate": 6.8541508362691315e-06, "loss": 0.3555, "step": 2995 }, { "epoch": 1.3246231526185825, "grad_norm": 0.2201020188852901, "learning_rate": 6.851761857380703e-06, "loss": 0.3017, "step": 2996 }, { "epoch": 1.3250654184940847, "grad_norm": 0.33109247120857105, "learning_rate": 6.849372388455099e-06, "loss": 0.3213, "step": 2997 }, { "epoch": 1.3255076843695868, "grad_norm": 0.33311530103685444, "learning_rate": 6.846982430124649e-06, "loss": 0.3297, "step": 2998 }, { "epoch": 1.3259499502450889, "grad_norm": 0.2634534033126953, "learning_rate": 6.844591983021814e-06, "loss": 0.3564, "step": 2999 }, { "epoch": 1.3263922161205912, "grad_norm": 0.310649746764634, "learning_rate": 6.842201047779185e-06, "loss": 0.3266, "step": 3000 }, { "epoch": 1.3268344819960933, "grad_norm": 0.34743807765676554, "learning_rate": 6.839809625029485e-06, "loss": 0.3508, "step": 3001 }, { "epoch": 1.3272767478715954, "grad_norm": 0.29293379279300275, "learning_rate": 6.8374177154055636e-06, "loss": 0.3554, "step": 3002 }, { "epoch": 1.3277190137470976, "grad_norm": 0.3178308900130121, "learning_rate": 6.8350253195403924e-06, "loss": 0.288, "step": 3003 }, { "epoch": 1.3281612796225999, "grad_norm": 0.2962816100082815, "learning_rate": 6.832632438067085e-06, "loss": 0.3921, "step": 3004 }, { "epoch": 1.328603545498102, "grad_norm": 0.33132082189732515, "learning_rate": 6.830239071618874e-06, "loss": 0.3135, "step": 3005 }, { "epoch": 1.3290458113736041, "grad_norm": 0.9255363830250658, "learning_rate": 6.8278452208291235e-06, "loss": 0.3319, "step": 3006 }, { "epoch": 1.3294880772491062, "grad_norm": 0.27818001249307195, "learning_rate": 6.825450886331324e-06, "loss": 0.3371, "step": 3007 }, { "epoch": 1.3299303431246083, "grad_norm": 0.30952211735405544, "learning_rate": 6.823056068759095e-06, "loss": 0.2887, "step": 3008 }, { "epoch": 1.3303726090001105, "grad_norm": 0.26335887898017557, "learning_rate": 6.8206607687461875e-06, "loss": 0.3277, "step": 3009 }, { "epoch": 1.3308148748756128, "grad_norm": 0.3005421015360579, "learning_rate": 6.818264986926475e-06, "loss": 0.3401, "step": 3010 }, { "epoch": 1.331257140751115, "grad_norm": 0.30011204350971216, "learning_rate": 6.815868723933958e-06, "loss": 0.3282, "step": 3011 }, { "epoch": 1.331699406626617, "grad_norm": 0.2937428496088316, "learning_rate": 6.813471980402768e-06, "loss": 0.2917, "step": 3012 }, { "epoch": 1.3321416725021191, "grad_norm": 0.3377388711121045, "learning_rate": 6.8110747569671645e-06, "loss": 0.3462, "step": 3013 }, { "epoch": 1.3325839383776215, "grad_norm": 0.3126239338463056, "learning_rate": 6.80867705426153e-06, "loss": 0.3337, "step": 3014 }, { "epoch": 1.3330262042531236, "grad_norm": 0.37731195625546565, "learning_rate": 6.806278872920376e-06, "loss": 0.3355, "step": 3015 }, { "epoch": 1.3334684701286257, "grad_norm": 0.2738879848507797, "learning_rate": 6.803880213578337e-06, "loss": 0.3263, "step": 3016 }, { "epoch": 1.3339107360041278, "grad_norm": 0.30428166657743294, "learning_rate": 6.801481076870184e-06, "loss": 0.3404, "step": 3017 }, { "epoch": 1.33435300187963, "grad_norm": 0.24631819117478299, "learning_rate": 6.799081463430801e-06, "loss": 0.3096, "step": 3018 }, { "epoch": 1.334795267755132, "grad_norm": 0.3491774829381162, "learning_rate": 6.796681373895208e-06, "loss": 0.3271, "step": 3019 }, { "epoch": 1.3352375336306344, "grad_norm": 0.27584995604840207, "learning_rate": 6.794280808898547e-06, "loss": 0.3394, "step": 3020 }, { "epoch": 1.3356797995061365, "grad_norm": 0.23567379265975147, "learning_rate": 6.791879769076087e-06, "loss": 0.2909, "step": 3021 }, { "epoch": 1.3361220653816386, "grad_norm": 0.3531911182518449, "learning_rate": 6.789478255063223e-06, "loss": 0.3292, "step": 3022 }, { "epoch": 1.3365643312571407, "grad_norm": 0.2679191561548284, "learning_rate": 6.787076267495474e-06, "loss": 0.3236, "step": 3023 }, { "epoch": 1.337006597132643, "grad_norm": 0.2865351996339857, "learning_rate": 6.784673807008481e-06, "loss": 0.3229, "step": 3024 }, { "epoch": 1.3374488630081451, "grad_norm": 0.28967667720447626, "learning_rate": 6.7822708742380225e-06, "loss": 0.3266, "step": 3025 }, { "epoch": 1.3378911288836473, "grad_norm": 0.36994228740607826, "learning_rate": 6.779867469819989e-06, "loss": 0.3228, "step": 3026 }, { "epoch": 1.3383333947591494, "grad_norm": 0.3662953169462776, "learning_rate": 6.777463594390396e-06, "loss": 0.31, "step": 3027 }, { "epoch": 1.3387756606346515, "grad_norm": 0.48125486215604923, "learning_rate": 6.775059248585397e-06, "loss": 0.3265, "step": 3028 }, { "epoch": 1.3392179265101536, "grad_norm": 0.31795738366848314, "learning_rate": 6.7726544330412545e-06, "loss": 0.3739, "step": 3029 }, { "epoch": 1.339660192385656, "grad_norm": 0.3104922822412524, "learning_rate": 6.770249148394366e-06, "loss": 0.3122, "step": 3030 }, { "epoch": 1.340102458261158, "grad_norm": 0.3277340074613926, "learning_rate": 6.767843395281246e-06, "loss": 0.3403, "step": 3031 }, { "epoch": 1.3405447241366601, "grad_norm": 0.32368339639344684, "learning_rate": 6.765437174338536e-06, "loss": 0.3117, "step": 3032 }, { "epoch": 1.3409869900121623, "grad_norm": 0.2848423403414679, "learning_rate": 6.763030486203006e-06, "loss": 0.2997, "step": 3033 }, { "epoch": 1.3414292558876646, "grad_norm": 0.29319765996299746, "learning_rate": 6.760623331511539e-06, "loss": 0.3552, "step": 3034 }, { "epoch": 1.3418715217631667, "grad_norm": 0.2771702029104755, "learning_rate": 6.7582157109011494e-06, "loss": 0.3336, "step": 3035 }, { "epoch": 1.3423137876386688, "grad_norm": 0.30581772305271415, "learning_rate": 6.755807625008974e-06, "loss": 0.3213, "step": 3036 }, { "epoch": 1.342756053514171, "grad_norm": 0.34259401151010704, "learning_rate": 6.753399074472271e-06, "loss": 0.3268, "step": 3037 }, { "epoch": 1.343198319389673, "grad_norm": 0.3301656298784308, "learning_rate": 6.750990059928424e-06, "loss": 0.3335, "step": 3038 }, { "epoch": 1.3436405852651752, "grad_norm": 0.2797143477113785, "learning_rate": 6.748580582014934e-06, "loss": 0.362, "step": 3039 }, { "epoch": 1.3440828511406773, "grad_norm": 0.3861552063274548, "learning_rate": 6.74617064136943e-06, "loss": 0.3714, "step": 3040 }, { "epoch": 1.3445251170161796, "grad_norm": 0.38324888996110607, "learning_rate": 6.743760238629662e-06, "loss": 0.3561, "step": 3041 }, { "epoch": 1.3449673828916817, "grad_norm": 0.3328754090250341, "learning_rate": 6.741349374433502e-06, "loss": 0.3671, "step": 3042 }, { "epoch": 1.3454096487671838, "grad_norm": 0.30794190624388895, "learning_rate": 6.738938049418943e-06, "loss": 0.3119, "step": 3043 }, { "epoch": 1.3458519146426862, "grad_norm": 0.29067658967072213, "learning_rate": 6.736526264224101e-06, "loss": 0.3395, "step": 3044 }, { "epoch": 1.3462941805181883, "grad_norm": 0.435941892050125, "learning_rate": 6.7341140194872165e-06, "loss": 0.3715, "step": 3045 }, { "epoch": 1.3467364463936904, "grad_norm": 0.2757706198603174, "learning_rate": 6.731701315846646e-06, "loss": 0.3273, "step": 3046 }, { "epoch": 1.3471787122691925, "grad_norm": 0.29120502536442333, "learning_rate": 6.729288153940871e-06, "loss": 0.301, "step": 3047 }, { "epoch": 1.3476209781446946, "grad_norm": 0.2748083841455336, "learning_rate": 6.726874534408496e-06, "loss": 0.3025, "step": 3048 }, { "epoch": 1.3480632440201967, "grad_norm": 0.317129588796724, "learning_rate": 6.72446045788824e-06, "loss": 0.3639, "step": 3049 }, { "epoch": 1.3485055098956988, "grad_norm": 0.3316069400568861, "learning_rate": 6.7220459250189505e-06, "loss": 0.3176, "step": 3050 }, { "epoch": 1.3489477757712012, "grad_norm": 0.38729877437688975, "learning_rate": 6.719630936439589e-06, "loss": 0.3262, "step": 3051 }, { "epoch": 1.3493900416467033, "grad_norm": 0.34543238842269675, "learning_rate": 6.717215492789246e-06, "loss": 0.355, "step": 3052 }, { "epoch": 1.3498323075222054, "grad_norm": 0.3314762237830678, "learning_rate": 6.714799594707124e-06, "loss": 0.3179, "step": 3053 }, { "epoch": 1.3502745733977075, "grad_norm": 0.3131911465333786, "learning_rate": 6.71238324283255e-06, "loss": 0.3182, "step": 3054 }, { "epoch": 1.3507168392732098, "grad_norm": 0.3081793375460918, "learning_rate": 6.709966437804969e-06, "loss": 0.3438, "step": 3055 }, { "epoch": 1.351159105148712, "grad_norm": 0.38082969269689265, "learning_rate": 6.707549180263949e-06, "loss": 0.3298, "step": 3056 }, { "epoch": 1.351601371024214, "grad_norm": 0.26334173594822013, "learning_rate": 6.7051314708491755e-06, "loss": 0.3068, "step": 3057 }, { "epoch": 1.3520436368997162, "grad_norm": 0.3019245188117046, "learning_rate": 6.702713310200455e-06, "loss": 0.322, "step": 3058 }, { "epoch": 1.3524859027752183, "grad_norm": 0.33258729590523695, "learning_rate": 6.700294698957709e-06, "loss": 0.308, "step": 3059 }, { "epoch": 1.3529281686507204, "grad_norm": 0.3529267792623306, "learning_rate": 6.697875637760986e-06, "loss": 0.3433, "step": 3060 }, { "epoch": 1.3533704345262227, "grad_norm": 0.3561158223585359, "learning_rate": 6.695456127250446e-06, "loss": 0.3513, "step": 3061 }, { "epoch": 1.3538127004017249, "grad_norm": 0.5854886673034961, "learning_rate": 6.6930361680663735e-06, "loss": 0.3834, "step": 3062 }, { "epoch": 1.354254966277227, "grad_norm": 0.3760066567071668, "learning_rate": 6.690615760849169e-06, "loss": 0.3462, "step": 3063 }, { "epoch": 1.354697232152729, "grad_norm": 0.2594097736338217, "learning_rate": 6.688194906239349e-06, "loss": 0.3125, "step": 3064 }, { "epoch": 1.3551394980282314, "grad_norm": 0.3344050599002679, "learning_rate": 6.685773604877556e-06, "loss": 0.3486, "step": 3065 }, { "epoch": 1.3555817639037335, "grad_norm": 0.2742547141683086, "learning_rate": 6.683351857404542e-06, "loss": 0.3583, "step": 3066 }, { "epoch": 1.3560240297792356, "grad_norm": 0.36837575862731015, "learning_rate": 6.680929664461184e-06, "loss": 0.3606, "step": 3067 }, { "epoch": 1.3564662956547378, "grad_norm": 0.3633691155123428, "learning_rate": 6.678507026688475e-06, "loss": 0.371, "step": 3068 }, { "epoch": 1.3569085615302399, "grad_norm": 0.3245765336868303, "learning_rate": 6.67608394472752e-06, "loss": 0.3469, "step": 3069 }, { "epoch": 1.357350827405742, "grad_norm": 0.32736063197952087, "learning_rate": 6.673660419219549e-06, "loss": 0.3105, "step": 3070 }, { "epoch": 1.3577930932812443, "grad_norm": 0.2835771677351122, "learning_rate": 6.671236450805907e-06, "loss": 0.3496, "step": 3071 }, { "epoch": 1.3582353591567464, "grad_norm": 0.2881734430472662, "learning_rate": 6.668812040128056e-06, "loss": 0.3065, "step": 3072 }, { "epoch": 1.3586776250322485, "grad_norm": 0.3009018063627328, "learning_rate": 6.666387187827575e-06, "loss": 0.3268, "step": 3073 }, { "epoch": 1.3591198909077506, "grad_norm": 0.2969472290079342, "learning_rate": 6.6639618945461585e-06, "loss": 0.3216, "step": 3074 }, { "epoch": 1.359562156783253, "grad_norm": 0.28371436035547404, "learning_rate": 6.66153616092562e-06, "loss": 0.3234, "step": 3075 }, { "epoch": 1.360004422658755, "grad_norm": 0.23686085712928717, "learning_rate": 6.65910998760789e-06, "loss": 0.3035, "step": 3076 }, { "epoch": 1.3604466885342572, "grad_norm": 0.3592438343963219, "learning_rate": 6.656683375235009e-06, "loss": 0.3563, "step": 3077 }, { "epoch": 1.3608889544097593, "grad_norm": 0.4652088163002835, "learning_rate": 6.654256324449145e-06, "loss": 0.3485, "step": 3078 }, { "epoch": 1.3613312202852614, "grad_norm": 0.30065937003187726, "learning_rate": 6.6518288358925695e-06, "loss": 0.3321, "step": 3079 }, { "epoch": 1.3617734861607635, "grad_norm": 0.2902075236274571, "learning_rate": 6.649400910207679e-06, "loss": 0.347, "step": 3080 }, { "epoch": 1.3622157520362659, "grad_norm": 0.32432095876704925, "learning_rate": 6.646972548036985e-06, "loss": 0.3191, "step": 3081 }, { "epoch": 1.362658017911768, "grad_norm": 0.28777337649387763, "learning_rate": 6.6445437500231055e-06, "loss": 0.2981, "step": 3082 }, { "epoch": 1.36310028378727, "grad_norm": 0.32418918814822517, "learning_rate": 6.642114516808788e-06, "loss": 0.3759, "step": 3083 }, { "epoch": 1.3635425496627722, "grad_norm": 0.3603845648644585, "learning_rate": 6.639684849036883e-06, "loss": 0.3138, "step": 3084 }, { "epoch": 1.3639848155382746, "grad_norm": 0.37307994757866486, "learning_rate": 6.63725474735036e-06, "loss": 0.3518, "step": 3085 }, { "epoch": 1.3644270814137767, "grad_norm": 0.32284407497378065, "learning_rate": 6.634824212392306e-06, "loss": 0.336, "step": 3086 }, { "epoch": 1.3648693472892788, "grad_norm": 0.27559524132810814, "learning_rate": 6.632393244805921e-06, "loss": 0.3156, "step": 3087 }, { "epoch": 1.365311613164781, "grad_norm": 0.3284014596225614, "learning_rate": 6.629961845234519e-06, "loss": 0.3199, "step": 3088 }, { "epoch": 1.365753879040283, "grad_norm": 0.36412533793132634, "learning_rate": 6.627530014321527e-06, "loss": 0.3533, "step": 3089 }, { "epoch": 1.3661961449157851, "grad_norm": 0.3295974720461419, "learning_rate": 6.625097752710488e-06, "loss": 0.3461, "step": 3090 }, { "epoch": 1.3666384107912872, "grad_norm": 0.4239404764455234, "learning_rate": 6.622665061045058e-06, "loss": 0.3473, "step": 3091 }, { "epoch": 1.3670806766667896, "grad_norm": 0.857630484926246, "learning_rate": 6.62023193996901e-06, "loss": 0.3303, "step": 3092 }, { "epoch": 1.3675229425422917, "grad_norm": 0.35350995539842883, "learning_rate": 6.617798390126225e-06, "loss": 0.3089, "step": 3093 }, { "epoch": 1.3679652084177938, "grad_norm": 0.3981901658839397, "learning_rate": 6.615364412160701e-06, "loss": 0.3203, "step": 3094 }, { "epoch": 1.3684074742932961, "grad_norm": 0.3042460463541901, "learning_rate": 6.612930006716549e-06, "loss": 0.2946, "step": 3095 }, { "epoch": 1.3688497401687982, "grad_norm": 0.3322651730696969, "learning_rate": 6.610495174437994e-06, "loss": 0.3449, "step": 3096 }, { "epoch": 1.3692920060443003, "grad_norm": 0.3282798330590284, "learning_rate": 6.6080599159693715e-06, "loss": 0.3347, "step": 3097 }, { "epoch": 1.3697342719198025, "grad_norm": 0.2722280083401233, "learning_rate": 6.6056242319551315e-06, "loss": 0.3798, "step": 3098 }, { "epoch": 1.3701765377953046, "grad_norm": 0.3896500341123285, "learning_rate": 6.603188123039838e-06, "loss": 0.3282, "step": 3099 }, { "epoch": 1.3706188036708067, "grad_norm": 0.30648751573108235, "learning_rate": 6.600751589868162e-06, "loss": 0.32, "step": 3100 }, { "epoch": 1.3710610695463088, "grad_norm": 0.31355213887532374, "learning_rate": 6.598314633084891e-06, "loss": 0.3148, "step": 3101 }, { "epoch": 1.3715033354218111, "grad_norm": 0.4748883935997744, "learning_rate": 6.595877253334928e-06, "loss": 0.3941, "step": 3102 }, { "epoch": 1.3719456012973132, "grad_norm": 0.3029525574628043, "learning_rate": 6.593439451263279e-06, "loss": 0.3249, "step": 3103 }, { "epoch": 1.3723878671728154, "grad_norm": 0.33363803578169066, "learning_rate": 6.591001227515072e-06, "loss": 0.3564, "step": 3104 }, { "epoch": 1.3728301330483175, "grad_norm": 0.3752365063701196, "learning_rate": 6.588562582735536e-06, "loss": 0.3255, "step": 3105 }, { "epoch": 1.3732723989238198, "grad_norm": 0.3087841625054505, "learning_rate": 6.586123517570019e-06, "loss": 0.3271, "step": 3106 }, { "epoch": 1.373714664799322, "grad_norm": 0.3291827987513017, "learning_rate": 6.58368403266398e-06, "loss": 0.3336, "step": 3107 }, { "epoch": 1.374156930674824, "grad_norm": 0.2945730598821484, "learning_rate": 6.5812441286629835e-06, "loss": 0.3286, "step": 3108 }, { "epoch": 1.3745991965503261, "grad_norm": 0.29283931041762834, "learning_rate": 6.57880380621271e-06, "loss": 0.3158, "step": 3109 }, { "epoch": 1.3750414624258283, "grad_norm": 0.2893078303264359, "learning_rate": 6.576363065958949e-06, "loss": 0.3344, "step": 3110 }, { "epoch": 1.3754837283013304, "grad_norm": 0.32252692764878205, "learning_rate": 6.573921908547603e-06, "loss": 0.3523, "step": 3111 }, { "epoch": 1.3759259941768327, "grad_norm": 0.2983256373231143, "learning_rate": 6.571480334624679e-06, "loss": 0.3243, "step": 3112 }, { "epoch": 1.3763682600523348, "grad_norm": 0.3816051052554249, "learning_rate": 6.5690383448363005e-06, "loss": 0.3178, "step": 3113 }, { "epoch": 1.376810525927837, "grad_norm": 0.3343256364094919, "learning_rate": 6.566595939828695e-06, "loss": 0.3292, "step": 3114 }, { "epoch": 1.377252791803339, "grad_norm": 0.3118672035451466, "learning_rate": 6.5641531202482086e-06, "loss": 0.3314, "step": 3115 }, { "epoch": 1.3776950576788414, "grad_norm": 0.3416205841090434, "learning_rate": 6.561709886741288e-06, "loss": 0.3623, "step": 3116 }, { "epoch": 1.3781373235543435, "grad_norm": 0.3267447138263183, "learning_rate": 6.5592662399544935e-06, "loss": 0.3678, "step": 3117 }, { "epoch": 1.3785795894298456, "grad_norm": 0.32527044490847473, "learning_rate": 6.5568221805344965e-06, "loss": 0.3555, "step": 3118 }, { "epoch": 1.3790218553053477, "grad_norm": 0.3544218538827545, "learning_rate": 6.5543777091280746e-06, "loss": 0.3368, "step": 3119 }, { "epoch": 1.3794641211808498, "grad_norm": 0.30196901170462703, "learning_rate": 6.5519328263821145e-06, "loss": 0.3436, "step": 3120 }, { "epoch": 1.379906387056352, "grad_norm": 0.23850508470761128, "learning_rate": 6.549487532943614e-06, "loss": 0.3012, "step": 3121 }, { "epoch": 1.3803486529318543, "grad_norm": 0.33748587833797544, "learning_rate": 6.547041829459676e-06, "loss": 0.3328, "step": 3122 }, { "epoch": 1.3807909188073564, "grad_norm": 0.31238158861359433, "learning_rate": 6.544595716577519e-06, "loss": 0.3338, "step": 3123 }, { "epoch": 1.3812331846828585, "grad_norm": 0.3379863154521874, "learning_rate": 6.54214919494446e-06, "loss": 0.3433, "step": 3124 }, { "epoch": 1.3816754505583606, "grad_norm": 0.27212814505030786, "learning_rate": 6.53970226520793e-06, "loss": 0.296, "step": 3125 }, { "epoch": 1.382117716433863, "grad_norm": 0.31978732051021785, "learning_rate": 6.53725492801547e-06, "loss": 0.3235, "step": 3126 }, { "epoch": 1.382559982309365, "grad_norm": 0.3437228751706653, "learning_rate": 6.534807184014724e-06, "loss": 0.331, "step": 3127 }, { "epoch": 1.3830022481848672, "grad_norm": 0.3698599795743542, "learning_rate": 6.532359033853447e-06, "loss": 0.3391, "step": 3128 }, { "epoch": 1.3834445140603693, "grad_norm": 0.2969180393764876, "learning_rate": 6.529910478179499e-06, "loss": 0.3306, "step": 3129 }, { "epoch": 1.3838867799358714, "grad_norm": 0.2824922731858298, "learning_rate": 6.527461517640847e-06, "loss": 0.3489, "step": 3130 }, { "epoch": 1.3843290458113735, "grad_norm": 0.3115694548116026, "learning_rate": 6.525012152885571e-06, "loss": 0.3154, "step": 3131 }, { "epoch": 1.3847713116868758, "grad_norm": 0.29016838587910015, "learning_rate": 6.522562384561849e-06, "loss": 0.3557, "step": 3132 }, { "epoch": 1.385213577562378, "grad_norm": 0.323543241369575, "learning_rate": 6.520112213317972e-06, "loss": 0.3355, "step": 3133 }, { "epoch": 1.38565584343788, "grad_norm": 0.33945744974091024, "learning_rate": 6.517661639802337e-06, "loss": 0.3219, "step": 3134 }, { "epoch": 1.3860981093133822, "grad_norm": 0.3287478247580169, "learning_rate": 6.515210664663446e-06, "loss": 0.3206, "step": 3135 }, { "epoch": 1.3865403751888845, "grad_norm": 0.5096526003964499, "learning_rate": 6.512759288549908e-06, "loss": 0.3704, "step": 3136 }, { "epoch": 1.3869826410643866, "grad_norm": 0.2986298909771948, "learning_rate": 6.5103075121104365e-06, "loss": 0.3258, "step": 3137 }, { "epoch": 1.3874249069398887, "grad_norm": 0.37614068202435735, "learning_rate": 6.507855335993854e-06, "loss": 0.3331, "step": 3138 }, { "epoch": 1.3878671728153908, "grad_norm": 0.3224589976380934, "learning_rate": 6.505402760849086e-06, "loss": 0.2976, "step": 3139 }, { "epoch": 1.388309438690893, "grad_norm": 0.2966714155101784, "learning_rate": 6.5029497873251636e-06, "loss": 0.2969, "step": 3140 }, { "epoch": 1.388751704566395, "grad_norm": 0.3543882770643326, "learning_rate": 6.5004964160712266e-06, "loss": 0.3354, "step": 3141 }, { "epoch": 1.3891939704418972, "grad_norm": 0.2788539997646968, "learning_rate": 6.498042647736518e-06, "loss": 0.2887, "step": 3142 }, { "epoch": 1.3896362363173995, "grad_norm": 0.27627052766101867, "learning_rate": 6.4955884829703834e-06, "loss": 0.3692, "step": 3143 }, { "epoch": 1.3900785021929016, "grad_norm": 0.2823200290994414, "learning_rate": 6.493133922422278e-06, "loss": 0.328, "step": 3144 }, { "epoch": 1.3905207680684037, "grad_norm": 0.3249483123940929, "learning_rate": 6.490678966741757e-06, "loss": 0.3232, "step": 3145 }, { "epoch": 1.390963033943906, "grad_norm": 0.33060446648982544, "learning_rate": 6.4882236165784855e-06, "loss": 0.3243, "step": 3146 }, { "epoch": 1.3914052998194082, "grad_norm": 0.29461398854371995, "learning_rate": 6.485767872582227e-06, "loss": 0.3467, "step": 3147 }, { "epoch": 1.3918475656949103, "grad_norm": 0.31017190645764575, "learning_rate": 6.483311735402853e-06, "loss": 0.3287, "step": 3148 }, { "epoch": 1.3922898315704124, "grad_norm": 0.3254663255560477, "learning_rate": 6.48085520569034e-06, "loss": 0.3378, "step": 3149 }, { "epoch": 1.3927320974459145, "grad_norm": 0.3680259351613457, "learning_rate": 6.478398284094766e-06, "loss": 0.3158, "step": 3150 }, { "epoch": 1.3931743633214166, "grad_norm": 0.31443516764552637, "learning_rate": 6.47594097126631e-06, "loss": 0.2984, "step": 3151 }, { "epoch": 1.3936166291969188, "grad_norm": 0.30151725695475073, "learning_rate": 6.473483267855264e-06, "loss": 0.3722, "step": 3152 }, { "epoch": 1.394058895072421, "grad_norm": 0.3293925995852447, "learning_rate": 6.471025174512012e-06, "loss": 0.2855, "step": 3153 }, { "epoch": 1.3945011609479232, "grad_norm": 0.441682506495598, "learning_rate": 6.468566691887051e-06, "loss": 0.3105, "step": 3154 }, { "epoch": 1.3949434268234253, "grad_norm": 0.34530129833401635, "learning_rate": 6.466107820630972e-06, "loss": 0.3327, "step": 3155 }, { "epoch": 1.3953856926989274, "grad_norm": 0.3524842332630875, "learning_rate": 6.463648561394476e-06, "loss": 0.3255, "step": 3156 }, { "epoch": 1.3958279585744298, "grad_norm": 0.3106203781941974, "learning_rate": 6.461188914828362e-06, "loss": 0.3331, "step": 3157 }, { "epoch": 1.3962702244499319, "grad_norm": 0.28539864462689074, "learning_rate": 6.458728881583537e-06, "loss": 0.29, "step": 3158 }, { "epoch": 1.396712490325434, "grad_norm": 0.31121268560436793, "learning_rate": 6.456268462311003e-06, "loss": 0.3355, "step": 3159 }, { "epoch": 1.397154756200936, "grad_norm": 0.31872009713920235, "learning_rate": 6.45380765766187e-06, "loss": 0.3525, "step": 3160 }, { "epoch": 1.3975970220764382, "grad_norm": 0.4664483013083476, "learning_rate": 6.451346468287348e-06, "loss": 0.3028, "step": 3161 }, { "epoch": 1.3980392879519403, "grad_norm": 0.32363482315496084, "learning_rate": 6.448884894838748e-06, "loss": 0.3254, "step": 3162 }, { "epoch": 1.3984815538274427, "grad_norm": 0.3281838808024036, "learning_rate": 6.446422937967485e-06, "loss": 0.3396, "step": 3163 }, { "epoch": 1.3989238197029448, "grad_norm": 0.27271772178935166, "learning_rate": 6.443960598325071e-06, "loss": 0.3414, "step": 3164 }, { "epoch": 1.3993660855784469, "grad_norm": 0.3359761738409486, "learning_rate": 6.441497876563124e-06, "loss": 0.3476, "step": 3165 }, { "epoch": 1.399808351453949, "grad_norm": 0.3027907682954855, "learning_rate": 6.439034773333363e-06, "loss": 0.3459, "step": 3166 }, { "epoch": 1.4002506173294513, "grad_norm": 0.33948848207114624, "learning_rate": 6.436571289287602e-06, "loss": 0.3607, "step": 3167 }, { "epoch": 1.4006928832049534, "grad_norm": 0.31985559825969323, "learning_rate": 6.434107425077763e-06, "loss": 0.3542, "step": 3168 }, { "epoch": 1.4011351490804556, "grad_norm": 0.305283133224083, "learning_rate": 6.431643181355866e-06, "loss": 0.307, "step": 3169 }, { "epoch": 1.4015774149559577, "grad_norm": 0.2725774924141863, "learning_rate": 6.429178558774029e-06, "loss": 0.3189, "step": 3170 }, { "epoch": 1.4020196808314598, "grad_norm": 0.2895226908104209, "learning_rate": 6.426713557984473e-06, "loss": 0.3473, "step": 3171 }, { "epoch": 1.402461946706962, "grad_norm": 0.2885496936477936, "learning_rate": 6.424248179639518e-06, "loss": 0.31, "step": 3172 }, { "epoch": 1.4029042125824642, "grad_norm": 0.29223408890625957, "learning_rate": 6.421782424391587e-06, "loss": 0.3652, "step": 3173 }, { "epoch": 1.4033464784579663, "grad_norm": 0.3132274257525985, "learning_rate": 6.419316292893199e-06, "loss": 0.3152, "step": 3174 }, { "epoch": 1.4037887443334685, "grad_norm": 0.29548210729227964, "learning_rate": 6.4168497857969705e-06, "loss": 0.3513, "step": 3175 }, { "epoch": 1.4042310102089706, "grad_norm": 0.2817966477821676, "learning_rate": 6.414382903755622e-06, "loss": 0.3441, "step": 3176 }, { "epoch": 1.404673276084473, "grad_norm": 0.280502896274072, "learning_rate": 6.411915647421976e-06, "loss": 0.329, "step": 3177 }, { "epoch": 1.405115541959975, "grad_norm": 0.3141085651257641, "learning_rate": 6.4094480174489436e-06, "loss": 0.3219, "step": 3178 }, { "epoch": 1.4055578078354771, "grad_norm": 0.267593715559483, "learning_rate": 6.406980014489545e-06, "loss": 0.3226, "step": 3179 }, { "epoch": 1.4060000737109792, "grad_norm": 0.3652063118944886, "learning_rate": 6.404511639196891e-06, "loss": 0.339, "step": 3180 }, { "epoch": 1.4064423395864813, "grad_norm": 0.3089972155388445, "learning_rate": 6.4020428922242e-06, "loss": 0.3225, "step": 3181 }, { "epoch": 1.4068846054619835, "grad_norm": 0.2969302310452808, "learning_rate": 6.399573774224782e-06, "loss": 0.3335, "step": 3182 }, { "epoch": 1.4073268713374858, "grad_norm": 0.2949183741414482, "learning_rate": 6.397104285852045e-06, "loss": 0.3253, "step": 3183 }, { "epoch": 1.407769137212988, "grad_norm": 0.38997988021411706, "learning_rate": 6.394634427759498e-06, "loss": 0.3171, "step": 3184 }, { "epoch": 1.40821140308849, "grad_norm": 0.3288604356982523, "learning_rate": 6.392164200600749e-06, "loss": 0.3278, "step": 3185 }, { "epoch": 1.4086536689639921, "grad_norm": 0.3585222476898772, "learning_rate": 6.389693605029499e-06, "loss": 0.3631, "step": 3186 }, { "epoch": 1.4090959348394945, "grad_norm": 0.3130389820233732, "learning_rate": 6.387222641699548e-06, "loss": 0.3239, "step": 3187 }, { "epoch": 1.4095382007149966, "grad_norm": 0.3376486479460788, "learning_rate": 6.384751311264797e-06, "loss": 0.3406, "step": 3188 }, { "epoch": 1.4099804665904987, "grad_norm": 0.31171217384174843, "learning_rate": 6.382279614379242e-06, "loss": 0.3411, "step": 3189 }, { "epoch": 1.4104227324660008, "grad_norm": 0.3049554678233742, "learning_rate": 6.379807551696973e-06, "loss": 0.3208, "step": 3190 }, { "epoch": 1.410864998341503, "grad_norm": 0.40752054126425963, "learning_rate": 6.377335123872177e-06, "loss": 0.3246, "step": 3191 }, { "epoch": 1.411307264217005, "grad_norm": 0.26053458398070706, "learning_rate": 6.374862331559146e-06, "loss": 0.3075, "step": 3192 }, { "epoch": 1.4117495300925071, "grad_norm": 0.3466813748360357, "learning_rate": 6.372389175412258e-06, "loss": 0.351, "step": 3193 }, { "epoch": 1.4121917959680095, "grad_norm": 0.3457876715640299, "learning_rate": 6.369915656085992e-06, "loss": 0.3124, "step": 3194 }, { "epoch": 1.4126340618435116, "grad_norm": 0.3180882873767701, "learning_rate": 6.367441774234924e-06, "loss": 0.3315, "step": 3195 }, { "epoch": 1.4130763277190137, "grad_norm": 0.2834072398940798, "learning_rate": 6.364967530513723e-06, "loss": 0.3037, "step": 3196 }, { "epoch": 1.413518593594516, "grad_norm": 0.33776366630407334, "learning_rate": 6.362492925577158e-06, "loss": 0.3156, "step": 3197 }, { "epoch": 1.4139608594700181, "grad_norm": 0.28863105128933614, "learning_rate": 6.360017960080088e-06, "loss": 0.3414, "step": 3198 }, { "epoch": 1.4144031253455203, "grad_norm": 0.31709804352197063, "learning_rate": 6.357542634677471e-06, "loss": 0.3147, "step": 3199 }, { "epoch": 1.4148453912210224, "grad_norm": 0.34926784547763345, "learning_rate": 6.3550669500243625e-06, "loss": 0.3312, "step": 3200 }, { "epoch": 1.4152876570965245, "grad_norm": 0.3187843595051889, "learning_rate": 6.3525909067759065e-06, "loss": 0.3111, "step": 3201 }, { "epoch": 1.4157299229720266, "grad_norm": 0.34897535969773447, "learning_rate": 6.350114505587347e-06, "loss": 0.3274, "step": 3202 }, { "epoch": 1.4161721888475287, "grad_norm": 0.26179138994099266, "learning_rate": 6.3476377471140215e-06, "loss": 0.3177, "step": 3203 }, { "epoch": 1.416614454723031, "grad_norm": 0.33998026546808324, "learning_rate": 6.345160632011362e-06, "loss": 0.2912, "step": 3204 }, { "epoch": 1.4170567205985332, "grad_norm": 0.38503718968240125, "learning_rate": 6.342683160934895e-06, "loss": 0.3468, "step": 3205 }, { "epoch": 1.4174989864740353, "grad_norm": 0.3242451195675309, "learning_rate": 6.34020533454024e-06, "loss": 0.3531, "step": 3206 }, { "epoch": 1.4179412523495374, "grad_norm": 0.3122517161020102, "learning_rate": 6.337727153483111e-06, "loss": 0.3289, "step": 3207 }, { "epoch": 1.4183835182250397, "grad_norm": 0.3460713966889776, "learning_rate": 6.3352486184193185e-06, "loss": 0.3197, "step": 3208 }, { "epoch": 1.4188257841005418, "grad_norm": 0.30594871860255907, "learning_rate": 6.332769730004763e-06, "loss": 0.3129, "step": 3209 }, { "epoch": 1.419268049976044, "grad_norm": 0.3121266952652128, "learning_rate": 6.3302904888954404e-06, "loss": 0.3358, "step": 3210 }, { "epoch": 1.419710315851546, "grad_norm": 0.3533542324423118, "learning_rate": 6.32781089574744e-06, "loss": 0.3658, "step": 3211 }, { "epoch": 1.4201525817270482, "grad_norm": 0.31685501645978187, "learning_rate": 6.325330951216942e-06, "loss": 0.3298, "step": 3212 }, { "epoch": 1.4205948476025503, "grad_norm": 0.3251746952306001, "learning_rate": 6.322850655960224e-06, "loss": 0.3616, "step": 3213 }, { "epoch": 1.4210371134780526, "grad_norm": 0.31727266659883274, "learning_rate": 6.320370010633652e-06, "loss": 0.32, "step": 3214 }, { "epoch": 1.4214793793535547, "grad_norm": 0.297094253514795, "learning_rate": 6.317889015893686e-06, "loss": 0.3104, "step": 3215 }, { "epoch": 1.4219216452290568, "grad_norm": 0.3322980314151764, "learning_rate": 6.31540767239688e-06, "loss": 0.3679, "step": 3216 }, { "epoch": 1.422363911104559, "grad_norm": 0.32800086577506476, "learning_rate": 6.31292598079988e-06, "loss": 0.3339, "step": 3217 }, { "epoch": 1.4228061769800613, "grad_norm": 0.28798541692793456, "learning_rate": 6.3104439417594224e-06, "loss": 0.3079, "step": 3218 }, { "epoch": 1.4232484428555634, "grad_norm": 0.3324220618713851, "learning_rate": 6.307961555932336e-06, "loss": 0.3324, "step": 3219 }, { "epoch": 1.4236907087310655, "grad_norm": 0.30487585156063146, "learning_rate": 6.305478823975543e-06, "loss": 0.3078, "step": 3220 }, { "epoch": 1.4241329746065676, "grad_norm": 0.3902831599883585, "learning_rate": 6.302995746546054e-06, "loss": 0.3463, "step": 3221 }, { "epoch": 1.4245752404820697, "grad_norm": 0.2890154529334093, "learning_rate": 6.300512324300975e-06, "loss": 0.301, "step": 3222 }, { "epoch": 1.4250175063575719, "grad_norm": 0.4573349313885337, "learning_rate": 6.298028557897501e-06, "loss": 0.358, "step": 3223 }, { "epoch": 1.4254597722330742, "grad_norm": 0.2883911108036119, "learning_rate": 6.295544447992918e-06, "loss": 0.3091, "step": 3224 }, { "epoch": 1.4259020381085763, "grad_norm": 0.2639277746743767, "learning_rate": 6.293059995244605e-06, "loss": 0.3643, "step": 3225 }, { "epoch": 1.4263443039840784, "grad_norm": 0.3188384724501906, "learning_rate": 6.290575200310027e-06, "loss": 0.3091, "step": 3226 }, { "epoch": 1.4267865698595805, "grad_norm": 0.4259112107679432, "learning_rate": 6.288090063846745e-06, "loss": 0.3104, "step": 3227 }, { "epoch": 1.4272288357350829, "grad_norm": 0.3216340692683756, "learning_rate": 6.285604586512406e-06, "loss": 0.3133, "step": 3228 }, { "epoch": 1.427671101610585, "grad_norm": 0.33837520327232595, "learning_rate": 6.283118768964751e-06, "loss": 0.3497, "step": 3229 }, { "epoch": 1.428113367486087, "grad_norm": 0.29243103378529506, "learning_rate": 6.2806326118616064e-06, "loss": 0.3104, "step": 3230 }, { "epoch": 1.4285556333615892, "grad_norm": 0.3008800318092811, "learning_rate": 6.278146115860896e-06, "loss": 0.326, "step": 3231 }, { "epoch": 1.4289978992370913, "grad_norm": 0.38715760682726935, "learning_rate": 6.275659281620625e-06, "loss": 0.3709, "step": 3232 }, { "epoch": 1.4294401651125934, "grad_norm": 0.34738024419157715, "learning_rate": 6.273172109798892e-06, "loss": 0.3231, "step": 3233 }, { "epoch": 1.4298824309880958, "grad_norm": 0.3421916524409462, "learning_rate": 6.270684601053884e-06, "loss": 0.3461, "step": 3234 }, { "epoch": 1.4303246968635979, "grad_norm": 0.37480775160994384, "learning_rate": 6.268196756043881e-06, "loss": 0.3735, "step": 3235 }, { "epoch": 1.4307669627391, "grad_norm": 0.3752857481180012, "learning_rate": 6.265708575427246e-06, "loss": 0.3306, "step": 3236 }, { "epoch": 1.431209228614602, "grad_norm": 0.34384245983351286, "learning_rate": 6.263220059862434e-06, "loss": 0.311, "step": 3237 }, { "epoch": 1.4316514944901044, "grad_norm": 0.3207093798678472, "learning_rate": 6.260731210007986e-06, "loss": 0.3003, "step": 3238 }, { "epoch": 1.4320937603656065, "grad_norm": 0.2741813254322843, "learning_rate": 6.258242026522538e-06, "loss": 0.3149, "step": 3239 }, { "epoch": 1.4325360262411087, "grad_norm": 0.3378919709016807, "learning_rate": 6.255752510064808e-06, "loss": 0.3287, "step": 3240 }, { "epoch": 1.4329782921166108, "grad_norm": 0.32797855752437105, "learning_rate": 6.2532626612936035e-06, "loss": 0.3489, "step": 3241 }, { "epoch": 1.4334205579921129, "grad_norm": 0.2858356528809759, "learning_rate": 6.2507724808678205e-06, "loss": 0.317, "step": 3242 }, { "epoch": 1.433862823867615, "grad_norm": 0.3070764161726117, "learning_rate": 6.248281969446443e-06, "loss": 0.3141, "step": 3243 }, { "epoch": 1.434305089743117, "grad_norm": 0.3061903608765578, "learning_rate": 6.245791127688543e-06, "loss": 0.3267, "step": 3244 }, { "epoch": 1.4347473556186194, "grad_norm": 0.33419002167827466, "learning_rate": 6.243299956253279e-06, "loss": 0.3396, "step": 3245 }, { "epoch": 1.4351896214941215, "grad_norm": 0.4479001933935145, "learning_rate": 6.240808455799896e-06, "loss": 0.3401, "step": 3246 }, { "epoch": 1.4356318873696237, "grad_norm": 0.3301418975785273, "learning_rate": 6.2383166269877285e-06, "loss": 0.3164, "step": 3247 }, { "epoch": 1.436074153245126, "grad_norm": 0.2987402098160475, "learning_rate": 6.235824470476196e-06, "loss": 0.3062, "step": 3248 }, { "epoch": 1.436516419120628, "grad_norm": 0.2536283410558376, "learning_rate": 6.233331986924803e-06, "loss": 0.3268, "step": 3249 }, { "epoch": 1.4369586849961302, "grad_norm": 0.3317957571339087, "learning_rate": 6.2308391769931465e-06, "loss": 0.3448, "step": 3250 }, { "epoch": 1.4374009508716323, "grad_norm": 0.3032063111303965, "learning_rate": 6.2283460413409034e-06, "loss": 0.329, "step": 3251 }, { "epoch": 1.4378432167471344, "grad_norm": 0.31489305227111125, "learning_rate": 6.2258525806278415e-06, "loss": 0.3147, "step": 3252 }, { "epoch": 1.4382854826226366, "grad_norm": 0.34025163980427026, "learning_rate": 6.223358795513812e-06, "loss": 0.3586, "step": 3253 }, { "epoch": 1.4387277484981387, "grad_norm": 0.3208729811732599, "learning_rate": 6.220864686658749e-06, "loss": 0.3414, "step": 3254 }, { "epoch": 1.439170014373641, "grad_norm": 0.29373155236994003, "learning_rate": 6.218370254722682e-06, "loss": 0.3615, "step": 3255 }, { "epoch": 1.4396122802491431, "grad_norm": 0.26990507459499663, "learning_rate": 6.215875500365718e-06, "loss": 0.3077, "step": 3256 }, { "epoch": 1.4400545461246452, "grad_norm": 0.3384749472113197, "learning_rate": 6.213380424248047e-06, "loss": 0.3141, "step": 3257 }, { "epoch": 1.4404968120001473, "grad_norm": 0.2490250957625159, "learning_rate": 6.210885027029953e-06, "loss": 0.3056, "step": 3258 }, { "epoch": 1.4409390778756497, "grad_norm": 0.27337747943740387, "learning_rate": 6.208389309371797e-06, "loss": 0.3166, "step": 3259 }, { "epoch": 1.4413813437511518, "grad_norm": 0.32797225883078684, "learning_rate": 6.205893271934032e-06, "loss": 0.2934, "step": 3260 }, { "epoch": 1.441823609626654, "grad_norm": 0.3132905816229255, "learning_rate": 6.2033969153771885e-06, "loss": 0.3548, "step": 3261 }, { "epoch": 1.442265875502156, "grad_norm": 0.34138841147257826, "learning_rate": 6.200900240361885e-06, "loss": 0.3603, "step": 3262 }, { "epoch": 1.4427081413776581, "grad_norm": 0.7647797046828707, "learning_rate": 6.198403247548824e-06, "loss": 0.3689, "step": 3263 }, { "epoch": 1.4431504072531602, "grad_norm": 0.41722269088701003, "learning_rate": 6.195905937598795e-06, "loss": 0.3224, "step": 3264 }, { "epoch": 1.4435926731286626, "grad_norm": 0.2738499223975412, "learning_rate": 6.193408311172664e-06, "loss": 0.3401, "step": 3265 }, { "epoch": 1.4440349390041647, "grad_norm": 0.31088726433536684, "learning_rate": 6.190910368931389e-06, "loss": 0.3315, "step": 3266 }, { "epoch": 1.4444772048796668, "grad_norm": 0.4210331171492957, "learning_rate": 6.1884121115360045e-06, "loss": 0.3545, "step": 3267 }, { "epoch": 1.444919470755169, "grad_norm": 0.32541983784768014, "learning_rate": 6.185913539647634e-06, "loss": 0.333, "step": 3268 }, { "epoch": 1.4453617366306712, "grad_norm": 0.2919760807438159, "learning_rate": 6.183414653927481e-06, "loss": 0.3398, "step": 3269 }, { "epoch": 1.4458040025061734, "grad_norm": 0.2745180537933913, "learning_rate": 6.180915455036833e-06, "loss": 0.3522, "step": 3270 }, { "epoch": 1.4462462683816755, "grad_norm": 0.33583433823326897, "learning_rate": 6.178415943637062e-06, "loss": 0.3509, "step": 3271 }, { "epoch": 1.4466885342571776, "grad_norm": 0.2950758074554019, "learning_rate": 6.175916120389618e-06, "loss": 0.3767, "step": 3272 }, { "epoch": 1.4471308001326797, "grad_norm": 0.26929106419822546, "learning_rate": 6.1734159859560375e-06, "loss": 0.3028, "step": 3273 }, { "epoch": 1.4475730660081818, "grad_norm": 0.36856689992582436, "learning_rate": 6.170915540997939e-06, "loss": 0.3246, "step": 3274 }, { "epoch": 1.4480153318836841, "grad_norm": 0.2925979713610701, "learning_rate": 6.168414786177024e-06, "loss": 0.3558, "step": 3275 }, { "epoch": 1.4484575977591863, "grad_norm": 0.3188215295807621, "learning_rate": 6.165913722155072e-06, "loss": 0.3473, "step": 3276 }, { "epoch": 1.4488998636346884, "grad_norm": 0.2581593195533637, "learning_rate": 6.163412349593949e-06, "loss": 0.3091, "step": 3277 }, { "epoch": 1.4493421295101905, "grad_norm": 0.30586911131664524, "learning_rate": 6.1609106691555985e-06, "loss": 0.3686, "step": 3278 }, { "epoch": 1.4497843953856928, "grad_norm": 0.30803037089687046, "learning_rate": 6.1584086815020515e-06, "loss": 0.3252, "step": 3279 }, { "epoch": 1.450226661261195, "grad_norm": 0.3429865568244395, "learning_rate": 6.155906387295411e-06, "loss": 0.3068, "step": 3280 }, { "epoch": 1.450668927136697, "grad_norm": 0.2610880276762138, "learning_rate": 6.153403787197869e-06, "loss": 0.3258, "step": 3281 }, { "epoch": 1.4511111930121992, "grad_norm": 0.3718468242480794, "learning_rate": 6.150900881871698e-06, "loss": 0.3346, "step": 3282 }, { "epoch": 1.4515534588877013, "grad_norm": 0.34274212314059216, "learning_rate": 6.148397671979245e-06, "loss": 0.3199, "step": 3283 }, { "epoch": 1.4519957247632034, "grad_norm": 0.33019120902504445, "learning_rate": 6.145894158182945e-06, "loss": 0.3361, "step": 3284 }, { "epoch": 1.4524379906387057, "grad_norm": 0.29620115841808436, "learning_rate": 6.143390341145309e-06, "loss": 0.2819, "step": 3285 }, { "epoch": 1.4528802565142078, "grad_norm": 0.32128600293620146, "learning_rate": 6.140886221528928e-06, "loss": 0.3197, "step": 3286 }, { "epoch": 1.45332252238971, "grad_norm": 0.30740014118229536, "learning_rate": 6.138381799996479e-06, "loss": 0.3798, "step": 3287 }, { "epoch": 1.453764788265212, "grad_norm": 0.28653426368397295, "learning_rate": 6.13587707721071e-06, "loss": 0.324, "step": 3288 }, { "epoch": 1.4542070541407144, "grad_norm": 0.28392383135354227, "learning_rate": 6.133372053834453e-06, "loss": 0.2935, "step": 3289 }, { "epoch": 1.4546493200162165, "grad_norm": 0.3096391971128859, "learning_rate": 6.130866730530623e-06, "loss": 0.2862, "step": 3290 }, { "epoch": 1.4550915858917186, "grad_norm": 0.29979094396704015, "learning_rate": 6.128361107962209e-06, "loss": 0.3211, "step": 3291 }, { "epoch": 1.4555338517672207, "grad_norm": 0.2669843281769505, "learning_rate": 6.125855186792283e-06, "loss": 0.3291, "step": 3292 }, { "epoch": 1.4559761176427228, "grad_norm": 0.3067422185275251, "learning_rate": 6.1233489676839916e-06, "loss": 0.333, "step": 3293 }, { "epoch": 1.456418383518225, "grad_norm": 0.4079691462753365, "learning_rate": 6.1208424513005636e-06, "loss": 0.3625, "step": 3294 }, { "epoch": 1.456860649393727, "grad_norm": 0.29306361528120584, "learning_rate": 6.118335638305309e-06, "loss": 0.2819, "step": 3295 }, { "epoch": 1.4573029152692294, "grad_norm": 0.2962597798194889, "learning_rate": 6.115828529361608e-06, "loss": 0.3368, "step": 3296 }, { "epoch": 1.4577451811447315, "grad_norm": 0.6609724691195966, "learning_rate": 6.113321125132928e-06, "loss": 0.3893, "step": 3297 }, { "epoch": 1.4581874470202336, "grad_norm": 0.3092894917356081, "learning_rate": 6.11081342628281e-06, "loss": 0.3172, "step": 3298 }, { "epoch": 1.458629712895736, "grad_norm": 0.3209668721434708, "learning_rate": 6.108305433474872e-06, "loss": 0.3531, "step": 3299 }, { "epoch": 1.459071978771238, "grad_norm": 0.32535070485848433, "learning_rate": 6.105797147372814e-06, "loss": 0.3396, "step": 3300 }, { "epoch": 1.4595142446467402, "grad_norm": 0.29759054397140944, "learning_rate": 6.103288568640408e-06, "loss": 0.3498, "step": 3301 }, { "epoch": 1.4599565105222423, "grad_norm": 0.3454891629297436, "learning_rate": 6.10077969794151e-06, "loss": 0.3088, "step": 3302 }, { "epoch": 1.4603987763977444, "grad_norm": 0.318698161237219, "learning_rate": 6.098270535940048e-06, "loss": 0.3225, "step": 3303 }, { "epoch": 1.4608410422732465, "grad_norm": 0.40108957712836735, "learning_rate": 6.095761083300028e-06, "loss": 0.3698, "step": 3304 }, { "epoch": 1.4612833081487486, "grad_norm": 0.32844275129770273, "learning_rate": 6.093251340685535e-06, "loss": 0.3613, "step": 3305 }, { "epoch": 1.461725574024251, "grad_norm": 0.295491451352214, "learning_rate": 6.090741308760732e-06, "loss": 0.3084, "step": 3306 }, { "epoch": 1.462167839899753, "grad_norm": 0.2798056202465525, "learning_rate": 6.08823098818985e-06, "loss": 0.3291, "step": 3307 }, { "epoch": 1.4626101057752552, "grad_norm": 0.2975444133137117, "learning_rate": 6.085720379637209e-06, "loss": 0.2915, "step": 3308 }, { "epoch": 1.4630523716507573, "grad_norm": 0.29525516096912596, "learning_rate": 6.083209483767195e-06, "loss": 0.3262, "step": 3309 }, { "epoch": 1.4634946375262596, "grad_norm": 0.2566490462563656, "learning_rate": 6.080698301244275e-06, "loss": 0.2926, "step": 3310 }, { "epoch": 1.4639369034017617, "grad_norm": 0.2684221703482616, "learning_rate": 6.078186832732989e-06, "loss": 0.3052, "step": 3311 }, { "epoch": 1.4643791692772639, "grad_norm": 1.617716879858056, "learning_rate": 6.075675078897955e-06, "loss": 0.3088, "step": 3312 }, { "epoch": 1.464821435152766, "grad_norm": 0.25611085439587294, "learning_rate": 6.0731630404038665e-06, "loss": 0.3242, "step": 3313 }, { "epoch": 1.465263701028268, "grad_norm": 0.31209729627598876, "learning_rate": 6.070650717915493e-06, "loss": 0.3453, "step": 3314 }, { "epoch": 1.4657059669037702, "grad_norm": 0.2850597463389476, "learning_rate": 6.0681381120976745e-06, "loss": 0.3509, "step": 3315 }, { "epoch": 1.4661482327792725, "grad_norm": 0.30556184964991984, "learning_rate": 6.065625223615333e-06, "loss": 0.344, "step": 3316 }, { "epoch": 1.4665904986547746, "grad_norm": 0.2868251420372117, "learning_rate": 6.063112053133458e-06, "loss": 0.3112, "step": 3317 }, { "epoch": 1.4670327645302768, "grad_norm": 0.30626669272776175, "learning_rate": 6.0605986013171185e-06, "loss": 0.3071, "step": 3318 }, { "epoch": 1.4674750304057789, "grad_norm": 0.3101716669418697, "learning_rate": 6.058084868831456e-06, "loss": 0.3456, "step": 3319 }, { "epoch": 1.4679172962812812, "grad_norm": 0.2656809155605547, "learning_rate": 6.055570856341688e-06, "loss": 0.3196, "step": 3320 }, { "epoch": 1.4683595621567833, "grad_norm": 0.3029635679735944, "learning_rate": 6.0530565645131056e-06, "loss": 0.3393, "step": 3321 }, { "epoch": 1.4688018280322854, "grad_norm": 0.2828473997433706, "learning_rate": 6.050541994011073e-06, "loss": 0.3199, "step": 3322 }, { "epoch": 1.4692440939077875, "grad_norm": 0.3557312454565777, "learning_rate": 6.048027145501027e-06, "loss": 0.3443, "step": 3323 }, { "epoch": 1.4696863597832897, "grad_norm": 0.3065820187311666, "learning_rate": 6.045512019648479e-06, "loss": 0.3007, "step": 3324 }, { "epoch": 1.4701286256587918, "grad_norm": 0.29550798687532276, "learning_rate": 6.042996617119016e-06, "loss": 0.3702, "step": 3325 }, { "epoch": 1.470570891534294, "grad_norm": 0.32614635044321294, "learning_rate": 6.040480938578296e-06, "loss": 0.3187, "step": 3326 }, { "epoch": 1.4710131574097962, "grad_norm": 0.3157002529046883, "learning_rate": 6.037964984692049e-06, "loss": 0.3583, "step": 3327 }, { "epoch": 1.4714554232852983, "grad_norm": 0.3949021850212591, "learning_rate": 6.03544875612608e-06, "loss": 0.3158, "step": 3328 }, { "epoch": 1.4718976891608004, "grad_norm": 0.3123474508063412, "learning_rate": 6.032932253546268e-06, "loss": 0.2985, "step": 3329 }, { "epoch": 1.4723399550363028, "grad_norm": 0.32673139855604666, "learning_rate": 6.030415477618561e-06, "loss": 0.3752, "step": 3330 }, { "epoch": 1.4727822209118049, "grad_norm": 0.3972942900145169, "learning_rate": 6.027898429008977e-06, "loss": 0.3518, "step": 3331 }, { "epoch": 1.473224486787307, "grad_norm": 0.2893406003453725, "learning_rate": 6.025381108383616e-06, "loss": 0.3328, "step": 3332 }, { "epoch": 1.473666752662809, "grad_norm": 0.2608209488866932, "learning_rate": 6.0228635164086415e-06, "loss": 0.2881, "step": 3333 }, { "epoch": 1.4741090185383112, "grad_norm": 0.41262147076642, "learning_rate": 6.02034565375029e-06, "loss": 0.3347, "step": 3334 }, { "epoch": 1.4745512844138133, "grad_norm": 0.27780151422826416, "learning_rate": 6.017827521074873e-06, "loss": 0.3048, "step": 3335 }, { "epoch": 1.4749935502893157, "grad_norm": 0.4373508864954902, "learning_rate": 6.0153091190487675e-06, "loss": 0.3481, "step": 3336 }, { "epoch": 1.4754358161648178, "grad_norm": 0.4410410806946462, "learning_rate": 6.0127904483384315e-06, "loss": 0.3303, "step": 3337 }, { "epoch": 1.47587808204032, "grad_norm": 0.2764030101043834, "learning_rate": 6.010271509610386e-06, "loss": 0.3437, "step": 3338 }, { "epoch": 1.476320347915822, "grad_norm": 0.37623736040074224, "learning_rate": 6.0077523035312216e-06, "loss": 0.355, "step": 3339 }, { "epoch": 1.4767626137913243, "grad_norm": 0.3106682934281559, "learning_rate": 6.005232830767607e-06, "loss": 0.2978, "step": 3340 }, { "epoch": 1.4772048796668265, "grad_norm": 0.33871576257749386, "learning_rate": 6.002713091986276e-06, "loss": 0.3715, "step": 3341 }, { "epoch": 1.4776471455423286, "grad_norm": 0.31289876756486146, "learning_rate": 6.000193087854036e-06, "loss": 0.337, "step": 3342 }, { "epoch": 1.4780894114178307, "grad_norm": 0.2987579572300145, "learning_rate": 5.997672819037763e-06, "loss": 0.3471, "step": 3343 }, { "epoch": 1.4785316772933328, "grad_norm": 0.3394324927709865, "learning_rate": 5.9951522862044e-06, "loss": 0.3386, "step": 3344 }, { "epoch": 1.478973943168835, "grad_norm": 0.25926263463287125, "learning_rate": 5.992631490020967e-06, "loss": 0.2997, "step": 3345 }, { "epoch": 1.479416209044337, "grad_norm": 0.26727585229979506, "learning_rate": 5.990110431154549e-06, "loss": 0.3294, "step": 3346 }, { "epoch": 1.4798584749198394, "grad_norm": 0.3655361239103565, "learning_rate": 5.987589110272297e-06, "loss": 0.3238, "step": 3347 }, { "epoch": 1.4803007407953415, "grad_norm": 0.3154256592994628, "learning_rate": 5.9850675280414415e-06, "loss": 0.3296, "step": 3348 }, { "epoch": 1.4807430066708436, "grad_norm": 0.4848524003471403, "learning_rate": 5.982545685129273e-06, "loss": 0.3575, "step": 3349 }, { "epoch": 1.481185272546346, "grad_norm": 0.290949020666619, "learning_rate": 5.980023582203154e-06, "loss": 0.3161, "step": 3350 }, { "epoch": 1.481627538421848, "grad_norm": 0.31947762244682215, "learning_rate": 5.977501219930517e-06, "loss": 0.3289, "step": 3351 }, { "epoch": 1.4820698042973501, "grad_norm": 0.32366358983689963, "learning_rate": 5.974978598978861e-06, "loss": 0.3323, "step": 3352 }, { "epoch": 1.4825120701728522, "grad_norm": 0.302556015954315, "learning_rate": 5.972455720015758e-06, "loss": 0.3379, "step": 3353 }, { "epoch": 1.4829543360483544, "grad_norm": 0.2862413670554005, "learning_rate": 5.969932583708841e-06, "loss": 0.3287, "step": 3354 }, { "epoch": 1.4833966019238565, "grad_norm": 0.2928408661706076, "learning_rate": 5.967409190725814e-06, "loss": 0.3535, "step": 3355 }, { "epoch": 1.4838388677993586, "grad_norm": 0.3605492189270206, "learning_rate": 5.964885541734454e-06, "loss": 0.3317, "step": 3356 }, { "epoch": 1.484281133674861, "grad_norm": 0.33370022833343527, "learning_rate": 5.9623616374026e-06, "loss": 0.3515, "step": 3357 }, { "epoch": 1.484723399550363, "grad_norm": 0.27488914365666145, "learning_rate": 5.9598374783981604e-06, "loss": 0.3216, "step": 3358 }, { "epoch": 1.4851656654258651, "grad_norm": 0.26986781545539074, "learning_rate": 5.95731306538911e-06, "loss": 0.3267, "step": 3359 }, { "epoch": 1.4856079313013673, "grad_norm": 0.2972541052563053, "learning_rate": 5.954788399043491e-06, "loss": 0.3294, "step": 3360 }, { "epoch": 1.4860501971768696, "grad_norm": 0.32200852608431435, "learning_rate": 5.952263480029417e-06, "loss": 0.2913, "step": 3361 }, { "epoch": 1.4864924630523717, "grad_norm": 0.35829323271502905, "learning_rate": 5.949738309015059e-06, "loss": 0.3774, "step": 3362 }, { "epoch": 1.4869347289278738, "grad_norm": 0.29536844263522605, "learning_rate": 5.947212886668664e-06, "loss": 0.3494, "step": 3363 }, { "epoch": 1.487376994803376, "grad_norm": 0.2802693193793304, "learning_rate": 5.944687213658542e-06, "loss": 0.3037, "step": 3364 }, { "epoch": 1.487819260678878, "grad_norm": 0.29307169089156265, "learning_rate": 5.9421612906530705e-06, "loss": 0.3239, "step": 3365 }, { "epoch": 1.4882615265543802, "grad_norm": 0.3376113472451062, "learning_rate": 5.939635118320688e-06, "loss": 0.3475, "step": 3366 }, { "epoch": 1.4887037924298825, "grad_norm": 0.31482687589803365, "learning_rate": 5.937108697329905e-06, "loss": 0.3159, "step": 3367 }, { "epoch": 1.4891460583053846, "grad_norm": 0.32210305079794294, "learning_rate": 5.934582028349295e-06, "loss": 0.3442, "step": 3368 }, { "epoch": 1.4895883241808867, "grad_norm": 0.30876544202879924, "learning_rate": 5.932055112047501e-06, "loss": 0.2906, "step": 3369 }, { "epoch": 1.4900305900563888, "grad_norm": 0.4144717259405973, "learning_rate": 5.929527949093223e-06, "loss": 0.3538, "step": 3370 }, { "epoch": 1.4904728559318912, "grad_norm": 0.33117962897365555, "learning_rate": 5.927000540155237e-06, "loss": 0.293, "step": 3371 }, { "epoch": 1.4909151218073933, "grad_norm": 0.3629763074367731, "learning_rate": 5.924472885902373e-06, "loss": 0.3607, "step": 3372 }, { "epoch": 1.4913573876828954, "grad_norm": 0.3104935765878214, "learning_rate": 5.921944987003537e-06, "loss": 0.3165, "step": 3373 }, { "epoch": 1.4917996535583975, "grad_norm": 0.3500659734808015, "learning_rate": 5.919416844127692e-06, "loss": 0.3638, "step": 3374 }, { "epoch": 1.4922419194338996, "grad_norm": 0.3064492807893673, "learning_rate": 5.916888457943867e-06, "loss": 0.3069, "step": 3375 }, { "epoch": 1.4926841853094017, "grad_norm": 0.3129713038883691, "learning_rate": 5.914359829121156e-06, "loss": 0.2998, "step": 3376 }, { "epoch": 1.493126451184904, "grad_norm": 0.33031821030678027, "learning_rate": 5.9118309583287205e-06, "loss": 0.3418, "step": 3377 }, { "epoch": 1.4935687170604062, "grad_norm": 0.2578304469145479, "learning_rate": 5.909301846235779e-06, "loss": 0.2911, "step": 3378 }, { "epoch": 1.4940109829359083, "grad_norm": 0.34393280297089, "learning_rate": 5.9067724935116214e-06, "loss": 0.3186, "step": 3379 }, { "epoch": 1.4944532488114104, "grad_norm": 0.4118720911515005, "learning_rate": 5.904242900825595e-06, "loss": 0.37, "step": 3380 }, { "epoch": 1.4948955146869127, "grad_norm": 0.34307648070901164, "learning_rate": 5.901713068847116e-06, "loss": 0.3332, "step": 3381 }, { "epoch": 1.4953377805624148, "grad_norm": 0.3159567633779197, "learning_rate": 5.899182998245658e-06, "loss": 0.3229, "step": 3382 }, { "epoch": 1.495780046437917, "grad_norm": 0.278891306081506, "learning_rate": 5.896652689690765e-06, "loss": 0.3451, "step": 3383 }, { "epoch": 1.496222312313419, "grad_norm": 0.3519482008133855, "learning_rate": 5.894122143852037e-06, "loss": 0.3486, "step": 3384 }, { "epoch": 1.4966645781889212, "grad_norm": 0.27607647506422794, "learning_rate": 5.89159136139914e-06, "loss": 0.3316, "step": 3385 }, { "epoch": 1.4971068440644233, "grad_norm": 0.33534965245913745, "learning_rate": 5.8890603430018015e-06, "loss": 0.3388, "step": 3386 }, { "epoch": 1.4975491099399256, "grad_norm": 0.36767841128561196, "learning_rate": 5.886529089329815e-06, "loss": 0.3652, "step": 3387 }, { "epoch": 1.4979913758154277, "grad_norm": 0.2830054649122808, "learning_rate": 5.883997601053033e-06, "loss": 0.3361, "step": 3388 }, { "epoch": 1.4984336416909299, "grad_norm": 0.31249783474369086, "learning_rate": 5.88146587884137e-06, "loss": 0.3482, "step": 3389 }, { "epoch": 1.498875907566432, "grad_norm": 0.24684106630596453, "learning_rate": 5.878933923364804e-06, "loss": 0.2971, "step": 3390 }, { "epoch": 1.4993181734419343, "grad_norm": 0.3591310998101924, "learning_rate": 5.876401735293371e-06, "loss": 0.3415, "step": 3391 }, { "epoch": 1.4997604393174364, "grad_norm": 0.3535604094890618, "learning_rate": 5.873869315297175e-06, "loss": 0.3607, "step": 3392 }, { "epoch": 1.5002027051929385, "grad_norm": 0.36006897626504986, "learning_rate": 5.871336664046375e-06, "loss": 0.3441, "step": 3393 }, { "epoch": 1.5006449710684406, "grad_norm": 0.3647608417073433, "learning_rate": 5.868803782211194e-06, "loss": 0.3697, "step": 3394 }, { "epoch": 1.5010872369439427, "grad_norm": 0.32597641349084056, "learning_rate": 5.866270670461919e-06, "loss": 0.3592, "step": 3395 }, { "epoch": 1.5015295028194449, "grad_norm": 0.2905375944831394, "learning_rate": 5.863737329468892e-06, "loss": 0.318, "step": 3396 }, { "epoch": 1.501971768694947, "grad_norm": 0.27399837096794893, "learning_rate": 5.861203759902519e-06, "loss": 0.3476, "step": 3397 }, { "epoch": 1.5024140345704493, "grad_norm": 0.35397015397063264, "learning_rate": 5.858669962433266e-06, "loss": 0.3289, "step": 3398 }, { "epoch": 1.5028563004459514, "grad_norm": 0.30266661561268154, "learning_rate": 5.856135937731661e-06, "loss": 0.3014, "step": 3399 }, { "epoch": 1.5032985663214535, "grad_norm": 0.2860189994340211, "learning_rate": 5.853601686468288e-06, "loss": 0.3136, "step": 3400 }, { "epoch": 1.5037408321969559, "grad_norm": 0.3915923440286897, "learning_rate": 5.851067209313793e-06, "loss": 0.3452, "step": 3401 }, { "epoch": 1.504183098072458, "grad_norm": 0.2896755918984749, "learning_rate": 5.8485325069388835e-06, "loss": 0.3254, "step": 3402 }, { "epoch": 1.50462536394796, "grad_norm": 0.399463177619155, "learning_rate": 5.845997580014326e-06, "loss": 0.3399, "step": 3403 }, { "epoch": 1.5050676298234622, "grad_norm": 0.2895740428408272, "learning_rate": 5.843462429210945e-06, "loss": 0.3253, "step": 3404 }, { "epoch": 1.5055098956989643, "grad_norm": 0.30451119099175616, "learning_rate": 5.8409270551996235e-06, "loss": 0.3096, "step": 3405 }, { "epoch": 1.5059521615744664, "grad_norm": 0.28551298948466597, "learning_rate": 5.8383914586513065e-06, "loss": 0.288, "step": 3406 }, { "epoch": 1.5063944274499685, "grad_norm": 0.3450340847247368, "learning_rate": 5.835855640236996e-06, "loss": 0.3325, "step": 3407 }, { "epoch": 1.5068366933254709, "grad_norm": 0.35089862068525546, "learning_rate": 5.8333196006277536e-06, "loss": 0.2997, "step": 3408 }, { "epoch": 1.507278959200973, "grad_norm": 0.2581082072223938, "learning_rate": 5.830783340494698e-06, "loss": 0.2987, "step": 3409 }, { "epoch": 1.507721225076475, "grad_norm": 0.3687358131577208, "learning_rate": 5.8282468605090055e-06, "loss": 0.3163, "step": 3410 }, { "epoch": 1.5081634909519774, "grad_norm": 0.323645743146135, "learning_rate": 5.825710161341917e-06, "loss": 0.3657, "step": 3411 }, { "epoch": 1.5086057568274795, "grad_norm": 0.3237662083910932, "learning_rate": 5.8231732436647235e-06, "loss": 0.3621, "step": 3412 }, { "epoch": 1.5090480227029817, "grad_norm": 0.3335367177350376, "learning_rate": 5.820636108148776e-06, "loss": 0.3205, "step": 3413 }, { "epoch": 1.5094902885784838, "grad_norm": 0.38791672470646116, "learning_rate": 5.818098755465487e-06, "loss": 0.3664, "step": 3414 }, { "epoch": 1.5099325544539859, "grad_norm": 0.36702739810355517, "learning_rate": 5.81556118628632e-06, "loss": 0.362, "step": 3415 }, { "epoch": 1.510374820329488, "grad_norm": 0.3043745509880406, "learning_rate": 5.813023401282804e-06, "loss": 0.315, "step": 3416 }, { "epoch": 1.5108170862049901, "grad_norm": 0.297693264277529, "learning_rate": 5.810485401126517e-06, "loss": 0.3187, "step": 3417 }, { "epoch": 1.5112593520804922, "grad_norm": 0.2938546274138852, "learning_rate": 5.807947186489097e-06, "loss": 0.3265, "step": 3418 }, { "epoch": 1.5117016179559946, "grad_norm": 0.3353507762457801, "learning_rate": 5.805408758042242e-06, "loss": 0.3402, "step": 3419 }, { "epoch": 1.5121438838314967, "grad_norm": 0.43964181108615036, "learning_rate": 5.802870116457703e-06, "loss": 0.3259, "step": 3420 }, { "epoch": 1.512586149706999, "grad_norm": 0.3240247649538191, "learning_rate": 5.800331262407285e-06, "loss": 0.3744, "step": 3421 }, { "epoch": 1.5130284155825011, "grad_norm": 0.29064503338537856, "learning_rate": 5.797792196562857e-06, "loss": 0.3392, "step": 3422 }, { "epoch": 1.5134706814580032, "grad_norm": 0.34225163238276246, "learning_rate": 5.795252919596337e-06, "loss": 0.3313, "step": 3423 }, { "epoch": 1.5139129473335053, "grad_norm": 0.35433018146735107, "learning_rate": 5.7927134321797015e-06, "loss": 0.3451, "step": 3424 }, { "epoch": 1.5143552132090075, "grad_norm": 0.31015288895103177, "learning_rate": 5.790173734984983e-06, "loss": 0.3416, "step": 3425 }, { "epoch": 1.5147974790845096, "grad_norm": 0.308882198175267, "learning_rate": 5.787633828684267e-06, "loss": 0.3449, "step": 3426 }, { "epoch": 1.5152397449600117, "grad_norm": 0.39834367411950294, "learning_rate": 5.7850937139496995e-06, "loss": 0.3732, "step": 3427 }, { "epoch": 1.5156820108355138, "grad_norm": 0.32077156646214094, "learning_rate": 5.782553391453479e-06, "loss": 0.3218, "step": 3428 }, { "epoch": 1.5161242767110161, "grad_norm": 0.34191490337634883, "learning_rate": 5.780012861867854e-06, "loss": 0.3422, "step": 3429 }, { "epoch": 1.5165665425865182, "grad_norm": 0.35897295301589444, "learning_rate": 5.7774721258651355e-06, "loss": 0.3654, "step": 3430 }, { "epoch": 1.5170088084620206, "grad_norm": 0.29592299992938315, "learning_rate": 5.774931184117686e-06, "loss": 0.3257, "step": 3431 }, { "epoch": 1.5174510743375227, "grad_norm": 0.2784468351913885, "learning_rate": 5.772390037297921e-06, "loss": 0.3213, "step": 3432 }, { "epoch": 1.5178933402130248, "grad_norm": 0.330412991509124, "learning_rate": 5.769848686078313e-06, "loss": 0.3666, "step": 3433 }, { "epoch": 1.518335606088527, "grad_norm": 0.27393578455003414, "learning_rate": 5.767307131131385e-06, "loss": 0.3293, "step": 3434 }, { "epoch": 1.518777871964029, "grad_norm": 0.30479080923950636, "learning_rate": 5.764765373129721e-06, "loss": 0.3137, "step": 3435 }, { "epoch": 1.5192201378395311, "grad_norm": 0.2769787786545293, "learning_rate": 5.762223412745948e-06, "loss": 0.3583, "step": 3436 }, { "epoch": 1.5196624037150332, "grad_norm": 0.2986590039665341, "learning_rate": 5.759681250652755e-06, "loss": 0.3352, "step": 3437 }, { "epoch": 1.5201046695905354, "grad_norm": 0.33124793752154025, "learning_rate": 5.757138887522884e-06, "loss": 0.3484, "step": 3438 }, { "epoch": 1.5205469354660377, "grad_norm": 0.29044702114479504, "learning_rate": 5.754596324029125e-06, "loss": 0.3723, "step": 3439 }, { "epoch": 1.5209892013415398, "grad_norm": 0.3044337454828915, "learning_rate": 5.752053560844324e-06, "loss": 0.3378, "step": 3440 }, { "epoch": 1.5214314672170421, "grad_norm": 0.33113934030785847, "learning_rate": 5.749510598641382e-06, "loss": 0.304, "step": 3441 }, { "epoch": 1.5218737330925443, "grad_norm": 0.27492398733452156, "learning_rate": 5.746967438093248e-06, "loss": 0.2954, "step": 3442 }, { "epoch": 1.5223159989680464, "grad_norm": 0.3600875325414294, "learning_rate": 5.74442407987293e-06, "loss": 0.3591, "step": 3443 }, { "epoch": 1.5227582648435485, "grad_norm": 0.3017440670259702, "learning_rate": 5.74188052465348e-06, "loss": 0.3177, "step": 3444 }, { "epoch": 1.5232005307190506, "grad_norm": 0.2972088887411886, "learning_rate": 5.7393367731080074e-06, "loss": 0.3436, "step": 3445 }, { "epoch": 1.5236427965945527, "grad_norm": 0.2610659448999932, "learning_rate": 5.736792825909675e-06, "loss": 0.3206, "step": 3446 }, { "epoch": 1.5240850624700548, "grad_norm": 0.32835389334172477, "learning_rate": 5.734248683731694e-06, "loss": 0.3648, "step": 3447 }, { "epoch": 1.524527328345557, "grad_norm": 0.4491387578437623, "learning_rate": 5.731704347247326e-06, "loss": 0.3708, "step": 3448 }, { "epoch": 1.5249695942210593, "grad_norm": 0.35056048765488385, "learning_rate": 5.72915981712989e-06, "loss": 0.3718, "step": 3449 }, { "epoch": 1.5254118600965614, "grad_norm": 0.29994641780537473, "learning_rate": 5.726615094052748e-06, "loss": 0.3535, "step": 3450 }, { "epoch": 1.5258541259720635, "grad_norm": 0.3554227530202575, "learning_rate": 5.724070178689323e-06, "loss": 0.3071, "step": 3451 }, { "epoch": 1.5262963918475658, "grad_norm": 0.28243046651978765, "learning_rate": 5.721525071713077e-06, "loss": 0.3103, "step": 3452 }, { "epoch": 1.526738657723068, "grad_norm": 0.3171102640867678, "learning_rate": 5.718979773797536e-06, "loss": 0.3248, "step": 3453 }, { "epoch": 1.52718092359857, "grad_norm": 0.29725533976254576, "learning_rate": 5.7164342856162645e-06, "loss": 0.3229, "step": 3454 }, { "epoch": 1.5276231894740722, "grad_norm": 0.36506537248582016, "learning_rate": 5.7138886078428845e-06, "loss": 0.359, "step": 3455 }, { "epoch": 1.5280654553495743, "grad_norm": 0.33862352011169616, "learning_rate": 5.711342741151067e-06, "loss": 0.3189, "step": 3456 }, { "epoch": 1.5285077212250764, "grad_norm": 0.3231170212413772, "learning_rate": 5.708796686214529e-06, "loss": 0.326, "step": 3457 }, { "epoch": 1.5289499871005785, "grad_norm": 0.30544815456361263, "learning_rate": 5.706250443707043e-06, "loss": 0.3181, "step": 3458 }, { "epoch": 1.5293922529760808, "grad_norm": 0.257157670853938, "learning_rate": 5.70370401430243e-06, "loss": 0.3335, "step": 3459 }, { "epoch": 1.529834518851583, "grad_norm": 0.412219127699329, "learning_rate": 5.701157398674556e-06, "loss": 0.3467, "step": 3460 }, { "epoch": 1.530276784727085, "grad_norm": 0.35692379716873923, "learning_rate": 5.698610597497341e-06, "loss": 0.3774, "step": 3461 }, { "epoch": 1.5307190506025874, "grad_norm": 0.3180824160439017, "learning_rate": 5.696063611444752e-06, "loss": 0.3162, "step": 3462 }, { "epoch": 1.5311613164780895, "grad_norm": 0.316319089760667, "learning_rate": 5.6935164411908065e-06, "loss": 0.3298, "step": 3463 }, { "epoch": 1.5316035823535916, "grad_norm": 0.3320249917413489, "learning_rate": 5.690969087409568e-06, "loss": 0.3006, "step": 3464 }, { "epoch": 1.5320458482290937, "grad_norm": 0.3001637899309255, "learning_rate": 5.68842155077515e-06, "loss": 0.3102, "step": 3465 }, { "epoch": 1.5324881141045958, "grad_norm": 0.29264703154713445, "learning_rate": 5.685873831961717e-06, "loss": 0.3154, "step": 3466 }, { "epoch": 1.532930379980098, "grad_norm": 0.28353161629765705, "learning_rate": 5.683325931643478e-06, "loss": 0.3128, "step": 3467 }, { "epoch": 1.5333726458556, "grad_norm": 0.3023430143344705, "learning_rate": 5.68077785049469e-06, "loss": 0.3033, "step": 3468 }, { "epoch": 1.5338149117311022, "grad_norm": 0.3770901723722238, "learning_rate": 5.67822958918966e-06, "loss": 0.3045, "step": 3469 }, { "epoch": 1.5342571776066045, "grad_norm": 0.27304848166327655, "learning_rate": 5.675681148402743e-06, "loss": 0.3406, "step": 3470 }, { "epoch": 1.5346994434821066, "grad_norm": 0.2887460535570763, "learning_rate": 5.673132528808341e-06, "loss": 0.3265, "step": 3471 }, { "epoch": 1.535141709357609, "grad_norm": 0.3199234947828661, "learning_rate": 5.6705837310809e-06, "loss": 0.3605, "step": 3472 }, { "epoch": 1.535583975233111, "grad_norm": 0.34130454923061493, "learning_rate": 5.668034755894916e-06, "loss": 0.3456, "step": 3473 }, { "epoch": 1.5360262411086132, "grad_norm": 0.3203821317345977, "learning_rate": 5.665485603924935e-06, "loss": 0.3089, "step": 3474 }, { "epoch": 1.5364685069841153, "grad_norm": 0.3079435345580405, "learning_rate": 5.662936275845544e-06, "loss": 0.3566, "step": 3475 }, { "epoch": 1.5369107728596174, "grad_norm": 0.30556245496465073, "learning_rate": 5.660386772331378e-06, "loss": 0.2938, "step": 3476 }, { "epoch": 1.5373530387351195, "grad_norm": 0.30307833718163896, "learning_rate": 5.657837094057123e-06, "loss": 0.3572, "step": 3477 }, { "epoch": 1.5377953046106216, "grad_norm": 0.3073330118726968, "learning_rate": 5.655287241697506e-06, "loss": 0.3268, "step": 3478 }, { "epoch": 1.5382375704861238, "grad_norm": 0.32674037468161343, "learning_rate": 5.652737215927302e-06, "loss": 0.3008, "step": 3479 }, { "epoch": 1.538679836361626, "grad_norm": 0.3105845412070139, "learning_rate": 5.6501870174213315e-06, "loss": 0.3291, "step": 3480 }, { "epoch": 1.5391221022371282, "grad_norm": 0.3077424948989133, "learning_rate": 5.647636646854461e-06, "loss": 0.3383, "step": 3481 }, { "epoch": 1.5395643681126305, "grad_norm": 0.360094545213666, "learning_rate": 5.645086104901604e-06, "loss": 0.3808, "step": 3482 }, { "epoch": 1.5400066339881326, "grad_norm": 0.3040162071528225, "learning_rate": 5.642535392237716e-06, "loss": 0.3277, "step": 3483 }, { "epoch": 1.5404488998636348, "grad_norm": 0.291879440842041, "learning_rate": 5.639984509537799e-06, "loss": 0.3, "step": 3484 }, { "epoch": 1.5408911657391369, "grad_norm": 0.4103884330279861, "learning_rate": 5.637433457476904e-06, "loss": 0.3589, "step": 3485 }, { "epoch": 1.541333431614639, "grad_norm": 0.3054420793681859, "learning_rate": 5.6348822367301225e-06, "loss": 0.3295, "step": 3486 }, { "epoch": 1.541775697490141, "grad_norm": 0.3332852794003616, "learning_rate": 5.632330847972588e-06, "loss": 0.3656, "step": 3487 }, { "epoch": 1.5422179633656432, "grad_norm": 0.31726901660941753, "learning_rate": 5.6297792918794856e-06, "loss": 0.3444, "step": 3488 }, { "epoch": 1.5426602292411453, "grad_norm": 0.29543726838971274, "learning_rate": 5.627227569126038e-06, "loss": 0.3312, "step": 3489 }, { "epoch": 1.5431024951166477, "grad_norm": 0.3409315465625976, "learning_rate": 5.6246756803875205e-06, "loss": 0.3305, "step": 3490 }, { "epoch": 1.5435447609921498, "grad_norm": 0.2673821678066559, "learning_rate": 5.622123626339241e-06, "loss": 0.2925, "step": 3491 }, { "epoch": 1.543987026867652, "grad_norm": 0.3053457805515216, "learning_rate": 5.619571407656559e-06, "loss": 0.3244, "step": 3492 }, { "epoch": 1.5444292927431542, "grad_norm": 0.41464291883621424, "learning_rate": 5.617019025014877e-06, "loss": 0.3362, "step": 3493 }, { "epoch": 1.5448715586186563, "grad_norm": 0.35579352639683814, "learning_rate": 5.61446647908964e-06, "loss": 0.3127, "step": 3494 }, { "epoch": 1.5453138244941584, "grad_norm": 0.32633487705397235, "learning_rate": 5.611913770556332e-06, "loss": 0.3777, "step": 3495 }, { "epoch": 1.5457560903696606, "grad_norm": 0.310826405291715, "learning_rate": 5.609360900090487e-06, "loss": 0.3218, "step": 3496 }, { "epoch": 1.5461983562451627, "grad_norm": 0.2978168736158961, "learning_rate": 5.606807868367679e-06, "loss": 0.3361, "step": 3497 }, { "epoch": 1.5466406221206648, "grad_norm": 0.3370991344091917, "learning_rate": 5.604254676063521e-06, "loss": 0.3523, "step": 3498 }, { "epoch": 1.5470828879961669, "grad_norm": 0.3656052865995458, "learning_rate": 5.6017013238536744e-06, "loss": 0.3611, "step": 3499 }, { "epoch": 1.5475251538716692, "grad_norm": 0.27590678633423504, "learning_rate": 5.599147812413839e-06, "loss": 0.3371, "step": 3500 }, { "epoch": 1.5479674197471713, "grad_norm": 0.34071584198196464, "learning_rate": 5.596594142419759e-06, "loss": 0.34, "step": 3501 }, { "epoch": 1.5484096856226734, "grad_norm": 0.29159856081211477, "learning_rate": 5.594040314547221e-06, "loss": 0.3226, "step": 3502 }, { "epoch": 1.5488519514981758, "grad_norm": 0.36512080886325887, "learning_rate": 5.591486329472048e-06, "loss": 0.3468, "step": 3503 }, { "epoch": 1.549294217373678, "grad_norm": 0.2671247283669101, "learning_rate": 5.588932187870111e-06, "loss": 0.2773, "step": 3504 }, { "epoch": 1.54973648324918, "grad_norm": 0.3451046532963557, "learning_rate": 5.586377890417321e-06, "loss": 0.3202, "step": 3505 }, { "epoch": 1.5501787491246821, "grad_norm": 0.38610814836191387, "learning_rate": 5.583823437789627e-06, "loss": 0.3271, "step": 3506 }, { "epoch": 1.5506210150001842, "grad_norm": 0.28761130706085586, "learning_rate": 5.581268830663023e-06, "loss": 0.3269, "step": 3507 }, { "epoch": 1.5510632808756863, "grad_norm": 0.3298526464006159, "learning_rate": 5.57871406971354e-06, "loss": 0.3528, "step": 3508 }, { "epoch": 1.5515055467511885, "grad_norm": 0.33876449817875315, "learning_rate": 5.576159155617254e-06, "loss": 0.3517, "step": 3509 }, { "epoch": 1.5519478126266908, "grad_norm": 0.31455417709544853, "learning_rate": 5.573604089050281e-06, "loss": 0.3197, "step": 3510 }, { "epoch": 1.552390078502193, "grad_norm": 0.3290784067620494, "learning_rate": 5.571048870688771e-06, "loss": 0.373, "step": 3511 }, { "epoch": 1.552832344377695, "grad_norm": 0.3773756055113683, "learning_rate": 5.5684935012089225e-06, "loss": 0.3217, "step": 3512 }, { "epoch": 1.5532746102531974, "grad_norm": 0.2714021395955187, "learning_rate": 5.56593798128697e-06, "loss": 0.3048, "step": 3513 }, { "epoch": 1.5537168761286995, "grad_norm": 0.34025693920754585, "learning_rate": 5.563382311599187e-06, "loss": 0.3566, "step": 3514 }, { "epoch": 1.5541591420042016, "grad_norm": 0.3021255675342062, "learning_rate": 5.560826492821889e-06, "loss": 0.3164, "step": 3515 }, { "epoch": 1.5546014078797037, "grad_norm": 0.2825555843343923, "learning_rate": 5.558270525631427e-06, "loss": 0.3044, "step": 3516 }, { "epoch": 1.5550436737552058, "grad_norm": 0.3458329631823241, "learning_rate": 5.555714410704201e-06, "loss": 0.3583, "step": 3517 }, { "epoch": 1.555485939630708, "grad_norm": 0.3609815099972233, "learning_rate": 5.5531581487166355e-06, "loss": 0.3351, "step": 3518 }, { "epoch": 1.55592820550621, "grad_norm": 0.27269513619306196, "learning_rate": 5.550601740345205e-06, "loss": 0.3199, "step": 3519 }, { "epoch": 1.5563704713817124, "grad_norm": 0.2851543363672283, "learning_rate": 5.54804518626642e-06, "loss": 0.3214, "step": 3520 }, { "epoch": 1.5568127372572145, "grad_norm": 0.29449174836110864, "learning_rate": 5.545488487156827e-06, "loss": 0.2872, "step": 3521 }, { "epoch": 1.5572550031327166, "grad_norm": 0.293723248686815, "learning_rate": 5.542931643693016e-06, "loss": 0.3282, "step": 3522 }, { "epoch": 1.557697269008219, "grad_norm": 0.2996148116954619, "learning_rate": 5.540374656551608e-06, "loss": 0.3179, "step": 3523 }, { "epoch": 1.558139534883721, "grad_norm": 0.3247066852667624, "learning_rate": 5.537817526409267e-06, "loss": 0.3376, "step": 3524 }, { "epoch": 1.5585818007592231, "grad_norm": 0.3325599559245042, "learning_rate": 5.5352602539426984e-06, "loss": 0.3597, "step": 3525 }, { "epoch": 1.5590240666347253, "grad_norm": 0.3563094623528098, "learning_rate": 5.532702839828633e-06, "loss": 0.3228, "step": 3526 }, { "epoch": 1.5594663325102274, "grad_norm": 0.33855220862840135, "learning_rate": 5.530145284743853e-06, "loss": 0.3324, "step": 3527 }, { "epoch": 1.5599085983857295, "grad_norm": 0.3120043130954907, "learning_rate": 5.527587589365169e-06, "loss": 0.3226, "step": 3528 }, { "epoch": 1.5603508642612316, "grad_norm": 0.2625951665165704, "learning_rate": 5.525029754369432e-06, "loss": 0.3132, "step": 3529 }, { "epoch": 1.5607931301367337, "grad_norm": 0.31498399624740936, "learning_rate": 5.522471780433531e-06, "loss": 0.3191, "step": 3530 }, { "epoch": 1.561235396012236, "grad_norm": 0.3375694844162439, "learning_rate": 5.519913668234387e-06, "loss": 0.3496, "step": 3531 }, { "epoch": 1.5616776618877382, "grad_norm": 0.291329686619264, "learning_rate": 5.517355418448961e-06, "loss": 0.3479, "step": 3532 }, { "epoch": 1.5621199277632405, "grad_norm": 0.3029012171931939, "learning_rate": 5.514797031754254e-06, "loss": 0.3681, "step": 3533 }, { "epoch": 1.5625621936387426, "grad_norm": 0.33585951642157824, "learning_rate": 5.512238508827294e-06, "loss": 0.3122, "step": 3534 }, { "epoch": 1.5630044595142447, "grad_norm": 0.27810575990496056, "learning_rate": 5.509679850345156e-06, "loss": 0.3269, "step": 3535 }, { "epoch": 1.5634467253897468, "grad_norm": 0.28171702930709774, "learning_rate": 5.507121056984942e-06, "loss": 0.2999, "step": 3536 }, { "epoch": 1.563888991265249, "grad_norm": 0.2908078452483837, "learning_rate": 5.5045621294237925e-06, "loss": 0.3041, "step": 3537 }, { "epoch": 1.564331257140751, "grad_norm": 0.3646849094075963, "learning_rate": 5.5020030683388866e-06, "loss": 0.3494, "step": 3538 }, { "epoch": 1.5647735230162532, "grad_norm": 0.3719853189290866, "learning_rate": 5.499443874407434e-06, "loss": 0.3367, "step": 3539 }, { "epoch": 1.5652157888917553, "grad_norm": 0.32099456163371437, "learning_rate": 5.496884548306681e-06, "loss": 0.292, "step": 3540 }, { "epoch": 1.5656580547672576, "grad_norm": 0.2979462497226126, "learning_rate": 5.4943250907139135e-06, "loss": 0.3011, "step": 3541 }, { "epoch": 1.5661003206427597, "grad_norm": 0.3305551620312058, "learning_rate": 5.491765502306443e-06, "loss": 0.3909, "step": 3542 }, { "epoch": 1.566542586518262, "grad_norm": 0.36445382014017386, "learning_rate": 5.489205783761625e-06, "loss": 0.319, "step": 3543 }, { "epoch": 1.5669848523937642, "grad_norm": 0.33219951239415557, "learning_rate": 5.486645935756845e-06, "loss": 0.3522, "step": 3544 }, { "epoch": 1.5674271182692663, "grad_norm": 0.3027404005059349, "learning_rate": 5.4840859589695205e-06, "loss": 0.3412, "step": 3545 }, { "epoch": 1.5678693841447684, "grad_norm": 0.3266290397497774, "learning_rate": 5.481525854077108e-06, "loss": 0.3352, "step": 3546 }, { "epoch": 1.5683116500202705, "grad_norm": 0.2606888153276826, "learning_rate": 5.478965621757094e-06, "loss": 0.3316, "step": 3547 }, { "epoch": 1.5687539158957726, "grad_norm": 0.2859376472406696, "learning_rate": 5.476405262687001e-06, "loss": 0.3358, "step": 3548 }, { "epoch": 1.5691961817712747, "grad_norm": 0.28034190620499355, "learning_rate": 5.473844777544385e-06, "loss": 0.3261, "step": 3549 }, { "epoch": 1.5696384476467768, "grad_norm": 0.2722775670588061, "learning_rate": 5.4712841670068325e-06, "loss": 0.3486, "step": 3550 }, { "epoch": 1.5700807135222792, "grad_norm": 0.4383821846171891, "learning_rate": 5.4687234317519686e-06, "loss": 0.3504, "step": 3551 }, { "epoch": 1.5705229793977813, "grad_norm": 0.34702930151215766, "learning_rate": 5.466162572457447e-06, "loss": 0.3359, "step": 3552 }, { "epoch": 1.5709652452732834, "grad_norm": 0.3365284723561623, "learning_rate": 5.4636015898009554e-06, "loss": 0.3461, "step": 3553 }, { "epoch": 1.5714075111487857, "grad_norm": 0.3298486842378478, "learning_rate": 5.461040484460213e-06, "loss": 0.3185, "step": 3554 }, { "epoch": 1.5718497770242879, "grad_norm": 0.30736313641152735, "learning_rate": 5.458479257112973e-06, "loss": 0.3492, "step": 3555 }, { "epoch": 1.57229204289979, "grad_norm": 0.3082559059645258, "learning_rate": 5.455917908437021e-06, "loss": 0.3218, "step": 3556 }, { "epoch": 1.572734308775292, "grad_norm": 0.3130543958536598, "learning_rate": 5.453356439110175e-06, "loss": 0.2859, "step": 3557 }, { "epoch": 1.5731765746507942, "grad_norm": 0.29159697284904146, "learning_rate": 5.450794849810282e-06, "loss": 0.3592, "step": 3558 }, { "epoch": 1.5736188405262963, "grad_norm": 0.25613853019399, "learning_rate": 5.448233141215223e-06, "loss": 0.3156, "step": 3559 }, { "epoch": 1.5740611064017984, "grad_norm": 0.310833971296688, "learning_rate": 5.445671314002914e-06, "loss": 0.338, "step": 3560 }, { "epoch": 1.5745033722773008, "grad_norm": 0.31987671349503166, "learning_rate": 5.443109368851296e-06, "loss": 0.3733, "step": 3561 }, { "epoch": 1.5749456381528029, "grad_norm": 0.2495141707019509, "learning_rate": 5.440547306438345e-06, "loss": 0.3238, "step": 3562 }, { "epoch": 1.575387904028305, "grad_norm": 0.39376019309699684, "learning_rate": 5.437985127442065e-06, "loss": 0.3353, "step": 3563 }, { "epoch": 1.5758301699038073, "grad_norm": 0.40076884026689513, "learning_rate": 5.435422832540497e-06, "loss": 0.3147, "step": 3564 }, { "epoch": 1.5762724357793094, "grad_norm": 0.3404624347939828, "learning_rate": 5.432860422411705e-06, "loss": 0.3638, "step": 3565 }, { "epoch": 1.5767147016548115, "grad_norm": 0.3316143810634266, "learning_rate": 5.430297897733787e-06, "loss": 0.3281, "step": 3566 }, { "epoch": 1.5771569675303136, "grad_norm": 0.3756355903544134, "learning_rate": 5.427735259184874e-06, "loss": 0.2826, "step": 3567 }, { "epoch": 1.5775992334058158, "grad_norm": 0.26760793477057293, "learning_rate": 5.425172507443125e-06, "loss": 0.3024, "step": 3568 }, { "epoch": 1.5780414992813179, "grad_norm": 0.2898311583565867, "learning_rate": 5.422609643186725e-06, "loss": 0.323, "step": 3569 }, { "epoch": 1.57848376515682, "grad_norm": 0.2896419239805612, "learning_rate": 5.420046667093894e-06, "loss": 0.3466, "step": 3570 }, { "epoch": 1.5789260310323223, "grad_norm": 0.33728209477262794, "learning_rate": 5.417483579842881e-06, "loss": 0.3503, "step": 3571 }, { "epoch": 1.5793682969078244, "grad_norm": 0.32126529912659085, "learning_rate": 5.4149203821119626e-06, "loss": 0.3443, "step": 3572 }, { "epoch": 1.5798105627833265, "grad_norm": 0.29529417408492703, "learning_rate": 5.412357074579444e-06, "loss": 0.3509, "step": 3573 }, { "epoch": 1.5802528286588289, "grad_norm": 0.28829617674774194, "learning_rate": 5.409793657923662e-06, "loss": 0.354, "step": 3574 }, { "epoch": 1.580695094534331, "grad_norm": 0.35165794390234506, "learning_rate": 5.4072301328229816e-06, "loss": 0.3064, "step": 3575 }, { "epoch": 1.581137360409833, "grad_norm": 0.289533636046186, "learning_rate": 5.404666499955796e-06, "loss": 0.3326, "step": 3576 }, { "epoch": 1.5815796262853352, "grad_norm": 0.2798358447158241, "learning_rate": 5.4021027600005245e-06, "loss": 0.3185, "step": 3577 }, { "epoch": 1.5820218921608373, "grad_norm": 0.27293761254693993, "learning_rate": 5.39953891363562e-06, "loss": 0.323, "step": 3578 }, { "epoch": 1.5824641580363394, "grad_norm": 0.33138598371324735, "learning_rate": 5.3969749615395594e-06, "loss": 0.3567, "step": 3579 }, { "epoch": 1.5829064239118416, "grad_norm": 0.3057413967606047, "learning_rate": 5.394410904390849e-06, "loss": 0.3115, "step": 3580 }, { "epoch": 1.5833486897873437, "grad_norm": 0.297964093276475, "learning_rate": 5.391846742868023e-06, "loss": 0.3163, "step": 3581 }, { "epoch": 1.583790955662846, "grad_norm": 0.2900258967098396, "learning_rate": 5.389282477649641e-06, "loss": 0.32, "step": 3582 }, { "epoch": 1.5842332215383481, "grad_norm": 0.28965116158583487, "learning_rate": 5.3867181094142965e-06, "loss": 0.344, "step": 3583 }, { "epoch": 1.5846754874138504, "grad_norm": 0.42238929922614965, "learning_rate": 5.384153638840604e-06, "loss": 0.3547, "step": 3584 }, { "epoch": 1.5851177532893526, "grad_norm": 0.43360345525246496, "learning_rate": 5.3815890666072026e-06, "loss": 0.3395, "step": 3585 }, { "epoch": 1.5855600191648547, "grad_norm": 0.36411693287182895, "learning_rate": 5.379024393392767e-06, "loss": 0.3217, "step": 3586 }, { "epoch": 1.5860022850403568, "grad_norm": 0.31858472166115004, "learning_rate": 5.376459619875994e-06, "loss": 0.3326, "step": 3587 }, { "epoch": 1.586444550915859, "grad_norm": 0.2937762133878651, "learning_rate": 5.373894746735606e-06, "loss": 0.3408, "step": 3588 }, { "epoch": 1.586886816791361, "grad_norm": 0.31462231978922534, "learning_rate": 5.371329774650353e-06, "loss": 0.3639, "step": 3589 }, { "epoch": 1.5873290826668631, "grad_norm": 0.36458444480250035, "learning_rate": 5.368764704299011e-06, "loss": 0.3392, "step": 3590 }, { "epoch": 1.5877713485423652, "grad_norm": 0.25745877316972593, "learning_rate": 5.366199536360383e-06, "loss": 0.2798, "step": 3591 }, { "epoch": 1.5882136144178676, "grad_norm": 0.30363532418837175, "learning_rate": 5.363634271513299e-06, "loss": 0.3218, "step": 3592 }, { "epoch": 1.5886558802933697, "grad_norm": 0.23674556622577533, "learning_rate": 5.361068910436607e-06, "loss": 0.3195, "step": 3593 }, { "epoch": 1.589098146168872, "grad_norm": 0.35327246785719785, "learning_rate": 5.3585034538091885e-06, "loss": 0.2992, "step": 3594 }, { "epoch": 1.5895404120443741, "grad_norm": 0.36014446127669236, "learning_rate": 5.35593790230995e-06, "loss": 0.3269, "step": 3595 }, { "epoch": 1.5899826779198762, "grad_norm": 0.23950743290559276, "learning_rate": 5.353372256617818e-06, "loss": 0.3068, "step": 3596 }, { "epoch": 1.5904249437953784, "grad_norm": 0.2898598441449508, "learning_rate": 5.350806517411748e-06, "loss": 0.3152, "step": 3597 }, { "epoch": 1.5908672096708805, "grad_norm": 0.313096457259036, "learning_rate": 5.34824068537072e-06, "loss": 0.3285, "step": 3598 }, { "epoch": 1.5913094755463826, "grad_norm": 0.41204306572224775, "learning_rate": 5.345674761173738e-06, "loss": 0.3011, "step": 3599 }, { "epoch": 1.5917517414218847, "grad_norm": 0.2861617538903444, "learning_rate": 5.343108745499826e-06, "loss": 0.3343, "step": 3600 }, { "epoch": 1.5921940072973868, "grad_norm": 0.3234175270532802, "learning_rate": 5.340542639028041e-06, "loss": 0.3299, "step": 3601 }, { "epoch": 1.5926362731728891, "grad_norm": 0.32018083409136655, "learning_rate": 5.337976442437458e-06, "loss": 0.3611, "step": 3602 }, { "epoch": 1.5930785390483913, "grad_norm": 0.40097298520109204, "learning_rate": 5.335410156407176e-06, "loss": 0.3475, "step": 3603 }, { "epoch": 1.5935208049238934, "grad_norm": 0.3759474331396036, "learning_rate": 5.332843781616319e-06, "loss": 0.3528, "step": 3604 }, { "epoch": 1.5939630707993957, "grad_norm": 0.3397907572044337, "learning_rate": 5.330277318744035e-06, "loss": 0.3156, "step": 3605 }, { "epoch": 1.5944053366748978, "grad_norm": 0.34677675534248475, "learning_rate": 5.3277107684694925e-06, "loss": 0.3172, "step": 3606 }, { "epoch": 1.5948476025504, "grad_norm": 0.2705773306338499, "learning_rate": 5.325144131471891e-06, "loss": 0.2983, "step": 3607 }, { "epoch": 1.595289868425902, "grad_norm": 0.36466143581455207, "learning_rate": 5.322577408430439e-06, "loss": 0.3602, "step": 3608 }, { "epoch": 1.5957321343014041, "grad_norm": 0.3016116036459552, "learning_rate": 5.320010600024384e-06, "loss": 0.3181, "step": 3609 }, { "epoch": 1.5961744001769063, "grad_norm": 0.3032149990279192, "learning_rate": 5.317443706932982e-06, "loss": 0.3185, "step": 3610 }, { "epoch": 1.5966166660524084, "grad_norm": 0.2764831787708463, "learning_rate": 5.3148767298355195e-06, "loss": 0.3464, "step": 3611 }, { "epoch": 1.5970589319279107, "grad_norm": 0.3439868599066557, "learning_rate": 5.312309669411304e-06, "loss": 0.3521, "step": 3612 }, { "epoch": 1.5975011978034128, "grad_norm": 0.37037906138563287, "learning_rate": 5.309742526339664e-06, "loss": 0.3649, "step": 3613 }, { "epoch": 1.597943463678915, "grad_norm": 0.3614650872723846, "learning_rate": 5.307175301299949e-06, "loss": 0.3358, "step": 3614 }, { "epoch": 1.5983857295544173, "grad_norm": 0.32564454499312223, "learning_rate": 5.304607994971533e-06, "loss": 0.3505, "step": 3615 }, { "epoch": 1.5988279954299194, "grad_norm": 0.29800788933457106, "learning_rate": 5.302040608033807e-06, "loss": 0.3451, "step": 3616 }, { "epoch": 1.5992702613054215, "grad_norm": 0.29057984738949133, "learning_rate": 5.29947314116619e-06, "loss": 0.3113, "step": 3617 }, { "epoch": 1.5997125271809236, "grad_norm": 0.2878775303905285, "learning_rate": 5.296905595048115e-06, "loss": 0.3345, "step": 3618 }, { "epoch": 1.6001547930564257, "grad_norm": 0.27435805457725265, "learning_rate": 5.294337970359042e-06, "loss": 0.3211, "step": 3619 }, { "epoch": 1.6005970589319278, "grad_norm": 0.29032105187018714, "learning_rate": 5.291770267778447e-06, "loss": 0.3171, "step": 3620 }, { "epoch": 1.60103932480743, "grad_norm": 0.3350617627253168, "learning_rate": 5.289202487985828e-06, "loss": 0.3228, "step": 3621 }, { "epoch": 1.6014815906829323, "grad_norm": 0.3473651081887665, "learning_rate": 5.2866346316607055e-06, "loss": 0.3552, "step": 3622 }, { "epoch": 1.6019238565584344, "grad_norm": 0.3632489225756051, "learning_rate": 5.284066699482622e-06, "loss": 0.3383, "step": 3623 }, { "epoch": 1.6023661224339365, "grad_norm": 0.2962237082006966, "learning_rate": 5.281498692131129e-06, "loss": 0.3169, "step": 3624 }, { "epoch": 1.6028083883094388, "grad_norm": 0.3308611835589382, "learning_rate": 5.278930610285813e-06, "loss": 0.3728, "step": 3625 }, { "epoch": 1.603250654184941, "grad_norm": 0.34306845517401974, "learning_rate": 5.27636245462627e-06, "loss": 0.3332, "step": 3626 }, { "epoch": 1.603692920060443, "grad_norm": 0.28752838467332986, "learning_rate": 5.273794225832121e-06, "loss": 0.3165, "step": 3627 }, { "epoch": 1.6041351859359452, "grad_norm": 0.3156805782572505, "learning_rate": 5.271225924583e-06, "loss": 0.3668, "step": 3628 }, { "epoch": 1.6045774518114473, "grad_norm": 0.28974848331060843, "learning_rate": 5.268657551558567e-06, "loss": 0.3083, "step": 3629 }, { "epoch": 1.6050197176869494, "grad_norm": 0.2627625696157416, "learning_rate": 5.266089107438497e-06, "loss": 0.3203, "step": 3630 }, { "epoch": 1.6054619835624515, "grad_norm": 0.25827745643889544, "learning_rate": 5.263520592902485e-06, "loss": 0.345, "step": 3631 }, { "epoch": 1.6059042494379536, "grad_norm": 0.33919472167940856, "learning_rate": 5.260952008630244e-06, "loss": 0.3531, "step": 3632 }, { "epoch": 1.606346515313456, "grad_norm": 0.37792892841991355, "learning_rate": 5.2583833553015075e-06, "loss": 0.3716, "step": 3633 }, { "epoch": 1.606788781188958, "grad_norm": 0.33309989498142756, "learning_rate": 5.255814633596025e-06, "loss": 0.3278, "step": 3634 }, { "epoch": 1.6072310470644604, "grad_norm": 0.37786072042531815, "learning_rate": 5.253245844193564e-06, "loss": 0.4084, "step": 3635 }, { "epoch": 1.6076733129399625, "grad_norm": 0.3027596191597167, "learning_rate": 5.250676987773913e-06, "loss": 0.3674, "step": 3636 }, { "epoch": 1.6081155788154646, "grad_norm": 0.2961773975875853, "learning_rate": 5.2481080650168734e-06, "loss": 0.2927, "step": 3637 }, { "epoch": 1.6085578446909667, "grad_norm": 0.24421749078659466, "learning_rate": 5.24553907660227e-06, "loss": 0.351, "step": 3638 }, { "epoch": 1.6090001105664689, "grad_norm": 0.41077903023940554, "learning_rate": 5.242970023209938e-06, "loss": 0.3273, "step": 3639 }, { "epoch": 1.609442376441971, "grad_norm": 0.3672675878799052, "learning_rate": 5.240400905519734e-06, "loss": 0.3468, "step": 3640 }, { "epoch": 1.609884642317473, "grad_norm": 0.3254246089721681, "learning_rate": 5.237831724211534e-06, "loss": 0.351, "step": 3641 }, { "epoch": 1.6103269081929752, "grad_norm": 0.3770225242116248, "learning_rate": 5.2352624799652275e-06, "loss": 0.3315, "step": 3642 }, { "epoch": 1.6107691740684775, "grad_norm": 0.3782396155453356, "learning_rate": 5.232693173460719e-06, "loss": 0.3517, "step": 3643 }, { "epoch": 1.6112114399439796, "grad_norm": 0.2684933801646838, "learning_rate": 5.2301238053779325e-06, "loss": 0.3062, "step": 3644 }, { "epoch": 1.611653705819482, "grad_norm": 0.29322998522999755, "learning_rate": 5.227554376396807e-06, "loss": 0.3658, "step": 3645 }, { "epoch": 1.612095971694984, "grad_norm": 0.3164223468933281, "learning_rate": 5.224984887197301e-06, "loss": 0.3375, "step": 3646 }, { "epoch": 1.6125382375704862, "grad_norm": 0.2798264674121568, "learning_rate": 5.222415338459382e-06, "loss": 0.3431, "step": 3647 }, { "epoch": 1.6129805034459883, "grad_norm": 0.40320403911174246, "learning_rate": 5.219845730863039e-06, "loss": 0.3503, "step": 3648 }, { "epoch": 1.6134227693214904, "grad_norm": 0.30463124368508643, "learning_rate": 5.217276065088275e-06, "loss": 0.3433, "step": 3649 }, { "epoch": 1.6138650351969925, "grad_norm": 0.3292771496440867, "learning_rate": 5.2147063418151085e-06, "loss": 0.2657, "step": 3650 }, { "epoch": 1.6143073010724946, "grad_norm": 0.34519796725577, "learning_rate": 5.212136561723571e-06, "loss": 0.3529, "step": 3651 }, { "epoch": 1.6147495669479968, "grad_norm": 0.28851080613999286, "learning_rate": 5.209566725493713e-06, "loss": 0.3283, "step": 3652 }, { "epoch": 1.615191832823499, "grad_norm": 0.3153596890187383, "learning_rate": 5.206996833805595e-06, "loss": 0.3177, "step": 3653 }, { "epoch": 1.6156340986990012, "grad_norm": 0.28124014561429006, "learning_rate": 5.204426887339301e-06, "loss": 0.3467, "step": 3654 }, { "epoch": 1.6160763645745033, "grad_norm": 0.30338297007603626, "learning_rate": 5.201856886774917e-06, "loss": 0.3236, "step": 3655 }, { "epoch": 1.6165186304500057, "grad_norm": 0.330055209960924, "learning_rate": 5.199286832792553e-06, "loss": 0.3079, "step": 3656 }, { "epoch": 1.6169608963255078, "grad_norm": 0.3209034996955531, "learning_rate": 5.196716726072329e-06, "loss": 0.3235, "step": 3657 }, { "epoch": 1.6174031622010099, "grad_norm": 0.33763924540944357, "learning_rate": 5.194146567294381e-06, "loss": 0.3828, "step": 3658 }, { "epoch": 1.617845428076512, "grad_norm": 0.4379525305776719, "learning_rate": 5.191576357138856e-06, "loss": 0.3, "step": 3659 }, { "epoch": 1.618287693952014, "grad_norm": 0.2876323785305917, "learning_rate": 5.189006096285917e-06, "loss": 0.2914, "step": 3660 }, { "epoch": 1.6187299598275162, "grad_norm": 0.2861829218643974, "learning_rate": 5.186435785415739e-06, "loss": 0.3451, "step": 3661 }, { "epoch": 1.6191722257030183, "grad_norm": 0.3434736045022986, "learning_rate": 5.183865425208513e-06, "loss": 0.3451, "step": 3662 }, { "epoch": 1.6196144915785207, "grad_norm": 0.3191370965968404, "learning_rate": 5.181295016344439e-06, "loss": 0.339, "step": 3663 }, { "epoch": 1.6200567574540228, "grad_norm": 0.34925837747616106, "learning_rate": 5.178724559503731e-06, "loss": 0.3678, "step": 3664 }, { "epoch": 1.620499023329525, "grad_norm": 0.28679268220382376, "learning_rate": 5.17615405536662e-06, "loss": 0.3179, "step": 3665 }, { "epoch": 1.6209412892050272, "grad_norm": 0.28655450316006226, "learning_rate": 5.173583504613345e-06, "loss": 0.328, "step": 3666 }, { "epoch": 1.6213835550805293, "grad_norm": 0.3458779200869171, "learning_rate": 5.171012907924156e-06, "loss": 0.3306, "step": 3667 }, { "epoch": 1.6218258209560314, "grad_norm": 0.28462317686743577, "learning_rate": 5.168442265979319e-06, "loss": 0.3292, "step": 3668 }, { "epoch": 1.6222680868315336, "grad_norm": 0.3134215316766922, "learning_rate": 5.16587157945911e-06, "loss": 0.3011, "step": 3669 }, { "epoch": 1.6227103527070357, "grad_norm": 0.33589870426564156, "learning_rate": 5.163300849043819e-06, "loss": 0.3056, "step": 3670 }, { "epoch": 1.6231526185825378, "grad_norm": 0.2696713446790209, "learning_rate": 5.160730075413745e-06, "loss": 0.3254, "step": 3671 }, { "epoch": 1.62359488445804, "grad_norm": 0.259847172292252, "learning_rate": 5.158159259249197e-06, "loss": 0.3344, "step": 3672 }, { "epoch": 1.6240371503335422, "grad_norm": 0.3058216207128465, "learning_rate": 5.155588401230501e-06, "loss": 0.3025, "step": 3673 }, { "epoch": 1.6244794162090443, "grad_norm": 0.33503560908072155, "learning_rate": 5.153017502037991e-06, "loss": 0.3461, "step": 3674 }, { "epoch": 1.6249216820845465, "grad_norm": 0.36769174076151234, "learning_rate": 5.150446562352008e-06, "loss": 0.3599, "step": 3675 }, { "epoch": 1.6253639479600488, "grad_norm": 0.2728351757654696, "learning_rate": 5.1478755828529105e-06, "loss": 0.3308, "step": 3676 }, { "epoch": 1.625806213835551, "grad_norm": 0.2917363457818481, "learning_rate": 5.1453045642210635e-06, "loss": 0.2981, "step": 3677 }, { "epoch": 1.626248479711053, "grad_norm": 0.2906638642005102, "learning_rate": 5.142733507136842e-06, "loss": 0.3292, "step": 3678 }, { "epoch": 1.6266907455865551, "grad_norm": 0.27507255953778414, "learning_rate": 5.140162412280635e-06, "loss": 0.3303, "step": 3679 }, { "epoch": 1.6271330114620572, "grad_norm": 0.35169776563413263, "learning_rate": 5.137591280332834e-06, "loss": 0.3416, "step": 3680 }, { "epoch": 1.6275752773375594, "grad_norm": 0.44030253781237805, "learning_rate": 5.135020111973853e-06, "loss": 0.362, "step": 3681 }, { "epoch": 1.6280175432130615, "grad_norm": 0.3290207604119807, "learning_rate": 5.1324489078841e-06, "loss": 0.327, "step": 3682 }, { "epoch": 1.6284598090885636, "grad_norm": 0.3445175477004971, "learning_rate": 5.129877668744005e-06, "loss": 0.3381, "step": 3683 }, { "epoch": 1.628902074964066, "grad_norm": 0.32434195362094703, "learning_rate": 5.127306395234001e-06, "loss": 0.3582, "step": 3684 }, { "epoch": 1.629344340839568, "grad_norm": 0.3784606395023504, "learning_rate": 5.1247350880345315e-06, "loss": 0.4121, "step": 3685 }, { "epoch": 1.6297866067150704, "grad_norm": 0.3409068584911197, "learning_rate": 5.122163747826049e-06, "loss": 0.3143, "step": 3686 }, { "epoch": 1.6302288725905725, "grad_norm": 0.32280001479653353, "learning_rate": 5.119592375289015e-06, "loss": 0.3076, "step": 3687 }, { "epoch": 1.6306711384660746, "grad_norm": 0.2634003000105853, "learning_rate": 5.1170209711039e-06, "loss": 0.3072, "step": 3688 }, { "epoch": 1.6311134043415767, "grad_norm": 0.30967474909614123, "learning_rate": 5.114449535951184e-06, "loss": 0.3742, "step": 3689 }, { "epoch": 1.6315556702170788, "grad_norm": 0.339005937458873, "learning_rate": 5.111878070511347e-06, "loss": 0.3281, "step": 3690 }, { "epoch": 1.631997936092581, "grad_norm": 0.29077774544417134, "learning_rate": 5.109306575464891e-06, "loss": 0.31, "step": 3691 }, { "epoch": 1.632440201968083, "grad_norm": 0.3287932226908382, "learning_rate": 5.106735051492313e-06, "loss": 0.3227, "step": 3692 }, { "epoch": 1.6328824678435852, "grad_norm": 0.32010021040549835, "learning_rate": 5.104163499274126e-06, "loss": 0.3381, "step": 3693 }, { "epoch": 1.6333247337190875, "grad_norm": 0.2561674081115152, "learning_rate": 5.101591919490847e-06, "loss": 0.3152, "step": 3694 }, { "epoch": 1.6337669995945896, "grad_norm": 0.3519172903146926, "learning_rate": 5.099020312822999e-06, "loss": 0.3018, "step": 3695 }, { "epoch": 1.634209265470092, "grad_norm": 0.36747135927832963, "learning_rate": 5.096448679951114e-06, "loss": 0.3556, "step": 3696 }, { "epoch": 1.634651531345594, "grad_norm": 0.26994248293280265, "learning_rate": 5.093877021555733e-06, "loss": 0.3085, "step": 3697 }, { "epoch": 1.6350937972210962, "grad_norm": 0.27907959888756895, "learning_rate": 5.0913053383174e-06, "loss": 0.3574, "step": 3698 }, { "epoch": 1.6355360630965983, "grad_norm": 0.3556758552301501, "learning_rate": 5.088733630916667e-06, "loss": 0.3289, "step": 3699 }, { "epoch": 1.6359783289721004, "grad_norm": 0.28355808966408746, "learning_rate": 5.086161900034092e-06, "loss": 0.3527, "step": 3700 }, { "epoch": 1.6364205948476025, "grad_norm": 0.3038695328110003, "learning_rate": 5.083590146350241e-06, "loss": 0.2845, "step": 3701 }, { "epoch": 1.6368628607231046, "grad_norm": 0.3031996306198581, "learning_rate": 5.081018370545684e-06, "loss": 0.3328, "step": 3702 }, { "epoch": 1.6373051265986067, "grad_norm": 0.3147680348583536, "learning_rate": 5.078446573300998e-06, "loss": 0.301, "step": 3703 }, { "epoch": 1.637747392474109, "grad_norm": 0.3122220911257422, "learning_rate": 5.075874755296764e-06, "loss": 0.3313, "step": 3704 }, { "epoch": 1.6381896583496112, "grad_norm": 0.3161834911555147, "learning_rate": 5.073302917213573e-06, "loss": 0.3378, "step": 3705 }, { "epoch": 1.6386319242251133, "grad_norm": 0.24156531469723644, "learning_rate": 5.070731059732012e-06, "loss": 0.3195, "step": 3706 }, { "epoch": 1.6390741901006156, "grad_norm": 0.3603454909004796, "learning_rate": 5.068159183532686e-06, "loss": 0.3128, "step": 3707 }, { "epoch": 1.6395164559761177, "grad_norm": 0.2881043703716912, "learning_rate": 5.065587289296194e-06, "loss": 0.3059, "step": 3708 }, { "epoch": 1.6399587218516198, "grad_norm": 0.33444430656512936, "learning_rate": 5.0630153777031464e-06, "loss": 0.3497, "step": 3709 }, { "epoch": 1.640400987727122, "grad_norm": 0.2985062792910012, "learning_rate": 5.060443449434154e-06, "loss": 0.2931, "step": 3710 }, { "epoch": 1.640843253602624, "grad_norm": 0.28866093383100533, "learning_rate": 5.057871505169833e-06, "loss": 0.3225, "step": 3711 }, { "epoch": 1.6412855194781262, "grad_norm": 0.3534970696391503, "learning_rate": 5.055299545590806e-06, "loss": 0.338, "step": 3712 }, { "epoch": 1.6417277853536283, "grad_norm": 0.3028804357160922, "learning_rate": 5.052727571377698e-06, "loss": 0.2991, "step": 3713 }, { "epoch": 1.6421700512291306, "grad_norm": 0.34189414339845314, "learning_rate": 5.050155583211137e-06, "loss": 0.3378, "step": 3714 }, { "epoch": 1.6426123171046327, "grad_norm": 0.2835890857417111, "learning_rate": 5.047583581771758e-06, "loss": 0.3432, "step": 3715 }, { "epoch": 1.6430545829801348, "grad_norm": 0.24985727789649081, "learning_rate": 5.045011567740196e-06, "loss": 0.315, "step": 3716 }, { "epoch": 1.6434968488556372, "grad_norm": 0.2832905599696625, "learning_rate": 5.042439541797091e-06, "loss": 0.3238, "step": 3717 }, { "epoch": 1.6439391147311393, "grad_norm": 0.31075712623000407, "learning_rate": 5.039867504623084e-06, "loss": 0.3172, "step": 3718 }, { "epoch": 1.6443813806066414, "grad_norm": 0.31698821728245263, "learning_rate": 5.037295456898822e-06, "loss": 0.3767, "step": 3719 }, { "epoch": 1.6448236464821435, "grad_norm": 0.3505508138017955, "learning_rate": 5.034723399304956e-06, "loss": 0.3206, "step": 3720 }, { "epoch": 1.6452659123576456, "grad_norm": 0.3209227162271587, "learning_rate": 5.032151332522134e-06, "loss": 0.312, "step": 3721 }, { "epoch": 1.6457081782331477, "grad_norm": 0.3298560505307982, "learning_rate": 5.0295792572310085e-06, "loss": 0.3415, "step": 3722 }, { "epoch": 1.6461504441086499, "grad_norm": 0.31683901888482685, "learning_rate": 5.02700717411224e-06, "loss": 0.3011, "step": 3723 }, { "epoch": 1.6465927099841522, "grad_norm": 0.3413239910271752, "learning_rate": 5.0244350838464815e-06, "loss": 0.3158, "step": 3724 }, { "epoch": 1.6470349758596543, "grad_norm": 0.34959450527499153, "learning_rate": 5.021862987114397e-06, "loss": 0.3468, "step": 3725 }, { "epoch": 1.6474772417351564, "grad_norm": 0.245336477823575, "learning_rate": 5.019290884596645e-06, "loss": 0.299, "step": 3726 }, { "epoch": 1.6479195076106588, "grad_norm": 0.39051705628866684, "learning_rate": 5.016718776973888e-06, "loss": 0.3268, "step": 3727 }, { "epoch": 1.6483617734861609, "grad_norm": 0.33784064602866437, "learning_rate": 5.0141466649267945e-06, "loss": 0.3483, "step": 3728 }, { "epoch": 1.648804039361663, "grad_norm": 0.2946082057555738, "learning_rate": 5.011574549136027e-06, "loss": 0.3231, "step": 3729 }, { "epoch": 1.649246305237165, "grad_norm": 0.2750744212748355, "learning_rate": 5.009002430282251e-06, "loss": 0.3328, "step": 3730 }, { "epoch": 1.6496885711126672, "grad_norm": 0.3667950737142476, "learning_rate": 5.006430309046138e-06, "loss": 0.3396, "step": 3731 }, { "epoch": 1.6501308369881693, "grad_norm": 0.27010312340457193, "learning_rate": 5.003858186108353e-06, "loss": 0.3229, "step": 3732 }, { "epoch": 1.6505731028636714, "grad_norm": 0.31775620636556773, "learning_rate": 5.001286062149563e-06, "loss": 0.3154, "step": 3733 }, { "epoch": 1.6510153687391735, "grad_norm": 0.2919785794601136, "learning_rate": 4.998713937850438e-06, "loss": 0.3456, "step": 3734 }, { "epoch": 1.6514576346146759, "grad_norm": 0.2882687435707963, "learning_rate": 4.996141813891649e-06, "loss": 0.3095, "step": 3735 }, { "epoch": 1.651899900490178, "grad_norm": 0.31742344276415896, "learning_rate": 4.993569690953865e-06, "loss": 0.311, "step": 3736 }, { "epoch": 1.6523421663656803, "grad_norm": 0.31723411462635154, "learning_rate": 4.99099756971775e-06, "loss": 0.3501, "step": 3737 }, { "epoch": 1.6527844322411824, "grad_norm": 0.3638552475402282, "learning_rate": 4.9884254508639755e-06, "loss": 0.3151, "step": 3738 }, { "epoch": 1.6532266981166845, "grad_norm": 0.3029780931186728, "learning_rate": 4.985853335073206e-06, "loss": 0.3692, "step": 3739 }, { "epoch": 1.6536689639921867, "grad_norm": 0.32228743890870126, "learning_rate": 4.983281223026113e-06, "loss": 0.3496, "step": 3740 }, { "epoch": 1.6541112298676888, "grad_norm": 0.3046138064968732, "learning_rate": 4.980709115403357e-06, "loss": 0.3694, "step": 3741 }, { "epoch": 1.6545534957431909, "grad_norm": 0.30901745019203036, "learning_rate": 4.978137012885605e-06, "loss": 0.2988, "step": 3742 }, { "epoch": 1.654995761618693, "grad_norm": 0.295055004664679, "learning_rate": 4.975564916153519e-06, "loss": 0.3681, "step": 3743 }, { "epoch": 1.655438027494195, "grad_norm": 0.31476532756067893, "learning_rate": 4.972992825887763e-06, "loss": 0.3406, "step": 3744 }, { "epoch": 1.6558802933696974, "grad_norm": 0.2763587976458054, "learning_rate": 4.9704207427689915e-06, "loss": 0.3111, "step": 3745 }, { "epoch": 1.6563225592451996, "grad_norm": 0.2996510736878437, "learning_rate": 4.967848667477868e-06, "loss": 0.3256, "step": 3746 }, { "epoch": 1.656764825120702, "grad_norm": 0.3612542359051443, "learning_rate": 4.965276600695045e-06, "loss": 0.3357, "step": 3747 }, { "epoch": 1.657207090996204, "grad_norm": 0.31856593715375975, "learning_rate": 4.96270454310118e-06, "loss": 0.3283, "step": 3748 }, { "epoch": 1.6576493568717061, "grad_norm": 0.2660842021235755, "learning_rate": 4.960132495376919e-06, "loss": 0.3062, "step": 3749 }, { "epoch": 1.6580916227472082, "grad_norm": 0.3262914880546213, "learning_rate": 4.95756045820291e-06, "loss": 0.3458, "step": 3750 }, { "epoch": 1.6585338886227103, "grad_norm": 0.30146329165298086, "learning_rate": 4.954988432259806e-06, "loss": 0.3618, "step": 3751 }, { "epoch": 1.6589761544982125, "grad_norm": 0.3461719922449565, "learning_rate": 4.9524164182282435e-06, "loss": 0.3399, "step": 3752 }, { "epoch": 1.6594184203737146, "grad_norm": 0.33563535503716346, "learning_rate": 4.9498444167888644e-06, "loss": 0.3362, "step": 3753 }, { "epoch": 1.6598606862492167, "grad_norm": 0.27956229733753335, "learning_rate": 4.9472724286223045e-06, "loss": 0.3044, "step": 3754 }, { "epoch": 1.660302952124719, "grad_norm": 0.3266013939179187, "learning_rate": 4.944700454409195e-06, "loss": 0.3179, "step": 3755 }, { "epoch": 1.6607452180002211, "grad_norm": 0.32599517395292893, "learning_rate": 4.942128494830168e-06, "loss": 0.3434, "step": 3756 }, { "epoch": 1.6611874838757232, "grad_norm": 0.2864295148616196, "learning_rate": 4.939556550565848e-06, "loss": 0.3209, "step": 3757 }, { "epoch": 1.6616297497512256, "grad_norm": 0.35511485696496825, "learning_rate": 4.936984622296855e-06, "loss": 0.3365, "step": 3758 }, { "epoch": 1.6620720156267277, "grad_norm": 0.33283863279853754, "learning_rate": 4.934412710703807e-06, "loss": 0.3469, "step": 3759 }, { "epoch": 1.6625142815022298, "grad_norm": 0.2826218835608126, "learning_rate": 4.931840816467314e-06, "loss": 0.2978, "step": 3760 }, { "epoch": 1.662956547377732, "grad_norm": 0.9233059501354478, "learning_rate": 4.929268940267987e-06, "loss": 0.3214, "step": 3761 }, { "epoch": 1.663398813253234, "grad_norm": 0.2897604596305762, "learning_rate": 4.9266970827864286e-06, "loss": 0.2877, "step": 3762 }, { "epoch": 1.6638410791287361, "grad_norm": 0.35990194118428603, "learning_rate": 4.924125244703237e-06, "loss": 0.3524, "step": 3763 }, { "epoch": 1.6642833450042382, "grad_norm": 0.3672600061136613, "learning_rate": 4.9215534266990045e-06, "loss": 0.3549, "step": 3764 }, { "epoch": 1.6647256108797406, "grad_norm": 0.28796654397122073, "learning_rate": 4.918981629454318e-06, "loss": 0.3453, "step": 3765 }, { "epoch": 1.6651678767552427, "grad_norm": 0.30200557052384636, "learning_rate": 4.916409853649759e-06, "loss": 0.32, "step": 3766 }, { "epoch": 1.6656101426307448, "grad_norm": 0.3269124877234133, "learning_rate": 4.9138380999659095e-06, "loss": 0.3421, "step": 3767 }, { "epoch": 1.6660524085062471, "grad_norm": 0.30345668215265525, "learning_rate": 4.9112663690833355e-06, "loss": 0.3517, "step": 3768 }, { "epoch": 1.6664946743817493, "grad_norm": 0.3415803233101787, "learning_rate": 4.908694661682602e-06, "loss": 0.3375, "step": 3769 }, { "epoch": 1.6669369402572514, "grad_norm": 0.25223283092287435, "learning_rate": 4.906122978444268e-06, "loss": 0.2996, "step": 3770 }, { "epoch": 1.6673792061327535, "grad_norm": 0.4425116053717259, "learning_rate": 4.903551320048887e-06, "loss": 0.3729, "step": 3771 }, { "epoch": 1.6678214720082556, "grad_norm": 0.3004501330411138, "learning_rate": 4.900979687177003e-06, "loss": 0.3233, "step": 3772 }, { "epoch": 1.6682637378837577, "grad_norm": 0.3594836652314782, "learning_rate": 4.898408080509155e-06, "loss": 0.375, "step": 3773 }, { "epoch": 1.6687060037592598, "grad_norm": 0.2840211376034278, "learning_rate": 4.8958365007258755e-06, "loss": 0.3004, "step": 3774 }, { "epoch": 1.6691482696347621, "grad_norm": 0.29985495459002676, "learning_rate": 4.8932649485076875e-06, "loss": 0.3091, "step": 3775 }, { "epoch": 1.6695905355102643, "grad_norm": 0.47043771362515546, "learning_rate": 4.89069342453511e-06, "loss": 0.3651, "step": 3776 }, { "epoch": 1.6700328013857664, "grad_norm": 0.2801909525491551, "learning_rate": 4.8881219294886525e-06, "loss": 0.334, "step": 3777 }, { "epoch": 1.6704750672612687, "grad_norm": 0.43476723648401633, "learning_rate": 4.885550464048818e-06, "loss": 0.3664, "step": 3778 }, { "epoch": 1.6709173331367708, "grad_norm": 0.3171923221390185, "learning_rate": 4.882979028896102e-06, "loss": 0.3139, "step": 3779 }, { "epoch": 1.671359599012273, "grad_norm": 0.3353946080038912, "learning_rate": 4.880407624710986e-06, "loss": 0.3417, "step": 3780 }, { "epoch": 1.671801864887775, "grad_norm": 0.3061984338132856, "learning_rate": 4.877836252173953e-06, "loss": 0.3445, "step": 3781 }, { "epoch": 1.6722441307632772, "grad_norm": 0.3260708593336216, "learning_rate": 4.87526491196547e-06, "loss": 0.3132, "step": 3782 }, { "epoch": 1.6726863966387793, "grad_norm": 0.2872855295432387, "learning_rate": 4.872693604766002e-06, "loss": 0.3272, "step": 3783 }, { "epoch": 1.6731286625142814, "grad_norm": 0.28365147623775794, "learning_rate": 4.870122331255997e-06, "loss": 0.3311, "step": 3784 }, { "epoch": 1.6735709283897835, "grad_norm": 0.3058033188931645, "learning_rate": 4.867551092115902e-06, "loss": 0.3426, "step": 3785 }, { "epoch": 1.6740131942652858, "grad_norm": 0.34298631548653036, "learning_rate": 4.86497988802615e-06, "loss": 0.3387, "step": 3786 }, { "epoch": 1.674455460140788, "grad_norm": 0.33674138609756094, "learning_rate": 4.862408719667166e-06, "loss": 0.3468, "step": 3787 }, { "epoch": 1.6748977260162903, "grad_norm": 0.31160993500109124, "learning_rate": 4.859837587719367e-06, "loss": 0.3211, "step": 3788 }, { "epoch": 1.6753399918917924, "grad_norm": 0.29467845310387875, "learning_rate": 4.857266492863159e-06, "loss": 0.347, "step": 3789 }, { "epoch": 1.6757822577672945, "grad_norm": 0.31323419801779584, "learning_rate": 4.854695435778938e-06, "loss": 0.3352, "step": 3790 }, { "epoch": 1.6762245236427966, "grad_norm": 0.3157658633094383, "learning_rate": 4.852124417147092e-06, "loss": 0.3495, "step": 3791 }, { "epoch": 1.6766667895182987, "grad_norm": 0.316377556348695, "learning_rate": 4.849553437647993e-06, "loss": 0.3114, "step": 3792 }, { "epoch": 1.6771090553938008, "grad_norm": 0.28297623856432913, "learning_rate": 4.846982497962011e-06, "loss": 0.2932, "step": 3793 }, { "epoch": 1.677551321269303, "grad_norm": 0.25912075919366256, "learning_rate": 4.8444115987695e-06, "loss": 0.3032, "step": 3794 }, { "epoch": 1.677993587144805, "grad_norm": 0.3079357766641423, "learning_rate": 4.841840740750805e-06, "loss": 0.3627, "step": 3795 }, { "epoch": 1.6784358530203074, "grad_norm": 0.39700053227062276, "learning_rate": 4.839269924586259e-06, "loss": 0.3167, "step": 3796 }, { "epoch": 1.6788781188958095, "grad_norm": 0.2729502783053691, "learning_rate": 4.836699150956182e-06, "loss": 0.3217, "step": 3797 }, { "epoch": 1.6793203847713118, "grad_norm": 0.3045976726442567, "learning_rate": 4.834128420540891e-06, "loss": 0.2831, "step": 3798 }, { "epoch": 1.679762650646814, "grad_norm": 0.30479404842424207, "learning_rate": 4.8315577340206835e-06, "loss": 0.3076, "step": 3799 }, { "epoch": 1.680204916522316, "grad_norm": 0.3356214658677951, "learning_rate": 4.828987092075847e-06, "loss": 0.3356, "step": 3800 }, { "epoch": 1.6806471823978182, "grad_norm": 0.2792814980326527, "learning_rate": 4.826416495386657e-06, "loss": 0.2826, "step": 3801 }, { "epoch": 1.6810894482733203, "grad_norm": 0.3195107035234557, "learning_rate": 4.823845944633382e-06, "loss": 0.3332, "step": 3802 }, { "epoch": 1.6815317141488224, "grad_norm": 0.3729707595105956, "learning_rate": 4.821275440496269e-06, "loss": 0.3344, "step": 3803 }, { "epoch": 1.6819739800243245, "grad_norm": 0.3138908734310869, "learning_rate": 4.818704983655563e-06, "loss": 0.3054, "step": 3804 }, { "epoch": 1.6824162458998266, "grad_norm": 0.3812094613304706, "learning_rate": 4.8161345747914885e-06, "loss": 0.3501, "step": 3805 }, { "epoch": 1.682858511775329, "grad_norm": 0.35146299166135214, "learning_rate": 4.813564214584262e-06, "loss": 0.3294, "step": 3806 }, { "epoch": 1.683300777650831, "grad_norm": 0.3071185422347154, "learning_rate": 4.810993903714087e-06, "loss": 0.3464, "step": 3807 }, { "epoch": 1.6837430435263332, "grad_norm": 0.2743590829482478, "learning_rate": 4.808423642861145e-06, "loss": 0.3056, "step": 3808 }, { "epoch": 1.6841853094018355, "grad_norm": 0.3276482172892194, "learning_rate": 4.8058534327056204e-06, "loss": 0.3243, "step": 3809 }, { "epoch": 1.6846275752773376, "grad_norm": 0.3417013188942843, "learning_rate": 4.803283273927672e-06, "loss": 0.3535, "step": 3810 }, { "epoch": 1.6850698411528398, "grad_norm": 0.2933965805141081, "learning_rate": 4.800713167207449e-06, "loss": 0.3513, "step": 3811 }, { "epoch": 1.6855121070283419, "grad_norm": 0.3934953001425071, "learning_rate": 4.798143113225085e-06, "loss": 0.3604, "step": 3812 }, { "epoch": 1.685954372903844, "grad_norm": 0.300992923266916, "learning_rate": 4.7955731126607e-06, "loss": 0.2955, "step": 3813 }, { "epoch": 1.686396638779346, "grad_norm": 0.3024666956835371, "learning_rate": 4.793003166194405e-06, "loss": 0.3162, "step": 3814 }, { "epoch": 1.6868389046548482, "grad_norm": 0.32983892058214864, "learning_rate": 4.790433274506289e-06, "loss": 0.2998, "step": 3815 }, { "epoch": 1.6872811705303505, "grad_norm": 0.3520708207973646, "learning_rate": 4.78786343827643e-06, "loss": 0.3334, "step": 3816 }, { "epoch": 1.6877234364058527, "grad_norm": 0.32675693740182044, "learning_rate": 4.785293658184894e-06, "loss": 0.3395, "step": 3817 }, { "epoch": 1.6881657022813548, "grad_norm": 0.3504888044300263, "learning_rate": 4.7827239349117265e-06, "loss": 0.3288, "step": 3818 }, { "epoch": 1.688607968156857, "grad_norm": 0.3003528832327105, "learning_rate": 4.7801542691369614e-06, "loss": 0.3271, "step": 3819 }, { "epoch": 1.6890502340323592, "grad_norm": 0.3240644184906291, "learning_rate": 4.777584661540619e-06, "loss": 0.3677, "step": 3820 }, { "epoch": 1.6894924999078613, "grad_norm": 0.34745267396242746, "learning_rate": 4.7750151128027e-06, "loss": 0.3162, "step": 3821 }, { "epoch": 1.6899347657833634, "grad_norm": 0.2657372330901539, "learning_rate": 4.7724456236031944e-06, "loss": 0.3314, "step": 3822 }, { "epoch": 1.6903770316588655, "grad_norm": 0.2855711011264239, "learning_rate": 4.769876194622069e-06, "loss": 0.3115, "step": 3823 }, { "epoch": 1.6908192975343677, "grad_norm": 0.4012502085506, "learning_rate": 4.767306826539281e-06, "loss": 0.3321, "step": 3824 }, { "epoch": 1.6912615634098698, "grad_norm": 0.34307035547322245, "learning_rate": 4.764737520034774e-06, "loss": 0.332, "step": 3825 }, { "epoch": 1.691703829285372, "grad_norm": 0.28436481552663767, "learning_rate": 4.762168275788467e-06, "loss": 0.3471, "step": 3826 }, { "epoch": 1.6921460951608742, "grad_norm": 0.2777645463593979, "learning_rate": 4.759599094480267e-06, "loss": 0.3304, "step": 3827 }, { "epoch": 1.6925883610363763, "grad_norm": 0.2857013438119835, "learning_rate": 4.757029976790066e-06, "loss": 0.3129, "step": 3828 }, { "epoch": 1.6930306269118787, "grad_norm": 0.33340989601411203, "learning_rate": 4.7544609233977326e-06, "loss": 0.3047, "step": 3829 }, { "epoch": 1.6934728927873808, "grad_norm": 0.3517584507887655, "learning_rate": 4.751891934983127e-06, "loss": 0.3054, "step": 3830 }, { "epoch": 1.693915158662883, "grad_norm": 0.3132549311434234, "learning_rate": 4.749323012226089e-06, "loss": 0.3123, "step": 3831 }, { "epoch": 1.694357424538385, "grad_norm": 0.35153746649225037, "learning_rate": 4.746754155806437e-06, "loss": 0.3432, "step": 3832 }, { "epoch": 1.6947996904138871, "grad_norm": 0.3063561112734395, "learning_rate": 4.744185366403977e-06, "loss": 0.3448, "step": 3833 }, { "epoch": 1.6952419562893892, "grad_norm": 0.26960118372270025, "learning_rate": 4.741616644698493e-06, "loss": 0.3298, "step": 3834 }, { "epoch": 1.6956842221648913, "grad_norm": 0.38383435152079, "learning_rate": 4.739047991369757e-06, "loss": 0.355, "step": 3835 }, { "epoch": 1.6961264880403935, "grad_norm": 0.28518804104864837, "learning_rate": 4.736479407097517e-06, "loss": 0.3305, "step": 3836 }, { "epoch": 1.6965687539158958, "grad_norm": 0.27241278387529694, "learning_rate": 4.7339108925615056e-06, "loss": 0.3111, "step": 3837 }, { "epoch": 1.697011019791398, "grad_norm": 0.3006329660248996, "learning_rate": 4.731342448441436e-06, "loss": 0.3108, "step": 3838 }, { "epoch": 1.6974532856669002, "grad_norm": 0.32196491140227834, "learning_rate": 4.728774075417002e-06, "loss": 0.3317, "step": 3839 }, { "epoch": 1.6978955515424023, "grad_norm": 0.29432158207515063, "learning_rate": 4.72620577416788e-06, "loss": 0.3018, "step": 3840 }, { "epoch": 1.6983378174179045, "grad_norm": 0.30694683163890424, "learning_rate": 4.7236375453737305e-06, "loss": 0.3461, "step": 3841 }, { "epoch": 1.6987800832934066, "grad_norm": 0.3195153421456572, "learning_rate": 4.721069389714188e-06, "loss": 0.3081, "step": 3842 }, { "epoch": 1.6992223491689087, "grad_norm": 0.3163755845215169, "learning_rate": 4.7185013078688716e-06, "loss": 0.3653, "step": 3843 }, { "epoch": 1.6996646150444108, "grad_norm": 0.28373810669392874, "learning_rate": 4.715933300517381e-06, "loss": 0.352, "step": 3844 }, { "epoch": 1.700106880919913, "grad_norm": 0.32402521601119555, "learning_rate": 4.7133653683392945e-06, "loss": 0.3249, "step": 3845 }, { "epoch": 1.700549146795415, "grad_norm": 0.301664355544197, "learning_rate": 4.710797512014173e-06, "loss": 0.3242, "step": 3846 }, { "epoch": 1.7009914126709174, "grad_norm": 0.3005165411391783, "learning_rate": 4.708229732221555e-06, "loss": 0.3346, "step": 3847 }, { "epoch": 1.7014336785464195, "grad_norm": 0.29842011574654076, "learning_rate": 4.705662029640959e-06, "loss": 0.3338, "step": 3848 }, { "epoch": 1.7018759444219218, "grad_norm": 0.26731012930611464, "learning_rate": 4.703094404951886e-06, "loss": 0.3306, "step": 3849 }, { "epoch": 1.702318210297424, "grad_norm": 0.27074533435702774, "learning_rate": 4.700526858833811e-06, "loss": 0.3497, "step": 3850 }, { "epoch": 1.702760476172926, "grad_norm": 0.4890631523386586, "learning_rate": 4.697959391966192e-06, "loss": 0.3408, "step": 3851 }, { "epoch": 1.7032027420484281, "grad_norm": 0.25190225633013275, "learning_rate": 4.695392005028468e-06, "loss": 0.3519, "step": 3852 }, { "epoch": 1.7036450079239303, "grad_norm": 0.30089692290595327, "learning_rate": 4.692824698700052e-06, "loss": 0.3228, "step": 3853 }, { "epoch": 1.7040872737994324, "grad_norm": 0.3542837669808669, "learning_rate": 4.690257473660338e-06, "loss": 0.3048, "step": 3854 }, { "epoch": 1.7045295396749345, "grad_norm": 0.30738402617065463, "learning_rate": 4.6876903305886975e-06, "loss": 0.3008, "step": 3855 }, { "epoch": 1.7049718055504366, "grad_norm": 0.30977779602385397, "learning_rate": 4.685123270164481e-06, "loss": 0.3493, "step": 3856 }, { "epoch": 1.705414071425939, "grad_norm": 0.3356720442553052, "learning_rate": 4.68255629306702e-06, "loss": 0.364, "step": 3857 }, { "epoch": 1.705856337301441, "grad_norm": 0.3011248794854905, "learning_rate": 4.679989399975619e-06, "loss": 0.3204, "step": 3858 }, { "epoch": 1.7062986031769432, "grad_norm": 0.30643026874671614, "learning_rate": 4.677422591569562e-06, "loss": 0.3677, "step": 3859 }, { "epoch": 1.7067408690524455, "grad_norm": 0.4745229164735908, "learning_rate": 4.674855868528112e-06, "loss": 0.3395, "step": 3860 }, { "epoch": 1.7071831349279476, "grad_norm": 0.32334702033041246, "learning_rate": 4.672289231530507e-06, "loss": 0.34, "step": 3861 }, { "epoch": 1.7076254008034497, "grad_norm": 0.3606051138768803, "learning_rate": 4.6697226812559665e-06, "loss": 0.3712, "step": 3862 }, { "epoch": 1.7080676666789518, "grad_norm": 0.26990683757084133, "learning_rate": 4.667156218383682e-06, "loss": 0.3358, "step": 3863 }, { "epoch": 1.708509932554454, "grad_norm": 0.3408484121414673, "learning_rate": 4.664589843592826e-06, "loss": 0.3861, "step": 3864 }, { "epoch": 1.708952198429956, "grad_norm": 0.31751532603643845, "learning_rate": 4.662023557562545e-06, "loss": 0.3097, "step": 3865 }, { "epoch": 1.7093944643054582, "grad_norm": 0.3561027545710492, "learning_rate": 4.659457360971959e-06, "loss": 0.3592, "step": 3866 }, { "epoch": 1.7098367301809605, "grad_norm": 0.37874814856737127, "learning_rate": 4.656891254500174e-06, "loss": 0.351, "step": 3867 }, { "epoch": 1.7102789960564626, "grad_norm": 0.31161960398765803, "learning_rate": 4.654325238826264e-06, "loss": 0.3554, "step": 3868 }, { "epoch": 1.7107212619319647, "grad_norm": 0.37904951177536406, "learning_rate": 4.651759314629282e-06, "loss": 0.3436, "step": 3869 }, { "epoch": 1.711163527807467, "grad_norm": 0.2937943350840743, "learning_rate": 4.6491934825882535e-06, "loss": 0.3277, "step": 3870 }, { "epoch": 1.7116057936829692, "grad_norm": 0.3395821858698644, "learning_rate": 4.646627743382182e-06, "loss": 0.3189, "step": 3871 }, { "epoch": 1.7120480595584713, "grad_norm": 0.23385959881401483, "learning_rate": 4.644062097690052e-06, "loss": 0.3238, "step": 3872 }, { "epoch": 1.7124903254339734, "grad_norm": 0.28434418372223846, "learning_rate": 4.641496546190813e-06, "loss": 0.3382, "step": 3873 }, { "epoch": 1.7129325913094755, "grad_norm": 0.3262524440481284, "learning_rate": 4.638931089563396e-06, "loss": 0.3461, "step": 3874 }, { "epoch": 1.7133748571849776, "grad_norm": 0.3158831491751921, "learning_rate": 4.636365728486704e-06, "loss": 0.3085, "step": 3875 }, { "epoch": 1.7138171230604797, "grad_norm": 0.29211014286523757, "learning_rate": 4.6338004636396175e-06, "loss": 0.342, "step": 3876 }, { "epoch": 1.714259388935982, "grad_norm": 0.3580406161036265, "learning_rate": 4.631235295700989e-06, "loss": 0.3324, "step": 3877 }, { "epoch": 1.7147016548114842, "grad_norm": 0.3645727249929362, "learning_rate": 4.628670225349648e-06, "loss": 0.3145, "step": 3878 }, { "epoch": 1.7151439206869863, "grad_norm": 0.3156122552353248, "learning_rate": 4.626105253264395e-06, "loss": 0.3324, "step": 3879 }, { "epoch": 1.7155861865624886, "grad_norm": 0.35161718375589346, "learning_rate": 4.6235403801240075e-06, "loss": 0.355, "step": 3880 }, { "epoch": 1.7160284524379907, "grad_norm": 0.41269850155646787, "learning_rate": 4.6209756066072355e-06, "loss": 0.3371, "step": 3881 }, { "epoch": 1.7164707183134928, "grad_norm": 0.4879343303082417, "learning_rate": 4.618410933392798e-06, "loss": 0.3247, "step": 3882 }, { "epoch": 1.716912984188995, "grad_norm": 0.3627514870920198, "learning_rate": 4.615846361159399e-06, "loss": 0.3799, "step": 3883 }, { "epoch": 1.717355250064497, "grad_norm": 0.5321125073304871, "learning_rate": 4.613281890585705e-06, "loss": 0.3925, "step": 3884 }, { "epoch": 1.7177975159399992, "grad_norm": 0.3803487069936657, "learning_rate": 4.6107175223503595e-06, "loss": 0.3532, "step": 3885 }, { "epoch": 1.7182397818155013, "grad_norm": 0.3963763999659194, "learning_rate": 4.60815325713198e-06, "loss": 0.3553, "step": 3886 }, { "epoch": 1.7186820476910034, "grad_norm": 0.3190934371256428, "learning_rate": 4.6055890956091526e-06, "loss": 0.3365, "step": 3887 }, { "epoch": 1.7191243135665057, "grad_norm": 0.3096576938590307, "learning_rate": 4.603025038460442e-06, "loss": 0.2895, "step": 3888 }, { "epoch": 1.7195665794420079, "grad_norm": 0.2558405038025071, "learning_rate": 4.600461086364382e-06, "loss": 0.3111, "step": 3889 }, { "epoch": 1.7200088453175102, "grad_norm": 0.28169188064863776, "learning_rate": 4.597897239999477e-06, "loss": 0.3295, "step": 3890 }, { "epoch": 1.7204511111930123, "grad_norm": 0.2571694145226686, "learning_rate": 4.5953335000442065e-06, "loss": 0.2962, "step": 3891 }, { "epoch": 1.7208933770685144, "grad_norm": 0.43471264059373027, "learning_rate": 4.59276986717702e-06, "loss": 0.3183, "step": 3892 }, { "epoch": 1.7213356429440165, "grad_norm": 0.30505720629794314, "learning_rate": 4.590206342076339e-06, "loss": 0.3462, "step": 3893 }, { "epoch": 1.7217779088195186, "grad_norm": 0.26228012479688234, "learning_rate": 4.5876429254205575e-06, "loss": 0.3174, "step": 3894 }, { "epoch": 1.7222201746950208, "grad_norm": 0.33124496492285627, "learning_rate": 4.585079617888039e-06, "loss": 0.3454, "step": 3895 }, { "epoch": 1.7226624405705229, "grad_norm": 0.24974390985056288, "learning_rate": 4.582516420157122e-06, "loss": 0.3311, "step": 3896 }, { "epoch": 1.723104706446025, "grad_norm": 0.3287954415445833, "learning_rate": 4.579953332906108e-06, "loss": 0.3396, "step": 3897 }, { "epoch": 1.7235469723215273, "grad_norm": 0.31787587373931636, "learning_rate": 4.577390356813276e-06, "loss": 0.3448, "step": 3898 }, { "epoch": 1.7239892381970294, "grad_norm": 0.3395046525132779, "learning_rate": 4.574827492556876e-06, "loss": 0.3348, "step": 3899 }, { "epoch": 1.7244315040725318, "grad_norm": 0.3589645971456806, "learning_rate": 4.572264740815127e-06, "loss": 0.3222, "step": 3900 }, { "epoch": 1.7248737699480339, "grad_norm": 0.28591380434303804, "learning_rate": 4.569702102266214e-06, "loss": 0.316, "step": 3901 }, { "epoch": 1.725316035823536, "grad_norm": 0.3432258163754354, "learning_rate": 4.567139577588298e-06, "loss": 0.3177, "step": 3902 }, { "epoch": 1.725758301699038, "grad_norm": 0.32224165736517657, "learning_rate": 4.5645771674595045e-06, "loss": 0.2984, "step": 3903 }, { "epoch": 1.7262005675745402, "grad_norm": 0.2912029361940647, "learning_rate": 4.562014872557936e-06, "loss": 0.2967, "step": 3904 }, { "epoch": 1.7266428334500423, "grad_norm": 0.29349478508397336, "learning_rate": 4.559452693561657e-06, "loss": 0.3231, "step": 3905 }, { "epoch": 1.7270850993255444, "grad_norm": 0.2732342639537418, "learning_rate": 4.556890631148705e-06, "loss": 0.3342, "step": 3906 }, { "epoch": 1.7275273652010465, "grad_norm": 0.2767482885522378, "learning_rate": 4.554328685997087e-06, "loss": 0.3054, "step": 3907 }, { "epoch": 1.7279696310765489, "grad_norm": 0.2704909920291074, "learning_rate": 4.551766858784776e-06, "loss": 0.3227, "step": 3908 }, { "epoch": 1.728411896952051, "grad_norm": 0.4831296587990118, "learning_rate": 4.549205150189719e-06, "loss": 0.3706, "step": 3909 }, { "epoch": 1.728854162827553, "grad_norm": 0.2980633806267141, "learning_rate": 4.546643560889827e-06, "loss": 0.3007, "step": 3910 }, { "epoch": 1.7292964287030554, "grad_norm": 0.31833970598286654, "learning_rate": 4.54408209156298e-06, "loss": 0.3201, "step": 3911 }, { "epoch": 1.7297386945785576, "grad_norm": 0.29463192729851345, "learning_rate": 4.54152074288703e-06, "loss": 0.337, "step": 3912 }, { "epoch": 1.7301809604540597, "grad_norm": 0.36149573894390535, "learning_rate": 4.53895951553979e-06, "loss": 0.3859, "step": 3913 }, { "epoch": 1.7306232263295618, "grad_norm": 0.28501428583314314, "learning_rate": 4.536398410199045e-06, "loss": 0.3335, "step": 3914 }, { "epoch": 1.731065492205064, "grad_norm": 0.33556833858639884, "learning_rate": 4.533837427542554e-06, "loss": 0.344, "step": 3915 }, { "epoch": 1.731507758080566, "grad_norm": 0.34127107000035406, "learning_rate": 4.531276568248032e-06, "loss": 0.3153, "step": 3916 }, { "epoch": 1.7319500239560681, "grad_norm": 0.3603247318676884, "learning_rate": 4.528715832993168e-06, "loss": 0.3327, "step": 3917 }, { "epoch": 1.7323922898315705, "grad_norm": 0.30951262218235426, "learning_rate": 4.526155222455617e-06, "loss": 0.3037, "step": 3918 }, { "epoch": 1.7328345557070726, "grad_norm": 0.3555929339552259, "learning_rate": 4.523594737313e-06, "loss": 0.3495, "step": 3919 }, { "epoch": 1.7332768215825747, "grad_norm": 0.31727548042638226, "learning_rate": 4.5210343782429074e-06, "loss": 0.3111, "step": 3920 }, { "epoch": 1.733719087458077, "grad_norm": 0.3681385231940353, "learning_rate": 4.518474145922895e-06, "loss": 0.3301, "step": 3921 }, { "epoch": 1.7341613533335791, "grad_norm": 0.5260348771186796, "learning_rate": 4.515914041030481e-06, "loss": 0.3291, "step": 3922 }, { "epoch": 1.7346036192090812, "grad_norm": 0.2880185983359292, "learning_rate": 4.513354064243157e-06, "loss": 0.3358, "step": 3923 }, { "epoch": 1.7350458850845834, "grad_norm": 0.31282582796486896, "learning_rate": 4.510794216238375e-06, "loss": 0.3031, "step": 3924 }, { "epoch": 1.7354881509600855, "grad_norm": 0.3751454445274682, "learning_rate": 4.508234497693557e-06, "loss": 0.3133, "step": 3925 }, { "epoch": 1.7359304168355876, "grad_norm": 0.3383838855746397, "learning_rate": 4.505674909286088e-06, "loss": 0.3414, "step": 3926 }, { "epoch": 1.7363726827110897, "grad_norm": 0.30044765521910416, "learning_rate": 4.5031154516933206e-06, "loss": 0.3382, "step": 3927 }, { "epoch": 1.736814948586592, "grad_norm": 0.3163614729540792, "learning_rate": 4.500556125592569e-06, "loss": 0.308, "step": 3928 }, { "epoch": 1.7372572144620941, "grad_norm": 0.2772619749350225, "learning_rate": 4.497996931661116e-06, "loss": 0.3239, "step": 3929 }, { "epoch": 1.7376994803375962, "grad_norm": 0.3360859696928452, "learning_rate": 4.495437870576207e-06, "loss": 0.3225, "step": 3930 }, { "epoch": 1.7381417462130986, "grad_norm": 0.34944401178859485, "learning_rate": 4.4928789430150596e-06, "loss": 0.3554, "step": 3931 }, { "epoch": 1.7385840120886007, "grad_norm": 0.29613169702720155, "learning_rate": 4.4903201496548456e-06, "loss": 0.3211, "step": 3932 }, { "epoch": 1.7390262779641028, "grad_norm": 0.3141705256129199, "learning_rate": 4.4877614911727064e-06, "loss": 0.3016, "step": 3933 }, { "epoch": 1.739468543839605, "grad_norm": 0.28286638564565686, "learning_rate": 4.4852029682457484e-06, "loss": 0.3126, "step": 3934 }, { "epoch": 1.739910809715107, "grad_norm": 0.3045557767744593, "learning_rate": 4.48264458155104e-06, "loss": 0.3152, "step": 3935 }, { "epoch": 1.7403530755906091, "grad_norm": 0.3239093620239275, "learning_rate": 4.4800863317656155e-06, "loss": 0.329, "step": 3936 }, { "epoch": 1.7407953414661113, "grad_norm": 0.3486292208055818, "learning_rate": 4.477528219566471e-06, "loss": 0.3357, "step": 3937 }, { "epoch": 1.7412376073416134, "grad_norm": 0.29998287189994843, "learning_rate": 4.4749702456305685e-06, "loss": 0.3023, "step": 3938 }, { "epoch": 1.7416798732171157, "grad_norm": 0.29666834924719004, "learning_rate": 4.472412410634832e-06, "loss": 0.313, "step": 3939 }, { "epoch": 1.7421221390926178, "grad_norm": 0.37566259106106414, "learning_rate": 4.469854715256148e-06, "loss": 0.3302, "step": 3940 }, { "epoch": 1.7425644049681202, "grad_norm": 0.2838061078325033, "learning_rate": 4.467297160171367e-06, "loss": 0.354, "step": 3941 }, { "epoch": 1.7430066708436223, "grad_norm": 0.40583373715128357, "learning_rate": 4.464739746057304e-06, "loss": 0.3449, "step": 3942 }, { "epoch": 1.7434489367191244, "grad_norm": 0.34968740141250726, "learning_rate": 4.462182473590734e-06, "loss": 0.3073, "step": 3943 }, { "epoch": 1.7438912025946265, "grad_norm": 0.28870857377118647, "learning_rate": 4.4596253434483944e-06, "loss": 0.303, "step": 3944 }, { "epoch": 1.7443334684701286, "grad_norm": 0.2730949304829127, "learning_rate": 4.457068356306985e-06, "loss": 0.3315, "step": 3945 }, { "epoch": 1.7447757343456307, "grad_norm": 0.2897946500755897, "learning_rate": 4.4545115128431736e-06, "loss": 0.3051, "step": 3946 }, { "epoch": 1.7452180002211328, "grad_norm": 0.3123225748198406, "learning_rate": 4.4519548137335816e-06, "loss": 0.3289, "step": 3947 }, { "epoch": 1.745660266096635, "grad_norm": 0.4080652483245326, "learning_rate": 4.449398259654796e-06, "loss": 0.324, "step": 3948 }, { "epoch": 1.7461025319721373, "grad_norm": 0.2619084773632476, "learning_rate": 4.446841851283367e-06, "loss": 0.3163, "step": 3949 }, { "epoch": 1.7465447978476394, "grad_norm": 0.28958501701155265, "learning_rate": 4.444285589295802e-06, "loss": 0.3436, "step": 3950 }, { "epoch": 1.7469870637231417, "grad_norm": 0.31098912010171104, "learning_rate": 4.441729474368572e-06, "loss": 0.3161, "step": 3951 }, { "epoch": 1.7474293295986438, "grad_norm": 0.27162928771612466, "learning_rate": 4.439173507178113e-06, "loss": 0.3102, "step": 3952 }, { "epoch": 1.747871595474146, "grad_norm": 0.3217866286690114, "learning_rate": 4.436617688400814e-06, "loss": 0.327, "step": 3953 }, { "epoch": 1.748313861349648, "grad_norm": 0.31406330489908707, "learning_rate": 4.434062018713032e-06, "loss": 0.3264, "step": 3954 }, { "epoch": 1.7487561272251502, "grad_norm": 0.34334416908930737, "learning_rate": 4.43150649879108e-06, "loss": 0.346, "step": 3955 }, { "epoch": 1.7491983931006523, "grad_norm": 0.2786596143008142, "learning_rate": 4.428951129311229e-06, "loss": 0.3515, "step": 3956 }, { "epoch": 1.7496406589761544, "grad_norm": 0.2918469046719794, "learning_rate": 4.42639591094972e-06, "loss": 0.3596, "step": 3957 }, { "epoch": 1.7500829248516565, "grad_norm": 0.30302887879575047, "learning_rate": 4.4238408443827465e-06, "loss": 0.328, "step": 3958 }, { "epoch": 1.7505251907271588, "grad_norm": 0.2558556640953865, "learning_rate": 4.421285930286461e-06, "loss": 0.3309, "step": 3959 }, { "epoch": 1.750967456602661, "grad_norm": 0.44162068268167015, "learning_rate": 4.41873116933698e-06, "loss": 0.3637, "step": 3960 }, { "epoch": 1.7514097224781633, "grad_norm": 0.29878666593838044, "learning_rate": 4.416176562210373e-06, "loss": 0.2958, "step": 3961 }, { "epoch": 1.7518519883536654, "grad_norm": 0.28530217611450753, "learning_rate": 4.413622109582681e-06, "loss": 0.3012, "step": 3962 }, { "epoch": 1.7522942542291675, "grad_norm": 0.3191989200164971, "learning_rate": 4.41106781212989e-06, "loss": 0.3183, "step": 3963 }, { "epoch": 1.7527365201046696, "grad_norm": 0.25960123771521476, "learning_rate": 4.408513670527954e-06, "loss": 0.3238, "step": 3964 }, { "epoch": 1.7531787859801717, "grad_norm": 0.2988574024072548, "learning_rate": 4.405959685452782e-06, "loss": 0.324, "step": 3965 }, { "epoch": 1.7536210518556739, "grad_norm": 0.290510025248516, "learning_rate": 4.403405857580243e-06, "loss": 0.3056, "step": 3966 }, { "epoch": 1.754063317731176, "grad_norm": 0.3128497681276004, "learning_rate": 4.400852187586161e-06, "loss": 0.3312, "step": 3967 }, { "epoch": 1.754505583606678, "grad_norm": 0.29697631340166813, "learning_rate": 4.398298676146327e-06, "loss": 0.3364, "step": 3968 }, { "epoch": 1.7549478494821804, "grad_norm": 0.32258924028263714, "learning_rate": 4.395745323936481e-06, "loss": 0.3075, "step": 3969 }, { "epoch": 1.7553901153576825, "grad_norm": 0.39885946412265705, "learning_rate": 4.393192131632324e-06, "loss": 0.3381, "step": 3970 }, { "epoch": 1.7558323812331846, "grad_norm": 0.3201924603716079, "learning_rate": 4.390639099909515e-06, "loss": 0.321, "step": 3971 }, { "epoch": 1.756274647108687, "grad_norm": 0.31038201185236874, "learning_rate": 4.3880862294436685e-06, "loss": 0.306, "step": 3972 }, { "epoch": 1.756716912984189, "grad_norm": 0.30934674133621415, "learning_rate": 4.385533520910361e-06, "loss": 0.3443, "step": 3973 }, { "epoch": 1.7571591788596912, "grad_norm": 0.29575673483896814, "learning_rate": 4.382980974985124e-06, "loss": 0.2792, "step": 3974 }, { "epoch": 1.7576014447351933, "grad_norm": 0.3094008625919546, "learning_rate": 4.380428592343442e-06, "loss": 0.3293, "step": 3975 }, { "epoch": 1.7580437106106954, "grad_norm": 0.3331622907018014, "learning_rate": 4.377876373660761e-06, "loss": 0.3237, "step": 3976 }, { "epoch": 1.7584859764861975, "grad_norm": 0.268806399037961, "learning_rate": 4.375324319612481e-06, "loss": 0.3229, "step": 3977 }, { "epoch": 1.7589282423616996, "grad_norm": 0.30964401434631306, "learning_rate": 4.372772430873962e-06, "loss": 0.3084, "step": 3978 }, { "epoch": 1.759370508237202, "grad_norm": 0.2947776370156596, "learning_rate": 4.370220708120516e-06, "loss": 0.3068, "step": 3979 }, { "epoch": 1.759812774112704, "grad_norm": 0.4081731843175963, "learning_rate": 4.367669152027413e-06, "loss": 0.3789, "step": 3980 }, { "epoch": 1.7602550399882062, "grad_norm": 0.3042717153365785, "learning_rate": 4.36511776326988e-06, "loss": 0.2944, "step": 3981 }, { "epoch": 1.7606973058637085, "grad_norm": 0.3174748592310244, "learning_rate": 4.362566542523097e-06, "loss": 0.3032, "step": 3982 }, { "epoch": 1.7611395717392107, "grad_norm": 0.2914240808447158, "learning_rate": 4.3600154904622e-06, "loss": 0.3459, "step": 3983 }, { "epoch": 1.7615818376147128, "grad_norm": 0.27016937504249405, "learning_rate": 4.357464607762286e-06, "loss": 0.3042, "step": 3984 }, { "epoch": 1.7620241034902149, "grad_norm": 0.30522123040314336, "learning_rate": 4.354913895098398e-06, "loss": 0.2915, "step": 3985 }, { "epoch": 1.762466369365717, "grad_norm": 0.29153687203025214, "learning_rate": 4.352363353145541e-06, "loss": 0.3148, "step": 3986 }, { "epoch": 1.762908635241219, "grad_norm": 0.31043187772396436, "learning_rate": 4.349812982578672e-06, "loss": 0.3157, "step": 3987 }, { "epoch": 1.7633509011167212, "grad_norm": 0.33409511648074436, "learning_rate": 4.347262784072699e-06, "loss": 0.3315, "step": 3988 }, { "epoch": 1.7637931669922233, "grad_norm": 0.34711270363190433, "learning_rate": 4.344712758302496e-06, "loss": 0.324, "step": 3989 }, { "epoch": 1.7642354328677257, "grad_norm": 0.31365732925225076, "learning_rate": 4.342162905942879e-06, "loss": 0.3041, "step": 3990 }, { "epoch": 1.7646776987432278, "grad_norm": 0.26471920559290957, "learning_rate": 4.339613227668623e-06, "loss": 0.3413, "step": 3991 }, { "epoch": 1.76511996461873, "grad_norm": 0.36703462867649106, "learning_rate": 4.337063724154459e-06, "loss": 0.3137, "step": 3992 }, { "epoch": 1.7655622304942322, "grad_norm": 0.34267411841806156, "learning_rate": 4.3345143960750665e-06, "loss": 0.317, "step": 3993 }, { "epoch": 1.7660044963697343, "grad_norm": 0.26657071264142945, "learning_rate": 4.3319652441050845e-06, "loss": 0.3632, "step": 3994 }, { "epoch": 1.7664467622452364, "grad_norm": 0.31225176313526765, "learning_rate": 4.3294162689191025e-06, "loss": 0.3051, "step": 3995 }, { "epoch": 1.7668890281207386, "grad_norm": 0.3267186854388556, "learning_rate": 4.326867471191661e-06, "loss": 0.3154, "step": 3996 }, { "epoch": 1.7673312939962407, "grad_norm": 0.26400851971395345, "learning_rate": 4.3243188515972575e-06, "loss": 0.3474, "step": 3997 }, { "epoch": 1.7677735598717428, "grad_norm": 0.3364792194280267, "learning_rate": 4.321770410810339e-06, "loss": 0.3445, "step": 3998 }, { "epoch": 1.768215825747245, "grad_norm": 0.315383690022799, "learning_rate": 4.319222149505311e-06, "loss": 0.314, "step": 3999 }, { "epoch": 1.7686580916227472, "grad_norm": 0.3305856239443015, "learning_rate": 4.316674068356523e-06, "loss": 0.305, "step": 4000 }, { "epoch": 1.7691003574982493, "grad_norm": 0.2987499771296853, "learning_rate": 4.314126168038284e-06, "loss": 0.3419, "step": 4001 }, { "epoch": 1.7695426233737517, "grad_norm": 0.2730939211295197, "learning_rate": 4.311578449224852e-06, "loss": 0.3417, "step": 4002 }, { "epoch": 1.7699848892492538, "grad_norm": 0.30147167921344337, "learning_rate": 4.309030912590435e-06, "loss": 0.3214, "step": 4003 }, { "epoch": 1.770427155124756, "grad_norm": 0.3168237286946974, "learning_rate": 4.306483558809194e-06, "loss": 0.3263, "step": 4004 }, { "epoch": 1.770869421000258, "grad_norm": 0.31015776524142635, "learning_rate": 4.3039363885552494e-06, "loss": 0.3259, "step": 4005 }, { "epoch": 1.7713116868757601, "grad_norm": 0.3409713010484931, "learning_rate": 4.30138940250266e-06, "loss": 0.3411, "step": 4006 }, { "epoch": 1.7717539527512622, "grad_norm": 0.32069198741663796, "learning_rate": 4.298842601325446e-06, "loss": 0.3025, "step": 4007 }, { "epoch": 1.7721962186267644, "grad_norm": 0.2750001445349048, "learning_rate": 4.296295985697572e-06, "loss": 0.3042, "step": 4008 }, { "epoch": 1.7726384845022665, "grad_norm": 0.29100677696502597, "learning_rate": 4.293749556292957e-06, "loss": 0.3369, "step": 4009 }, { "epoch": 1.7730807503777688, "grad_norm": 0.29298498980509297, "learning_rate": 4.291203313785472e-06, "loss": 0.3212, "step": 4010 }, { "epoch": 1.773523016253271, "grad_norm": 0.28883531373138693, "learning_rate": 4.288657258848935e-06, "loss": 0.3416, "step": 4011 }, { "epoch": 1.7739652821287732, "grad_norm": 0.271629599077632, "learning_rate": 4.286111392157117e-06, "loss": 0.2873, "step": 4012 }, { "epoch": 1.7744075480042754, "grad_norm": 0.37471524697938596, "learning_rate": 4.283565714383736e-06, "loss": 0.2968, "step": 4013 }, { "epoch": 1.7748498138797775, "grad_norm": 0.33377751881600226, "learning_rate": 4.281020226202465e-06, "loss": 0.3166, "step": 4014 }, { "epoch": 1.7752920797552796, "grad_norm": 0.3215655181089847, "learning_rate": 4.278474928286922e-06, "loss": 0.3348, "step": 4015 }, { "epoch": 1.7757343456307817, "grad_norm": 0.25994822549750446, "learning_rate": 4.275929821310678e-06, "loss": 0.3483, "step": 4016 }, { "epoch": 1.7761766115062838, "grad_norm": 0.37491631620774063, "learning_rate": 4.273384905947253e-06, "loss": 0.337, "step": 4017 }, { "epoch": 1.776618877381786, "grad_norm": 0.2979339474228781, "learning_rate": 4.270840182870113e-06, "loss": 0.3582, "step": 4018 }, { "epoch": 1.777061143257288, "grad_norm": 0.30378923025051635, "learning_rate": 4.268295652752676e-06, "loss": 0.3168, "step": 4019 }, { "epoch": 1.7775034091327904, "grad_norm": 0.31502150158675324, "learning_rate": 4.265751316268308e-06, "loss": 0.3112, "step": 4020 }, { "epoch": 1.7779456750082925, "grad_norm": 0.2727123805687449, "learning_rate": 4.263207174090326e-06, "loss": 0.3735, "step": 4021 }, { "epoch": 1.7783879408837946, "grad_norm": 0.3039715761611664, "learning_rate": 4.260663226891993e-06, "loss": 0.3093, "step": 4022 }, { "epoch": 1.778830206759297, "grad_norm": 0.27555702505575536, "learning_rate": 4.258119475346522e-06, "loss": 0.3044, "step": 4023 }, { "epoch": 1.779272472634799, "grad_norm": 0.28081605288447864, "learning_rate": 4.255575920127072e-06, "loss": 0.2979, "step": 4024 }, { "epoch": 1.7797147385103012, "grad_norm": 0.27518697234173706, "learning_rate": 4.253032561906752e-06, "loss": 0.3212, "step": 4025 }, { "epoch": 1.7801570043858033, "grad_norm": 0.35435073370531245, "learning_rate": 4.250489401358619e-06, "loss": 0.339, "step": 4026 }, { "epoch": 1.7805992702613054, "grad_norm": 0.3443700627814528, "learning_rate": 4.247946439155677e-06, "loss": 0.332, "step": 4027 }, { "epoch": 1.7810415361368075, "grad_norm": 0.30220181196231083, "learning_rate": 4.245403675970877e-06, "loss": 0.3425, "step": 4028 }, { "epoch": 1.7814838020123096, "grad_norm": 0.28167431246464625, "learning_rate": 4.2428611124771184e-06, "loss": 0.3279, "step": 4029 }, { "epoch": 1.781926067887812, "grad_norm": 0.31997890319997824, "learning_rate": 4.2403187493472446e-06, "loss": 0.322, "step": 4030 }, { "epoch": 1.782368333763314, "grad_norm": 0.34447797416365117, "learning_rate": 4.2377765872540525e-06, "loss": 0.3187, "step": 4031 }, { "epoch": 1.7828105996388162, "grad_norm": 0.3217827679001557, "learning_rate": 4.23523462687028e-06, "loss": 0.3134, "step": 4032 }, { "epoch": 1.7832528655143185, "grad_norm": 0.3116402530661999, "learning_rate": 4.2326928688686155e-06, "loss": 0.3015, "step": 4033 }, { "epoch": 1.7836951313898206, "grad_norm": 0.3174556056967228, "learning_rate": 4.230151313921688e-06, "loss": 0.3493, "step": 4034 }, { "epoch": 1.7841373972653227, "grad_norm": 0.337968639057394, "learning_rate": 4.227609962702079e-06, "loss": 0.3019, "step": 4035 }, { "epoch": 1.7845796631408248, "grad_norm": 0.28955392990095735, "learning_rate": 4.225068815882316e-06, "loss": 0.3178, "step": 4036 }, { "epoch": 1.785021929016327, "grad_norm": 0.3643648610853629, "learning_rate": 4.222527874134865e-06, "loss": 0.3388, "step": 4037 }, { "epoch": 1.785464194891829, "grad_norm": 0.3030609275611852, "learning_rate": 4.219987138132148e-06, "loss": 0.3175, "step": 4038 }, { "epoch": 1.7859064607673312, "grad_norm": 0.28393747171942946, "learning_rate": 4.217446608546524e-06, "loss": 0.3343, "step": 4039 }, { "epoch": 1.7863487266428333, "grad_norm": 0.3617278359770635, "learning_rate": 4.214906286050301e-06, "loss": 0.3601, "step": 4040 }, { "epoch": 1.7867909925183356, "grad_norm": 0.2923397433190261, "learning_rate": 4.212366171315733e-06, "loss": 0.331, "step": 4041 }, { "epoch": 1.7872332583938377, "grad_norm": 0.3474236105709054, "learning_rate": 4.2098262650150185e-06, "loss": 0.3779, "step": 4042 }, { "epoch": 1.78767552426934, "grad_norm": 0.32216537488778435, "learning_rate": 4.2072865678203e-06, "loss": 0.3306, "step": 4043 }, { "epoch": 1.7881177901448422, "grad_norm": 0.2787480991390978, "learning_rate": 4.204747080403665e-06, "loss": 0.3365, "step": 4044 }, { "epoch": 1.7885600560203443, "grad_norm": 0.31742132521169786, "learning_rate": 4.202207803437146e-06, "loss": 0.3516, "step": 4045 }, { "epoch": 1.7890023218958464, "grad_norm": 0.33679873224639456, "learning_rate": 4.199668737592715e-06, "loss": 0.3363, "step": 4046 }, { "epoch": 1.7894445877713485, "grad_norm": 0.3212950014113431, "learning_rate": 4.197129883542299e-06, "loss": 0.3502, "step": 4047 }, { "epoch": 1.7898868536468506, "grad_norm": 0.30466350311273627, "learning_rate": 4.19459124195776e-06, "loss": 0.3234, "step": 4048 }, { "epoch": 1.7903291195223527, "grad_norm": 0.337733658285673, "learning_rate": 4.192052813510905e-06, "loss": 0.3253, "step": 4049 }, { "epoch": 1.7907713853978549, "grad_norm": 0.3042618250802373, "learning_rate": 4.189514598873486e-06, "loss": 0.301, "step": 4050 }, { "epoch": 1.7912136512733572, "grad_norm": 0.2956553282330182, "learning_rate": 4.186976598717198e-06, "loss": 0.3242, "step": 4051 }, { "epoch": 1.7916559171488593, "grad_norm": 0.36642772807144147, "learning_rate": 4.1844388137136805e-06, "loss": 0.3427, "step": 4052 }, { "epoch": 1.7920981830243616, "grad_norm": 0.29253818976780743, "learning_rate": 4.1819012445345155e-06, "loss": 0.3303, "step": 4053 }, { "epoch": 1.7925404488998637, "grad_norm": 0.3137232248135909, "learning_rate": 4.1793638918512255e-06, "loss": 0.2589, "step": 4054 }, { "epoch": 1.7929827147753659, "grad_norm": 0.2889662912216577, "learning_rate": 4.176826756335279e-06, "loss": 0.3387, "step": 4055 }, { "epoch": 1.793424980650868, "grad_norm": 0.2788312274122398, "learning_rate": 4.174289838658085e-06, "loss": 0.3419, "step": 4056 }, { "epoch": 1.79386724652637, "grad_norm": 0.2917198102522168, "learning_rate": 4.1717531394909945e-06, "loss": 0.3067, "step": 4057 }, { "epoch": 1.7943095124018722, "grad_norm": 0.2641918229251641, "learning_rate": 4.169216659505304e-06, "loss": 0.2693, "step": 4058 }, { "epoch": 1.7947517782773743, "grad_norm": 0.27218353710061244, "learning_rate": 4.166680399372248e-06, "loss": 0.3055, "step": 4059 }, { "epoch": 1.7951940441528764, "grad_norm": 0.3339588265462104, "learning_rate": 4.164144359763006e-06, "loss": 0.3388, "step": 4060 }, { "epoch": 1.7956363100283788, "grad_norm": 0.29647752587823917, "learning_rate": 4.161608541348696e-06, "loss": 0.3304, "step": 4061 }, { "epoch": 1.7960785759038809, "grad_norm": 0.3640168859437094, "learning_rate": 4.159072944800377e-06, "loss": 0.3079, "step": 4062 }, { "epoch": 1.7965208417793832, "grad_norm": 0.35452659598547304, "learning_rate": 4.156537570789056e-06, "loss": 0.3233, "step": 4063 }, { "epoch": 1.7969631076548853, "grad_norm": 0.38030368557802247, "learning_rate": 4.154002419985675e-06, "loss": 0.3304, "step": 4064 }, { "epoch": 1.7974053735303874, "grad_norm": 0.34734069658233735, "learning_rate": 4.151467493061118e-06, "loss": 0.3221, "step": 4065 }, { "epoch": 1.7978476394058895, "grad_norm": 0.2618258164957656, "learning_rate": 4.148932790686209e-06, "loss": 0.3144, "step": 4066 }, { "epoch": 1.7982899052813917, "grad_norm": 0.3338751667468583, "learning_rate": 4.146398313531715e-06, "loss": 0.3155, "step": 4067 }, { "epoch": 1.7987321711568938, "grad_norm": 0.3713612675577529, "learning_rate": 4.143864062268341e-06, "loss": 0.3185, "step": 4068 }, { "epoch": 1.7991744370323959, "grad_norm": 0.36955511257942675, "learning_rate": 4.1413300375667346e-06, "loss": 0.3263, "step": 4069 }, { "epoch": 1.799616702907898, "grad_norm": 0.29175786375324647, "learning_rate": 4.1387962400974825e-06, "loss": 0.3176, "step": 4070 }, { "epoch": 1.8000589687834003, "grad_norm": 0.33049107169993686, "learning_rate": 4.136262670531109e-06, "loss": 0.3242, "step": 4071 }, { "epoch": 1.8005012346589024, "grad_norm": 0.41637254342842694, "learning_rate": 4.133729329538082e-06, "loss": 0.3281, "step": 4072 }, { "epoch": 1.8009435005344046, "grad_norm": 0.29297289850936575, "learning_rate": 4.131196217788806e-06, "loss": 0.3185, "step": 4073 }, { "epoch": 1.8013857664099069, "grad_norm": 0.3040562464113474, "learning_rate": 4.128663335953627e-06, "loss": 0.3142, "step": 4074 }, { "epoch": 1.801828032285409, "grad_norm": 0.32219931282482994, "learning_rate": 4.1261306847028265e-06, "loss": 0.323, "step": 4075 }, { "epoch": 1.802270298160911, "grad_norm": 0.2802788106693779, "learning_rate": 4.123598264706631e-06, "loss": 0.333, "step": 4076 }, { "epoch": 1.8027125640364132, "grad_norm": 0.23288626039093932, "learning_rate": 4.121066076635199e-06, "loss": 0.3246, "step": 4077 }, { "epoch": 1.8031548299119153, "grad_norm": 0.30769365124031384, "learning_rate": 4.11853412115863e-06, "loss": 0.3067, "step": 4078 }, { "epoch": 1.8035970957874174, "grad_norm": 0.29332807401062766, "learning_rate": 4.116002398946968e-06, "loss": 0.3133, "step": 4079 }, { "epoch": 1.8040393616629196, "grad_norm": 0.2717435104596825, "learning_rate": 4.113470910670186e-06, "loss": 0.321, "step": 4080 }, { "epoch": 1.804481627538422, "grad_norm": 0.2665689585774161, "learning_rate": 4.110939656998199e-06, "loss": 0.3151, "step": 4081 }, { "epoch": 1.804923893413924, "grad_norm": 0.3062702678634196, "learning_rate": 4.108408638600863e-06, "loss": 0.3362, "step": 4082 }, { "epoch": 1.8053661592894261, "grad_norm": 0.3276402245630096, "learning_rate": 4.105877856147965e-06, "loss": 0.3442, "step": 4083 }, { "epoch": 1.8058084251649285, "grad_norm": 0.33174695726426456, "learning_rate": 4.1033473103092365e-06, "loss": 0.351, "step": 4084 }, { "epoch": 1.8062506910404306, "grad_norm": 0.3514064874310781, "learning_rate": 4.100817001754343e-06, "loss": 0.347, "step": 4085 }, { "epoch": 1.8066929569159327, "grad_norm": 0.3912742981734132, "learning_rate": 4.0982869311528856e-06, "loss": 0.35, "step": 4086 }, { "epoch": 1.8071352227914348, "grad_norm": 0.2631948429173992, "learning_rate": 4.095757099174406e-06, "loss": 0.336, "step": 4087 }, { "epoch": 1.807577488666937, "grad_norm": 0.27609969687493285, "learning_rate": 4.093227506488379e-06, "loss": 0.3085, "step": 4088 }, { "epoch": 1.808019754542439, "grad_norm": 0.30775152503141084, "learning_rate": 4.090698153764221e-06, "loss": 0.3349, "step": 4089 }, { "epoch": 1.8084620204179411, "grad_norm": 0.2902330643105185, "learning_rate": 4.08816904167128e-06, "loss": 0.3466, "step": 4090 }, { "epoch": 1.8089042862934432, "grad_norm": 0.3934897247817246, "learning_rate": 4.085640170878846e-06, "loss": 0.3467, "step": 4091 }, { "epoch": 1.8093465521689456, "grad_norm": 0.3201592768688444, "learning_rate": 4.083111542056136e-06, "loss": 0.3292, "step": 4092 }, { "epoch": 1.8097888180444477, "grad_norm": 0.3462174178223063, "learning_rate": 4.0805831558723105e-06, "loss": 0.3392, "step": 4093 }, { "epoch": 1.81023108391995, "grad_norm": 0.2762090973492716, "learning_rate": 4.078055012996463e-06, "loss": 0.329, "step": 4094 }, { "epoch": 1.8106733497954521, "grad_norm": 0.31337378394003407, "learning_rate": 4.075527114097627e-06, "loss": 0.3419, "step": 4095 }, { "epoch": 1.8111156156709542, "grad_norm": 0.3443977975277074, "learning_rate": 4.072999459844765e-06, "loss": 0.3689, "step": 4096 }, { "epoch": 1.8115578815464564, "grad_norm": 0.2701645344314619, "learning_rate": 4.070472050906778e-06, "loss": 0.3199, "step": 4097 }, { "epoch": 1.8120001474219585, "grad_norm": 0.2914405508402354, "learning_rate": 4.067944887952502e-06, "loss": 0.3222, "step": 4098 }, { "epoch": 1.8124424132974606, "grad_norm": 0.2648549130230038, "learning_rate": 4.065417971650705e-06, "loss": 0.3039, "step": 4099 }, { "epoch": 1.8128846791729627, "grad_norm": 0.2864434365237143, "learning_rate": 4.062891302670097e-06, "loss": 0.326, "step": 4100 }, { "epoch": 1.8133269450484648, "grad_norm": 0.32415702034954347, "learning_rate": 4.060364881679314e-06, "loss": 0.3357, "step": 4101 }, { "epoch": 1.8137692109239671, "grad_norm": 0.2988992802801565, "learning_rate": 4.057838709346932e-06, "loss": 0.3166, "step": 4102 }, { "epoch": 1.8142114767994693, "grad_norm": 0.25887361337079806, "learning_rate": 4.055312786341459e-06, "loss": 0.3257, "step": 4103 }, { "epoch": 1.8146537426749716, "grad_norm": 0.2695143402063767, "learning_rate": 4.052787113331335e-06, "loss": 0.3266, "step": 4104 }, { "epoch": 1.8150960085504737, "grad_norm": 0.31033373331553316, "learning_rate": 4.0502616909849415e-06, "loss": 0.3424, "step": 4105 }, { "epoch": 1.8155382744259758, "grad_norm": 0.307294379354953, "learning_rate": 4.047736519970586e-06, "loss": 0.3426, "step": 4106 }, { "epoch": 1.815980540301478, "grad_norm": 0.34429598250500554, "learning_rate": 4.045211600956511e-06, "loss": 0.3611, "step": 4107 }, { "epoch": 1.81642280617698, "grad_norm": 0.3370861562586942, "learning_rate": 4.0426869346108925e-06, "loss": 0.3458, "step": 4108 }, { "epoch": 1.8168650720524822, "grad_norm": 0.3459999181369897, "learning_rate": 4.04016252160184e-06, "loss": 0.308, "step": 4109 }, { "epoch": 1.8173073379279843, "grad_norm": 0.2884297641175724, "learning_rate": 4.037638362597401e-06, "loss": 0.3193, "step": 4110 }, { "epoch": 1.8177496038034864, "grad_norm": 0.28824514916261257, "learning_rate": 4.035114458265547e-06, "loss": 0.339, "step": 4111 }, { "epoch": 1.8181918696789887, "grad_norm": 0.27445162423979225, "learning_rate": 4.032590809274187e-06, "loss": 0.3176, "step": 4112 }, { "epoch": 1.8186341355544908, "grad_norm": 0.4349469756634812, "learning_rate": 4.030067416291163e-06, "loss": 0.364, "step": 4113 }, { "epoch": 1.8190764014299932, "grad_norm": 0.244111172442544, "learning_rate": 4.027544279984245e-06, "loss": 0.2705, "step": 4114 }, { "epoch": 1.8195186673054953, "grad_norm": 0.319157720800852, "learning_rate": 4.025021401021139e-06, "loss": 0.3162, "step": 4115 }, { "epoch": 1.8199609331809974, "grad_norm": 0.30496958072464786, "learning_rate": 4.022498780069484e-06, "loss": 0.3133, "step": 4116 }, { "epoch": 1.8204031990564995, "grad_norm": 0.26884285689933307, "learning_rate": 4.019976417796848e-06, "loss": 0.3085, "step": 4117 }, { "epoch": 1.8208454649320016, "grad_norm": 0.30178012425391976, "learning_rate": 4.017454314870729e-06, "loss": 0.3238, "step": 4118 }, { "epoch": 1.8212877308075037, "grad_norm": 0.2734724264308239, "learning_rate": 4.014932471958562e-06, "loss": 0.2738, "step": 4119 }, { "epoch": 1.8217299966830058, "grad_norm": 0.3241742479815345, "learning_rate": 4.012410889727703e-06, "loss": 0.3489, "step": 4120 }, { "epoch": 1.822172262558508, "grad_norm": 0.3131528799067561, "learning_rate": 4.009889568845453e-06, "loss": 0.3243, "step": 4121 }, { "epoch": 1.8226145284340103, "grad_norm": 0.2759375653170965, "learning_rate": 4.0073685099790345e-06, "loss": 0.3097, "step": 4122 }, { "epoch": 1.8230567943095124, "grad_norm": 0.2775455042921857, "learning_rate": 4.004847713795602e-06, "loss": 0.2932, "step": 4123 }, { "epoch": 1.8234990601850145, "grad_norm": 0.2588804223166069, "learning_rate": 4.00232718096224e-06, "loss": 0.3289, "step": 4124 }, { "epoch": 1.8239413260605168, "grad_norm": 0.35295790071322786, "learning_rate": 3.999806912145964e-06, "loss": 0.3376, "step": 4125 }, { "epoch": 1.824383591936019, "grad_norm": 0.33045618099425766, "learning_rate": 3.997286908013725e-06, "loss": 0.2907, "step": 4126 }, { "epoch": 1.824825857811521, "grad_norm": 0.33633784090441576, "learning_rate": 3.994767169232395e-06, "loss": 0.3255, "step": 4127 }, { "epoch": 1.8252681236870232, "grad_norm": 0.40682603944470414, "learning_rate": 3.992247696468781e-06, "loss": 0.3515, "step": 4128 }, { "epoch": 1.8257103895625253, "grad_norm": 0.30435764583266606, "learning_rate": 3.989728490389617e-06, "loss": 0.3406, "step": 4129 }, { "epoch": 1.8261526554380274, "grad_norm": 0.3412000522658064, "learning_rate": 3.98720955166157e-06, "loss": 0.305, "step": 4130 }, { "epoch": 1.8265949213135295, "grad_norm": 0.27751798493972385, "learning_rate": 3.9846908809512325e-06, "loss": 0.295, "step": 4131 }, { "epoch": 1.8270371871890319, "grad_norm": 0.29423246987930685, "learning_rate": 3.9821724789251295e-06, "loss": 0.322, "step": 4132 }, { "epoch": 1.827479453064534, "grad_norm": 0.30490153786397145, "learning_rate": 3.979654346249712e-06, "loss": 0.33, "step": 4133 }, { "epoch": 1.827921718940036, "grad_norm": 0.36142857820426644, "learning_rate": 3.977136483591361e-06, "loss": 0.3499, "step": 4134 }, { "epoch": 1.8283639848155384, "grad_norm": 0.24370537076998713, "learning_rate": 3.974618891616387e-06, "loss": 0.323, "step": 4135 }, { "epoch": 1.8288062506910405, "grad_norm": 0.28770089950821065, "learning_rate": 3.9721015709910235e-06, "loss": 0.2838, "step": 4136 }, { "epoch": 1.8292485165665426, "grad_norm": 0.30822755675186286, "learning_rate": 3.9695845223814414e-06, "loss": 0.3301, "step": 4137 }, { "epoch": 1.8296907824420447, "grad_norm": 0.3379605077534898, "learning_rate": 3.967067746453734e-06, "loss": 0.3369, "step": 4138 }, { "epoch": 1.8301330483175469, "grad_norm": 0.37517351053191983, "learning_rate": 3.964551243873921e-06, "loss": 0.3041, "step": 4139 }, { "epoch": 1.830575314193049, "grad_norm": 0.3476983269237696, "learning_rate": 3.962035015307953e-06, "loss": 0.3262, "step": 4140 }, { "epoch": 1.831017580068551, "grad_norm": 0.30561983773808643, "learning_rate": 3.959519061421706e-06, "loss": 0.3267, "step": 4141 }, { "epoch": 1.8314598459440534, "grad_norm": 0.30186299902512626, "learning_rate": 3.957003382880985e-06, "loss": 0.3318, "step": 4142 }, { "epoch": 1.8319021118195555, "grad_norm": 0.3570518783606524, "learning_rate": 3.954487980351522e-06, "loss": 0.3414, "step": 4143 }, { "epoch": 1.8323443776950576, "grad_norm": 0.2810810446137309, "learning_rate": 3.951972854498976e-06, "loss": 0.3603, "step": 4144 }, { "epoch": 1.83278664357056, "grad_norm": 0.309660018354429, "learning_rate": 3.94945800598893e-06, "loss": 0.3481, "step": 4145 }, { "epoch": 1.833228909446062, "grad_norm": 0.33727997699199613, "learning_rate": 3.946943435486894e-06, "loss": 0.3464, "step": 4146 }, { "epoch": 1.8336711753215642, "grad_norm": 0.28777870730223976, "learning_rate": 3.9444291436583126e-06, "loss": 0.3618, "step": 4147 }, { "epoch": 1.8341134411970663, "grad_norm": 0.3727807463374909, "learning_rate": 3.9419151311685446e-06, "loss": 0.3186, "step": 4148 }, { "epoch": 1.8345557070725684, "grad_norm": 0.26616666592367505, "learning_rate": 3.939401398682884e-06, "loss": 0.3246, "step": 4149 }, { "epoch": 1.8349979729480705, "grad_norm": 0.3531857986078344, "learning_rate": 3.936887946866546e-06, "loss": 0.3218, "step": 4150 }, { "epoch": 1.8354402388235727, "grad_norm": 0.3131727919600656, "learning_rate": 3.934374776384671e-06, "loss": 0.337, "step": 4151 }, { "epoch": 1.8358825046990748, "grad_norm": 0.3764761412618856, "learning_rate": 3.9318618879023255e-06, "loss": 0.3143, "step": 4152 }, { "epoch": 1.836324770574577, "grad_norm": 0.3099679512849301, "learning_rate": 3.929349282084509e-06, "loss": 0.313, "step": 4153 }, { "epoch": 1.8367670364500792, "grad_norm": 0.27334455645968114, "learning_rate": 3.926836959596134e-06, "loss": 0.3212, "step": 4154 }, { "epoch": 1.8372093023255816, "grad_norm": 0.5258613657624621, "learning_rate": 3.924324921102047e-06, "loss": 0.3167, "step": 4155 }, { "epoch": 1.8376515682010837, "grad_norm": 0.27704475183182103, "learning_rate": 3.921813167267014e-06, "loss": 0.3196, "step": 4156 }, { "epoch": 1.8380938340765858, "grad_norm": 0.3298350003166911, "learning_rate": 3.919301698755728e-06, "loss": 0.3278, "step": 4157 }, { "epoch": 1.8385360999520879, "grad_norm": 0.24510000522432557, "learning_rate": 3.916790516232806e-06, "loss": 0.3156, "step": 4158 }, { "epoch": 1.83897836582759, "grad_norm": 0.3191023003268501, "learning_rate": 3.914279620362793e-06, "loss": 0.305, "step": 4159 }, { "epoch": 1.8394206317030921, "grad_norm": 0.38915059701680293, "learning_rate": 3.9117690118101505e-06, "loss": 0.3822, "step": 4160 }, { "epoch": 1.8398628975785942, "grad_norm": 0.33218833313229473, "learning_rate": 3.909258691239271e-06, "loss": 0.3361, "step": 4161 }, { "epoch": 1.8403051634540963, "grad_norm": 0.37286826830850917, "learning_rate": 3.906748659314464e-06, "loss": 0.3021, "step": 4162 }, { "epoch": 1.8407474293295987, "grad_norm": 0.27973287898809396, "learning_rate": 3.904238916699972e-06, "loss": 0.3119, "step": 4163 }, { "epoch": 1.8411896952051008, "grad_norm": 0.2833381057242619, "learning_rate": 3.901729464059953e-06, "loss": 0.3568, "step": 4164 }, { "epoch": 1.8416319610806031, "grad_norm": 0.3020804100919771, "learning_rate": 3.899220302058491e-06, "loss": 0.3359, "step": 4165 }, { "epoch": 1.8420742269561052, "grad_norm": 0.27947979657471544, "learning_rate": 3.896711431359593e-06, "loss": 0.3099, "step": 4166 }, { "epoch": 1.8425164928316073, "grad_norm": 0.2899851776328315, "learning_rate": 3.894202852627189e-06, "loss": 0.3174, "step": 4167 }, { "epoch": 1.8429587587071095, "grad_norm": 0.28492118417710827, "learning_rate": 3.891694566525129e-06, "loss": 0.3106, "step": 4168 }, { "epoch": 1.8434010245826116, "grad_norm": 0.40264165033438626, "learning_rate": 3.8891865737171915e-06, "loss": 0.3619, "step": 4169 }, { "epoch": 1.8438432904581137, "grad_norm": 0.33215087015760625, "learning_rate": 3.886678874867073e-06, "loss": 0.3208, "step": 4170 }, { "epoch": 1.8442855563336158, "grad_norm": 0.31008993229013, "learning_rate": 3.8841714706383936e-06, "loss": 0.3256, "step": 4171 }, { "epoch": 1.844727822209118, "grad_norm": 0.3047221644010014, "learning_rate": 3.8816643616946935e-06, "loss": 0.3066, "step": 4172 }, { "epoch": 1.8451700880846202, "grad_norm": 0.32744352591538356, "learning_rate": 3.8791575486994356e-06, "loss": 0.3427, "step": 4173 }, { "epoch": 1.8456123539601224, "grad_norm": 0.31190126699873044, "learning_rate": 3.876651032316009e-06, "loss": 0.3593, "step": 4174 }, { "epoch": 1.8460546198356245, "grad_norm": 0.2913824254004258, "learning_rate": 3.8741448132077184e-06, "loss": 0.3204, "step": 4175 }, { "epoch": 1.8464968857111268, "grad_norm": 0.31304898449459645, "learning_rate": 3.871638892037792e-06, "loss": 0.2876, "step": 4176 }, { "epoch": 1.846939151586629, "grad_norm": 0.26333176946134346, "learning_rate": 3.8691332694693785e-06, "loss": 0.318, "step": 4177 }, { "epoch": 1.847381417462131, "grad_norm": 0.29909475485304726, "learning_rate": 3.866627946165547e-06, "loss": 0.3283, "step": 4178 }, { "epoch": 1.8478236833376331, "grad_norm": 0.3196195913541604, "learning_rate": 3.8641229227892915e-06, "loss": 0.3439, "step": 4179 }, { "epoch": 1.8482659492131353, "grad_norm": 0.3894252388041125, "learning_rate": 3.861618200003523e-06, "loss": 0.3535, "step": 4180 }, { "epoch": 1.8487082150886374, "grad_norm": 0.3666784029177572, "learning_rate": 3.859113778471074e-06, "loss": 0.2985, "step": 4181 }, { "epoch": 1.8491504809641395, "grad_norm": 0.3120359278696083, "learning_rate": 3.856609658854694e-06, "loss": 0.3379, "step": 4182 }, { "epoch": 1.8495927468396418, "grad_norm": 0.27873279286692665, "learning_rate": 3.854105841817056e-06, "loss": 0.3031, "step": 4183 }, { "epoch": 1.850035012715144, "grad_norm": 0.3598293084812541, "learning_rate": 3.851602328020756e-06, "loss": 0.3484, "step": 4184 }, { "epoch": 1.850477278590646, "grad_norm": 0.31940087478643714, "learning_rate": 3.849099118128304e-06, "loss": 0.3486, "step": 4185 }, { "epoch": 1.8509195444661484, "grad_norm": 0.3021711535156108, "learning_rate": 3.8465962128021325e-06, "loss": 0.368, "step": 4186 }, { "epoch": 1.8513618103416505, "grad_norm": 0.2659457124202404, "learning_rate": 3.844093612704591e-06, "loss": 0.2982, "step": 4187 }, { "epoch": 1.8518040762171526, "grad_norm": 0.33905073323988344, "learning_rate": 3.841591318497951e-06, "loss": 0.3212, "step": 4188 }, { "epoch": 1.8522463420926547, "grad_norm": 0.3205277198256347, "learning_rate": 3.839089330844401e-06, "loss": 0.3434, "step": 4189 }, { "epoch": 1.8526886079681568, "grad_norm": 0.24675050373724894, "learning_rate": 3.836587650406052e-06, "loss": 0.329, "step": 4190 }, { "epoch": 1.853130873843659, "grad_norm": 0.29440484635355857, "learning_rate": 3.8340862778449284e-06, "loss": 0.3113, "step": 4191 }, { "epoch": 1.853573139719161, "grad_norm": 0.2674439937186099, "learning_rate": 3.8315852138229775e-06, "loss": 0.335, "step": 4192 }, { "epoch": 1.8540154055946634, "grad_norm": 0.30147598636251094, "learning_rate": 3.829084459002063e-06, "loss": 0.3097, "step": 4193 }, { "epoch": 1.8544576714701655, "grad_norm": 0.29499076073934605, "learning_rate": 3.826584014043963e-06, "loss": 0.2873, "step": 4194 }, { "epoch": 1.8548999373456676, "grad_norm": 0.309945328685767, "learning_rate": 3.824083879610384e-06, "loss": 0.3164, "step": 4195 }, { "epoch": 1.85534220322117, "grad_norm": 0.2967570030762046, "learning_rate": 3.821584056362941e-06, "loss": 0.3046, "step": 4196 }, { "epoch": 1.855784469096672, "grad_norm": 0.3421164032660365, "learning_rate": 3.819084544963168e-06, "loss": 0.318, "step": 4197 }, { "epoch": 1.8562267349721742, "grad_norm": 0.26346434850139927, "learning_rate": 3.816585346072521e-06, "loss": 0.3601, "step": 4198 }, { "epoch": 1.8566690008476763, "grad_norm": 0.2575941184520919, "learning_rate": 3.8140864603523666e-06, "loss": 0.3395, "step": 4199 }, { "epoch": 1.8571112667231784, "grad_norm": 0.30028642471431277, "learning_rate": 3.8115878884639968e-06, "loss": 0.3116, "step": 4200 }, { "epoch": 1.8575535325986805, "grad_norm": 0.3107144945351682, "learning_rate": 3.8090896310686133e-06, "loss": 0.361, "step": 4201 }, { "epoch": 1.8579957984741826, "grad_norm": 0.31159413489125937, "learning_rate": 3.806591688827337e-06, "loss": 0.316, "step": 4202 }, { "epoch": 1.8584380643496847, "grad_norm": 0.2816225187487507, "learning_rate": 3.804094062401207e-06, "loss": 0.3244, "step": 4203 }, { "epoch": 1.858880330225187, "grad_norm": 0.2689175593259988, "learning_rate": 3.8015967524511766e-06, "loss": 0.3184, "step": 4204 }, { "epoch": 1.8593225961006892, "grad_norm": 0.4865551549949413, "learning_rate": 3.7990997596381163e-06, "loss": 0.3398, "step": 4205 }, { "epoch": 1.8597648619761915, "grad_norm": 0.2756574064939213, "learning_rate": 3.796603084622813e-06, "loss": 0.3066, "step": 4206 }, { "epoch": 1.8602071278516936, "grad_norm": 0.30997577600450704, "learning_rate": 3.7941067280659695e-06, "loss": 0.3324, "step": 4207 }, { "epoch": 1.8606493937271957, "grad_norm": 0.26570688264405357, "learning_rate": 3.7916106906282035e-06, "loss": 0.2708, "step": 4208 }, { "epoch": 1.8610916596026978, "grad_norm": 0.31991862963414525, "learning_rate": 3.78911497297005e-06, "loss": 0.3244, "step": 4209 }, { "epoch": 1.8615339254782, "grad_norm": 0.3122902793662207, "learning_rate": 3.7866195757519538e-06, "loss": 0.3535, "step": 4210 }, { "epoch": 1.861976191353702, "grad_norm": 0.2988285212482233, "learning_rate": 3.784124499634284e-06, "loss": 0.3238, "step": 4211 }, { "epoch": 1.8624184572292042, "grad_norm": 0.2660256191730495, "learning_rate": 3.781629745277319e-06, "loss": 0.3064, "step": 4212 }, { "epoch": 1.8628607231047063, "grad_norm": 0.32969466708911804, "learning_rate": 3.779135313341252e-06, "loss": 0.3628, "step": 4213 }, { "epoch": 1.8633029889802086, "grad_norm": 0.29275280597188913, "learning_rate": 3.776641204486191e-06, "loss": 0.3173, "step": 4214 }, { "epoch": 1.8637452548557107, "grad_norm": 0.29323808788110595, "learning_rate": 3.7741474193721585e-06, "loss": 0.3193, "step": 4215 }, { "epoch": 1.864187520731213, "grad_norm": 0.3392459982078097, "learning_rate": 3.7716539586590974e-06, "loss": 0.3298, "step": 4216 }, { "epoch": 1.8646297866067152, "grad_norm": 0.2979859805327914, "learning_rate": 3.7691608230068548e-06, "loss": 0.3028, "step": 4217 }, { "epoch": 1.8650720524822173, "grad_norm": 0.3157747426676835, "learning_rate": 3.766668013075198e-06, "loss": 0.3452, "step": 4218 }, { "epoch": 1.8655143183577194, "grad_norm": 0.3271374645180258, "learning_rate": 3.764175529523807e-06, "loss": 0.3339, "step": 4219 }, { "epoch": 1.8659565842332215, "grad_norm": 0.37092011060231955, "learning_rate": 3.7616833730122727e-06, "loss": 0.367, "step": 4220 }, { "epoch": 1.8663988501087236, "grad_norm": 0.30023700701169737, "learning_rate": 3.759191544200105e-06, "loss": 0.3165, "step": 4221 }, { "epoch": 1.8668411159842258, "grad_norm": 0.30015192530289825, "learning_rate": 3.756700043746722e-06, "loss": 0.3233, "step": 4222 }, { "epoch": 1.8672833818597279, "grad_norm": 0.2743942119847933, "learning_rate": 3.754208872311458e-06, "loss": 0.3362, "step": 4223 }, { "epoch": 1.8677256477352302, "grad_norm": 0.37006490195632186, "learning_rate": 3.751718030553559e-06, "loss": 0.3164, "step": 4224 }, { "epoch": 1.8681679136107323, "grad_norm": 0.3232552416846354, "learning_rate": 3.749227519132182e-06, "loss": 0.333, "step": 4225 }, { "epoch": 1.8686101794862344, "grad_norm": 0.3692680533444253, "learning_rate": 3.7467373387063973e-06, "loss": 0.3529, "step": 4226 }, { "epoch": 1.8690524453617368, "grad_norm": 0.3034429716004852, "learning_rate": 3.744247489935192e-06, "loss": 0.3014, "step": 4227 }, { "epoch": 1.8694947112372389, "grad_norm": 0.3170175954124096, "learning_rate": 3.7417579734774634e-06, "loss": 0.3219, "step": 4228 }, { "epoch": 1.869936977112741, "grad_norm": 0.3162118666154732, "learning_rate": 3.739268789992015e-06, "loss": 0.3239, "step": 4229 }, { "epoch": 1.870379242988243, "grad_norm": 0.2559945448714664, "learning_rate": 3.736779940137569e-06, "loss": 0.3175, "step": 4230 }, { "epoch": 1.8708215088637452, "grad_norm": 0.2959464343770229, "learning_rate": 3.734291424572756e-06, "loss": 0.2884, "step": 4231 }, { "epoch": 1.8712637747392473, "grad_norm": 0.2727810282754067, "learning_rate": 3.7318032439561202e-06, "loss": 0.287, "step": 4232 }, { "epoch": 1.8717060406147494, "grad_norm": 0.3088682453632494, "learning_rate": 3.729315398946116e-06, "loss": 0.2771, "step": 4233 }, { "epoch": 1.8721483064902518, "grad_norm": 0.28056435939966246, "learning_rate": 3.7268278902011095e-06, "loss": 0.2887, "step": 4234 }, { "epoch": 1.8725905723657539, "grad_norm": 0.2771748486285794, "learning_rate": 3.724340718379377e-06, "loss": 0.3049, "step": 4235 }, { "epoch": 1.873032838241256, "grad_norm": 0.34018177745204914, "learning_rate": 3.721853884139105e-06, "loss": 0.3446, "step": 4236 }, { "epoch": 1.8734751041167583, "grad_norm": 0.2784845215601045, "learning_rate": 3.7193673881383935e-06, "loss": 0.3425, "step": 4237 }, { "epoch": 1.8739173699922604, "grad_norm": 0.28773040348049456, "learning_rate": 3.7168812310352507e-06, "loss": 0.3059, "step": 4238 }, { "epoch": 1.8743596358677626, "grad_norm": 0.32842848692484583, "learning_rate": 3.7143954134875955e-06, "loss": 0.3305, "step": 4239 }, { "epoch": 1.8748019017432647, "grad_norm": 0.36036843574438193, "learning_rate": 3.7119099361532584e-06, "loss": 0.3538, "step": 4240 }, { "epoch": 1.8752441676187668, "grad_norm": 0.2863875569992021, "learning_rate": 3.7094247996899755e-06, "loss": 0.312, "step": 4241 }, { "epoch": 1.875686433494269, "grad_norm": 0.3495133588101194, "learning_rate": 3.706940004755396e-06, "loss": 0.3496, "step": 4242 }, { "epoch": 1.876128699369771, "grad_norm": 0.29770784144255774, "learning_rate": 3.704455552007083e-06, "loss": 0.2948, "step": 4243 }, { "epoch": 1.8765709652452733, "grad_norm": 0.2827394569293132, "learning_rate": 3.7019714421025006e-06, "loss": 0.3238, "step": 4244 }, { "epoch": 1.8770132311207754, "grad_norm": 0.3229705683918989, "learning_rate": 3.699487675699027e-06, "loss": 0.3133, "step": 4245 }, { "epoch": 1.8774554969962776, "grad_norm": 0.2892554797340664, "learning_rate": 3.697004253453948e-06, "loss": 0.3564, "step": 4246 }, { "epoch": 1.87789776287178, "grad_norm": 0.3585128117495336, "learning_rate": 3.6945211760244595e-06, "loss": 0.336, "step": 4247 }, { "epoch": 1.878340028747282, "grad_norm": 0.32590022526026413, "learning_rate": 3.692038444067666e-06, "loss": 0.3464, "step": 4248 }, { "epoch": 1.8787822946227841, "grad_norm": 0.2963024308547752, "learning_rate": 3.6895560582405797e-06, "loss": 0.3152, "step": 4249 }, { "epoch": 1.8792245604982862, "grad_norm": 0.2828523258731132, "learning_rate": 3.687074019200122e-06, "loss": 0.3118, "step": 4250 }, { "epoch": 1.8796668263737883, "grad_norm": 0.2866093991860008, "learning_rate": 3.6845923276031214e-06, "loss": 0.3122, "step": 4251 }, { "epoch": 1.8801090922492905, "grad_norm": 0.3474404778741619, "learning_rate": 3.6821109841063152e-06, "loss": 0.3518, "step": 4252 }, { "epoch": 1.8805513581247926, "grad_norm": 0.27241725876655126, "learning_rate": 3.67962998936635e-06, "loss": 0.3252, "step": 4253 }, { "epoch": 1.8809936240002947, "grad_norm": 0.288997992375145, "learning_rate": 3.6771493440397777e-06, "loss": 0.3301, "step": 4254 }, { "epoch": 1.881435889875797, "grad_norm": 0.3921224650884588, "learning_rate": 3.67466904878306e-06, "loss": 0.3418, "step": 4255 }, { "epoch": 1.8818781557512991, "grad_norm": 0.2673239279822446, "learning_rate": 3.6721891042525624e-06, "loss": 0.3161, "step": 4256 }, { "epoch": 1.8823204216268015, "grad_norm": 0.26142152805995444, "learning_rate": 3.66970951110456e-06, "loss": 0.3246, "step": 4257 }, { "epoch": 1.8827626875023036, "grad_norm": 0.3002945605803395, "learning_rate": 3.667230269995237e-06, "loss": 0.3125, "step": 4258 }, { "epoch": 1.8832049533778057, "grad_norm": 0.3297901149376504, "learning_rate": 3.6647513815806823e-06, "loss": 0.3286, "step": 4259 }, { "epoch": 1.8836472192533078, "grad_norm": 0.2966522992472042, "learning_rate": 3.66227284651689e-06, "loss": 0.3169, "step": 4260 }, { "epoch": 1.88408948512881, "grad_norm": 0.31673300194386794, "learning_rate": 3.6597946654597616e-06, "loss": 0.3708, "step": 4261 }, { "epoch": 1.884531751004312, "grad_norm": 0.3585685622609409, "learning_rate": 3.657316839065107e-06, "loss": 0.3296, "step": 4262 }, { "epoch": 1.8849740168798141, "grad_norm": 0.34265272527645924, "learning_rate": 3.6548393679886385e-06, "loss": 0.3652, "step": 4263 }, { "epoch": 1.8854162827553163, "grad_norm": 0.3006132247983935, "learning_rate": 3.6523622528859793e-06, "loss": 0.3049, "step": 4264 }, { "epoch": 1.8858585486308186, "grad_norm": 0.2645657238820384, "learning_rate": 3.649885494412654e-06, "loss": 0.3217, "step": 4265 }, { "epoch": 1.8863008145063207, "grad_norm": 0.29826624706267346, "learning_rate": 3.6474090932240947e-06, "loss": 0.3296, "step": 4266 }, { "epoch": 1.886743080381823, "grad_norm": 0.3554194980467564, "learning_rate": 3.6449330499756396e-06, "loss": 0.3586, "step": 4267 }, { "epoch": 1.8871853462573251, "grad_norm": 0.32627491653361473, "learning_rate": 3.6424573653225286e-06, "loss": 0.3297, "step": 4268 }, { "epoch": 1.8876276121328273, "grad_norm": 0.3064729240393607, "learning_rate": 3.6399820399199126e-06, "loss": 0.3172, "step": 4269 }, { "epoch": 1.8880698780083294, "grad_norm": 0.29726261876230947, "learning_rate": 3.6375070744228434e-06, "loss": 0.3049, "step": 4270 }, { "epoch": 1.8885121438838315, "grad_norm": 0.2903085864008561, "learning_rate": 3.635032469486278e-06, "loss": 0.332, "step": 4271 }, { "epoch": 1.8889544097593336, "grad_norm": 0.38288532117964436, "learning_rate": 3.6325582257650784e-06, "loss": 0.3353, "step": 4272 }, { "epoch": 1.8893966756348357, "grad_norm": 0.2882409195894174, "learning_rate": 3.630084343914008e-06, "loss": 0.3199, "step": 4273 }, { "epoch": 1.8898389415103378, "grad_norm": 0.2828456131050346, "learning_rate": 3.627610824587744e-06, "loss": 0.3457, "step": 4274 }, { "epoch": 1.8902812073858402, "grad_norm": 0.26584489492688257, "learning_rate": 3.6251376684408557e-06, "loss": 0.321, "step": 4275 }, { "epoch": 1.8907234732613423, "grad_norm": 0.28274518851167263, "learning_rate": 3.6226648761278238e-06, "loss": 0.317, "step": 4276 }, { "epoch": 1.8911657391368444, "grad_norm": 0.3634560356670914, "learning_rate": 3.6201924483030305e-06, "loss": 0.3541, "step": 4277 }, { "epoch": 1.8916080050123467, "grad_norm": 0.31835232255307083, "learning_rate": 3.6177203856207598e-06, "loss": 0.3136, "step": 4278 }, { "epoch": 1.8920502708878488, "grad_norm": 0.3238001529599436, "learning_rate": 3.6152486887352027e-06, "loss": 0.3172, "step": 4279 }, { "epoch": 1.892492536763351, "grad_norm": 0.3824543858294771, "learning_rate": 3.612777358300452e-06, "loss": 0.3623, "step": 4280 }, { "epoch": 1.892934802638853, "grad_norm": 0.31158886696641946, "learning_rate": 3.6103063949705025e-06, "loss": 0.2881, "step": 4281 }, { "epoch": 1.8933770685143552, "grad_norm": 0.4049688540776919, "learning_rate": 3.607835799399252e-06, "loss": 0.3748, "step": 4282 }, { "epoch": 1.8938193343898573, "grad_norm": 0.39323290987847703, "learning_rate": 3.6053655722405034e-06, "loss": 0.3449, "step": 4283 }, { "epoch": 1.8942616002653594, "grad_norm": 0.2543802501263667, "learning_rate": 3.6028957141479552e-06, "loss": 0.2938, "step": 4284 }, { "epoch": 1.8947038661408617, "grad_norm": 0.3378436925611174, "learning_rate": 3.6004262257752186e-06, "loss": 0.3457, "step": 4285 }, { "epoch": 1.8951461320163638, "grad_norm": 0.3001336376780124, "learning_rate": 3.5979571077758e-06, "loss": 0.3125, "step": 4286 }, { "epoch": 1.895588397891866, "grad_norm": 0.2801424180727597, "learning_rate": 3.59548836080311e-06, "loss": 0.3025, "step": 4287 }, { "epoch": 1.8960306637673683, "grad_norm": 0.2627816615235408, "learning_rate": 3.593019985510458e-06, "loss": 0.3453, "step": 4288 }, { "epoch": 1.8964729296428704, "grad_norm": 0.31863283533916453, "learning_rate": 3.5905519825510573e-06, "loss": 0.34, "step": 4289 }, { "epoch": 1.8969151955183725, "grad_norm": 0.30448583233728943, "learning_rate": 3.5880843525780263e-06, "loss": 0.3379, "step": 4290 }, { "epoch": 1.8973574613938746, "grad_norm": 0.4094112860277661, "learning_rate": 3.585617096244378e-06, "loss": 0.3332, "step": 4291 }, { "epoch": 1.8977997272693767, "grad_norm": 0.4684932688135683, "learning_rate": 3.583150214203031e-06, "loss": 0.3559, "step": 4292 }, { "epoch": 1.8982419931448788, "grad_norm": 0.34336833000949807, "learning_rate": 3.5806837071068034e-06, "loss": 0.3173, "step": 4293 }, { "epoch": 1.898684259020381, "grad_norm": 0.2894111557518692, "learning_rate": 3.578217575608413e-06, "loss": 0.3235, "step": 4294 }, { "epoch": 1.8991265248958833, "grad_norm": 0.40132286580748827, "learning_rate": 3.5757518203604814e-06, "loss": 0.3418, "step": 4295 }, { "epoch": 1.8995687907713854, "grad_norm": 0.2781525330703849, "learning_rate": 3.5732864420155277e-06, "loss": 0.3052, "step": 4296 }, { "epoch": 1.9000110566468875, "grad_norm": 0.3963422699045132, "learning_rate": 3.5708214412259722e-06, "loss": 0.3571, "step": 4297 }, { "epoch": 1.9004533225223899, "grad_norm": 0.32104073834558355, "learning_rate": 3.5683568186441362e-06, "loss": 0.3776, "step": 4298 }, { "epoch": 1.900895588397892, "grad_norm": 0.2988942641902786, "learning_rate": 3.5658925749222393e-06, "loss": 0.2841, "step": 4299 }, { "epoch": 1.901337854273394, "grad_norm": 0.41880186972967465, "learning_rate": 3.563428710712399e-06, "loss": 0.3167, "step": 4300 }, { "epoch": 1.9017801201488962, "grad_norm": 0.3418061626900832, "learning_rate": 3.5609652266666384e-06, "loss": 0.3475, "step": 4301 }, { "epoch": 1.9022223860243983, "grad_norm": 0.31884003423013063, "learning_rate": 3.5585021234368777e-06, "loss": 0.3204, "step": 4302 }, { "epoch": 1.9026646518999004, "grad_norm": 0.30700445114923886, "learning_rate": 3.556039401674931e-06, "loss": 0.3314, "step": 4303 }, { "epoch": 1.9031069177754025, "grad_norm": 0.3897228383169585, "learning_rate": 3.5535770620325178e-06, "loss": 0.3631, "step": 4304 }, { "epoch": 1.9035491836509046, "grad_norm": 0.2791879116678652, "learning_rate": 3.5511151051612524e-06, "loss": 0.336, "step": 4305 }, { "epoch": 1.903991449526407, "grad_norm": 0.41583742514419764, "learning_rate": 3.548653531712653e-06, "loss": 0.3644, "step": 4306 }, { "epoch": 1.904433715401909, "grad_norm": 0.3424060138791318, "learning_rate": 3.5461923423381313e-06, "loss": 0.309, "step": 4307 }, { "epoch": 1.9048759812774114, "grad_norm": 0.3135268543629483, "learning_rate": 3.543731537688998e-06, "loss": 0.3267, "step": 4308 }, { "epoch": 1.9053182471529135, "grad_norm": 0.28903101526767755, "learning_rate": 3.541271118416465e-06, "loss": 0.3241, "step": 4309 }, { "epoch": 1.9057605130284156, "grad_norm": 0.38472353532533315, "learning_rate": 3.5388110851716383e-06, "loss": 0.3315, "step": 4310 }, { "epoch": 1.9062027789039178, "grad_norm": 0.30302122533695103, "learning_rate": 3.536351438605525e-06, "loss": 0.3242, "step": 4311 }, { "epoch": 1.9066450447794199, "grad_norm": 0.33238468331800297, "learning_rate": 3.5338921793690297e-06, "loss": 0.3003, "step": 4312 }, { "epoch": 1.907087310654922, "grad_norm": 0.34643110896942036, "learning_rate": 3.5314333081129513e-06, "loss": 0.3323, "step": 4313 }, { "epoch": 1.907529576530424, "grad_norm": 0.2749794683617592, "learning_rate": 3.5289748254879897e-06, "loss": 0.3167, "step": 4314 }, { "epoch": 1.9079718424059262, "grad_norm": 0.35225444851432847, "learning_rate": 3.526516732144739e-06, "loss": 0.3024, "step": 4315 }, { "epoch": 1.9084141082814285, "grad_norm": 0.2686555110478273, "learning_rate": 3.5240590287336897e-06, "loss": 0.3151, "step": 4316 }, { "epoch": 1.9088563741569307, "grad_norm": 0.32285008837977247, "learning_rate": 3.5216017159052362e-06, "loss": 0.3451, "step": 4317 }, { "epoch": 1.909298640032433, "grad_norm": 0.264370014009796, "learning_rate": 3.5191447943096625e-06, "loss": 0.3235, "step": 4318 }, { "epoch": 1.909740905907935, "grad_norm": 0.353099471515757, "learning_rate": 3.516688264597149e-06, "loss": 0.3226, "step": 4319 }, { "epoch": 1.9101831717834372, "grad_norm": 0.334704683441813, "learning_rate": 3.514232127417776e-06, "loss": 0.3152, "step": 4320 }, { "epoch": 1.9106254376589393, "grad_norm": 0.30062456643329266, "learning_rate": 3.511776383421517e-06, "loss": 0.3138, "step": 4321 }, { "epoch": 1.9110677035344414, "grad_norm": 0.3654545587642683, "learning_rate": 3.509321033258244e-06, "loss": 0.2955, "step": 4322 }, { "epoch": 1.9115099694099436, "grad_norm": 0.30265158428249256, "learning_rate": 3.5068660775777242e-06, "loss": 0.3422, "step": 4323 }, { "epoch": 1.9119522352854457, "grad_norm": 0.3937123694584135, "learning_rate": 3.504411517029618e-06, "loss": 0.3401, "step": 4324 }, { "epoch": 1.9123945011609478, "grad_norm": 0.2967534473160213, "learning_rate": 3.5019573522634836e-06, "loss": 0.2931, "step": 4325 }, { "epoch": 1.9128367670364501, "grad_norm": 0.2784111478633448, "learning_rate": 3.4995035839287734e-06, "loss": 0.319, "step": 4326 }, { "epoch": 1.9132790329119522, "grad_norm": 0.2925343177656228, "learning_rate": 3.497050212674837e-06, "loss": 0.324, "step": 4327 }, { "epoch": 1.9137212987874543, "grad_norm": 0.34538828377983183, "learning_rate": 3.4945972391509152e-06, "loss": 0.326, "step": 4328 }, { "epoch": 1.9141635646629567, "grad_norm": 0.320668809259325, "learning_rate": 3.4921446640061476e-06, "loss": 0.3002, "step": 4329 }, { "epoch": 1.9146058305384588, "grad_norm": 0.323177986140297, "learning_rate": 3.4896924878895656e-06, "loss": 0.3444, "step": 4330 }, { "epoch": 1.915048096413961, "grad_norm": 0.2885815799610063, "learning_rate": 3.487240711450093e-06, "loss": 0.3176, "step": 4331 }, { "epoch": 1.915490362289463, "grad_norm": 0.28233938122479507, "learning_rate": 3.484789335336554e-06, "loss": 0.3427, "step": 4332 }, { "epoch": 1.9159326281649651, "grad_norm": 0.3076991823520504, "learning_rate": 3.482338360197664e-06, "loss": 0.3365, "step": 4333 }, { "epoch": 1.9163748940404672, "grad_norm": 0.3128955603028553, "learning_rate": 3.479887786682029e-06, "loss": 0.3559, "step": 4334 }, { "epoch": 1.9168171599159693, "grad_norm": 0.3474380149683165, "learning_rate": 3.477437615438153e-06, "loss": 0.3804, "step": 4335 }, { "epoch": 1.9172594257914717, "grad_norm": 0.2946290582609789, "learning_rate": 3.4749878471144314e-06, "loss": 0.3047, "step": 4336 }, { "epoch": 1.9177016916669738, "grad_norm": 0.3381829994287831, "learning_rate": 3.4725384823591537e-06, "loss": 0.3127, "step": 4337 }, { "epoch": 1.918143957542476, "grad_norm": 0.2902752227678209, "learning_rate": 3.4700895218205026e-06, "loss": 0.3262, "step": 4338 }, { "epoch": 1.9185862234179782, "grad_norm": 0.3311676131567623, "learning_rate": 3.4676409661465543e-06, "loss": 0.3536, "step": 4339 }, { "epoch": 1.9190284892934804, "grad_norm": 0.3213856861867306, "learning_rate": 3.4651928159852765e-06, "loss": 0.3243, "step": 4340 }, { "epoch": 1.9194707551689825, "grad_norm": 0.32338388195095724, "learning_rate": 3.4627450719845314e-06, "loss": 0.3168, "step": 4341 }, { "epoch": 1.9199130210444846, "grad_norm": 0.3153162427908935, "learning_rate": 3.4602977347920696e-06, "loss": 0.3359, "step": 4342 }, { "epoch": 1.9203552869199867, "grad_norm": 0.3387158405081305, "learning_rate": 3.4578508050555413e-06, "loss": 0.3596, "step": 4343 }, { "epoch": 1.9207975527954888, "grad_norm": 0.45700257663768645, "learning_rate": 3.4554042834224826e-06, "loss": 0.3579, "step": 4344 }, { "epoch": 1.921239818670991, "grad_norm": 0.3608215464800119, "learning_rate": 3.4529581705403248e-06, "loss": 0.3549, "step": 4345 }, { "epoch": 1.9216820845464933, "grad_norm": 0.3357882059437387, "learning_rate": 3.4505124670563885e-06, "loss": 0.3097, "step": 4346 }, { "epoch": 1.9221243504219954, "grad_norm": 0.31875896084032784, "learning_rate": 3.4480671736178854e-06, "loss": 0.3298, "step": 4347 }, { "epoch": 1.9225666162974975, "grad_norm": 0.2956069906858791, "learning_rate": 3.4456222908719262e-06, "loss": 0.311, "step": 4348 }, { "epoch": 1.9230088821729998, "grad_norm": 0.26274519297005394, "learning_rate": 3.4431778194655048e-06, "loss": 0.3444, "step": 4349 }, { "epoch": 1.923451148048502, "grad_norm": 0.4020103472077496, "learning_rate": 3.4407337600455073e-06, "loss": 0.309, "step": 4350 }, { "epoch": 1.923893413924004, "grad_norm": 0.35641647593817455, "learning_rate": 3.4382901132587137e-06, "loss": 0.3285, "step": 4351 }, { "epoch": 1.9243356797995061, "grad_norm": 0.33673624856436996, "learning_rate": 3.435846879751793e-06, "loss": 0.3196, "step": 4352 }, { "epoch": 1.9247779456750083, "grad_norm": 0.29818680605026193, "learning_rate": 3.4334040601713047e-06, "loss": 0.3418, "step": 4353 }, { "epoch": 1.9252202115505104, "grad_norm": 0.3271409925950567, "learning_rate": 3.4309616551637016e-06, "loss": 0.3662, "step": 4354 }, { "epoch": 1.9256624774260125, "grad_norm": 0.2969585038429979, "learning_rate": 3.4285196653753228e-06, "loss": 0.2915, "step": 4355 }, { "epoch": 1.9261047433015146, "grad_norm": 0.3898849861185277, "learning_rate": 3.426078091452399e-06, "loss": 0.3214, "step": 4356 }, { "epoch": 1.926547009177017, "grad_norm": 0.33969633293924667, "learning_rate": 3.423636934041052e-06, "loss": 0.3326, "step": 4357 }, { "epoch": 1.926989275052519, "grad_norm": 0.4322057426047124, "learning_rate": 3.4211961937872905e-06, "loss": 0.3118, "step": 4358 }, { "epoch": 1.9274315409280214, "grad_norm": 0.3352143488826055, "learning_rate": 3.4187558713370174e-06, "loss": 0.3643, "step": 4359 }, { "epoch": 1.9278738068035235, "grad_norm": 0.42164486687231617, "learning_rate": 3.416315967336021e-06, "loss": 0.3889, "step": 4360 }, { "epoch": 1.9283160726790256, "grad_norm": 0.2728037034312862, "learning_rate": 3.4138764824299824e-06, "loss": 0.333, "step": 4361 }, { "epoch": 1.9287583385545277, "grad_norm": 0.246658420229474, "learning_rate": 3.411437417264466e-06, "loss": 0.3207, "step": 4362 }, { "epoch": 1.9292006044300298, "grad_norm": 0.32014082023671925, "learning_rate": 3.4089987724849294e-06, "loss": 0.3205, "step": 4363 }, { "epoch": 1.929642870305532, "grad_norm": 0.3280771165927599, "learning_rate": 3.4065605487367215e-06, "loss": 0.321, "step": 4364 }, { "epoch": 1.930085136181034, "grad_norm": 0.2500003511838968, "learning_rate": 3.404122746665074e-06, "loss": 0.3258, "step": 4365 }, { "epoch": 1.9305274020565362, "grad_norm": 0.39597637042442435, "learning_rate": 3.4016853669151103e-06, "loss": 0.3468, "step": 4366 }, { "epoch": 1.9309696679320385, "grad_norm": 0.3319930553602217, "learning_rate": 3.3992484101318407e-06, "loss": 0.3189, "step": 4367 }, { "epoch": 1.9314119338075406, "grad_norm": 0.27505159882407926, "learning_rate": 3.3968118769601645e-06, "loss": 0.3269, "step": 4368 }, { "epoch": 1.931854199683043, "grad_norm": 0.33169409199573296, "learning_rate": 3.3943757680448697e-06, "loss": 0.3661, "step": 4369 }, { "epoch": 1.932296465558545, "grad_norm": 0.33107776401259964, "learning_rate": 3.3919400840306294e-06, "loss": 0.3169, "step": 4370 }, { "epoch": 1.9327387314340472, "grad_norm": 0.26707703267546473, "learning_rate": 3.3895048255620076e-06, "loss": 0.3368, "step": 4371 }, { "epoch": 1.9331809973095493, "grad_norm": 0.3314990125483876, "learning_rate": 3.3870699932834524e-06, "loss": 0.3281, "step": 4372 }, { "epoch": 1.9336232631850514, "grad_norm": 0.275196990513953, "learning_rate": 3.3846355878393023e-06, "loss": 0.3527, "step": 4373 }, { "epoch": 1.9340655290605535, "grad_norm": 0.2957900374866727, "learning_rate": 3.3822016098737765e-06, "loss": 0.3194, "step": 4374 }, { "epoch": 1.9345077949360556, "grad_norm": 0.28663329487264083, "learning_rate": 3.3797680600309912e-06, "loss": 0.2987, "step": 4375 }, { "epoch": 1.9349500608115577, "grad_norm": 0.34175198063012313, "learning_rate": 3.377334938954944e-06, "loss": 0.3458, "step": 4376 }, { "epoch": 1.93539232668706, "grad_norm": 0.31307655571262466, "learning_rate": 3.3749022472895145e-06, "loss": 0.3281, "step": 4377 }, { "epoch": 1.9358345925625622, "grad_norm": 0.2912435021004476, "learning_rate": 3.3724699856784753e-06, "loss": 0.3118, "step": 4378 }, { "epoch": 1.9362768584380643, "grad_norm": 0.29385380359754043, "learning_rate": 3.3700381547654815e-06, "loss": 0.3269, "step": 4379 }, { "epoch": 1.9367191243135666, "grad_norm": 0.2952839332443341, "learning_rate": 3.3676067551940796e-06, "loss": 0.3103, "step": 4380 }, { "epoch": 1.9371613901890687, "grad_norm": 0.28750189402936016, "learning_rate": 3.3651757876076944e-06, "loss": 0.3656, "step": 4381 }, { "epoch": 1.9376036560645709, "grad_norm": 0.3017193670827897, "learning_rate": 3.362745252649641e-06, "loss": 0.335, "step": 4382 }, { "epoch": 1.938045921940073, "grad_norm": 0.28219247661128427, "learning_rate": 3.360315150963119e-06, "loss": 0.303, "step": 4383 }, { "epoch": 1.938488187815575, "grad_norm": 0.3312176166598465, "learning_rate": 3.3578854831912135e-06, "loss": 0.3153, "step": 4384 }, { "epoch": 1.9389304536910772, "grad_norm": 0.40385682766863445, "learning_rate": 3.355456249976894e-06, "loss": 0.3333, "step": 4385 }, { "epoch": 1.9393727195665793, "grad_norm": 0.27317335433304496, "learning_rate": 3.353027451963017e-06, "loss": 0.3248, "step": 4386 }, { "epoch": 1.9398149854420816, "grad_norm": 0.4350116646856476, "learning_rate": 3.350599089792321e-06, "loss": 0.3802, "step": 4387 }, { "epoch": 1.9402572513175838, "grad_norm": 0.28223659053497646, "learning_rate": 3.3481711641074317e-06, "loss": 0.3183, "step": 4388 }, { "epoch": 1.9406995171930859, "grad_norm": 0.27413822939308574, "learning_rate": 3.3457436755508587e-06, "loss": 0.3586, "step": 4389 }, { "epoch": 1.9411417830685882, "grad_norm": 0.49236968760166205, "learning_rate": 3.3433166247649906e-06, "loss": 0.3466, "step": 4390 }, { "epoch": 1.9415840489440903, "grad_norm": 0.28655545625195694, "learning_rate": 3.3408900123921117e-06, "loss": 0.3393, "step": 4391 }, { "epoch": 1.9420263148195924, "grad_norm": 0.24937225075269723, "learning_rate": 3.338463839074381e-06, "loss": 0.3501, "step": 4392 }, { "epoch": 1.9424685806950945, "grad_norm": 0.4317019627946105, "learning_rate": 3.3360381054538428e-06, "loss": 0.343, "step": 4393 }, { "epoch": 1.9429108465705967, "grad_norm": 0.32396067143768736, "learning_rate": 3.333612812172427e-06, "loss": 0.3237, "step": 4394 }, { "epoch": 1.9433531124460988, "grad_norm": 0.27433613129976664, "learning_rate": 3.331187959871944e-06, "loss": 0.3289, "step": 4395 }, { "epoch": 1.9437953783216009, "grad_norm": 0.5026772144098985, "learning_rate": 3.328763549194094e-06, "loss": 0.3533, "step": 4396 }, { "epoch": 1.9442376441971032, "grad_norm": 0.2588543914350296, "learning_rate": 3.326339580780452e-06, "loss": 0.3136, "step": 4397 }, { "epoch": 1.9446799100726053, "grad_norm": 0.29322179217868144, "learning_rate": 3.3239160552724824e-06, "loss": 0.3035, "step": 4398 }, { "epoch": 1.9451221759481074, "grad_norm": 0.32628934459510023, "learning_rate": 3.3214929733115282e-06, "loss": 0.3163, "step": 4399 }, { "epoch": 1.9455644418236098, "grad_norm": 0.30728589367942016, "learning_rate": 3.3190703355388165e-06, "loss": 0.3317, "step": 4400 }, { "epoch": 1.9460067076991119, "grad_norm": 0.30729195391148356, "learning_rate": 3.3166481425954583e-06, "loss": 0.3592, "step": 4401 }, { "epoch": 1.946448973574614, "grad_norm": 0.28610113684869504, "learning_rate": 3.314226395122445e-06, "loss": 0.3499, "step": 4402 }, { "epoch": 1.946891239450116, "grad_norm": 0.2626654706693482, "learning_rate": 3.3118050937606516e-06, "loss": 0.3213, "step": 4403 }, { "epoch": 1.9473335053256182, "grad_norm": 0.31767891749305854, "learning_rate": 3.3093842391508347e-06, "loss": 0.3432, "step": 4404 }, { "epoch": 1.9477757712011203, "grad_norm": 0.26039101458842445, "learning_rate": 3.306963831933629e-06, "loss": 0.2762, "step": 4405 }, { "epoch": 1.9482180370766224, "grad_norm": 0.32552705945999255, "learning_rate": 3.3045438727495545e-06, "loss": 0.3411, "step": 4406 }, { "epoch": 1.9486603029521246, "grad_norm": 0.30871325041264297, "learning_rate": 3.3021243622390164e-06, "loss": 0.3253, "step": 4407 }, { "epoch": 1.949102568827627, "grad_norm": 0.31385730353101793, "learning_rate": 3.2997053010422924e-06, "loss": 0.3027, "step": 4408 }, { "epoch": 1.949544834703129, "grad_norm": 0.29696951572026486, "learning_rate": 3.297286689799548e-06, "loss": 0.2948, "step": 4409 }, { "epoch": 1.9499871005786313, "grad_norm": 0.3686226649009672, "learning_rate": 3.2948685291508266e-06, "loss": 0.349, "step": 4410 }, { "epoch": 1.9504293664541335, "grad_norm": 0.2730230065129761, "learning_rate": 3.2924508197360515e-06, "loss": 0.308, "step": 4411 }, { "epoch": 1.9508716323296356, "grad_norm": 0.3307954776137592, "learning_rate": 3.2900335621950317e-06, "loss": 0.2999, "step": 4412 }, { "epoch": 1.9513138982051377, "grad_norm": 0.31174164641584534, "learning_rate": 3.2876167571674523e-06, "loss": 0.3319, "step": 4413 }, { "epoch": 1.9517561640806398, "grad_norm": 0.8388389590043706, "learning_rate": 3.2852004052928777e-06, "loss": 0.4199, "step": 4414 }, { "epoch": 1.952198429956142, "grad_norm": 0.3448910058650346, "learning_rate": 3.2827845072107557e-06, "loss": 0.3138, "step": 4415 }, { "epoch": 1.952640695831644, "grad_norm": 0.28252650195347956, "learning_rate": 3.28036906356041e-06, "loss": 0.3083, "step": 4416 }, { "epoch": 1.9530829617071461, "grad_norm": 0.282992179568135, "learning_rate": 3.2779540749810508e-06, "loss": 0.304, "step": 4417 }, { "epoch": 1.9535252275826485, "grad_norm": 0.28728741632962723, "learning_rate": 3.275539542111761e-06, "loss": 0.2933, "step": 4418 }, { "epoch": 1.9539674934581506, "grad_norm": 0.3273451855480747, "learning_rate": 3.2731254655915055e-06, "loss": 0.3112, "step": 4419 }, { "epoch": 1.954409759333653, "grad_norm": 0.28075161532807336, "learning_rate": 3.2707118460591304e-06, "loss": 0.3084, "step": 4420 }, { "epoch": 1.954852025209155, "grad_norm": 0.2965578273709476, "learning_rate": 3.2682986841533536e-06, "loss": 0.3487, "step": 4421 }, { "epoch": 1.9552942910846571, "grad_norm": 0.29288882520320086, "learning_rate": 3.2658859805127835e-06, "loss": 0.3298, "step": 4422 }, { "epoch": 1.9557365569601592, "grad_norm": 0.28715478247434956, "learning_rate": 3.2634737357758994e-06, "loss": 0.3352, "step": 4423 }, { "epoch": 1.9561788228356614, "grad_norm": 0.2909494760060591, "learning_rate": 3.2610619505810585e-06, "loss": 0.323, "step": 4424 }, { "epoch": 1.9566210887111635, "grad_norm": 0.31403265098601035, "learning_rate": 3.258650625566501e-06, "loss": 0.3474, "step": 4425 }, { "epoch": 1.9570633545866656, "grad_norm": 0.29529235214604704, "learning_rate": 3.25623976137034e-06, "loss": 0.3256, "step": 4426 }, { "epoch": 1.9575056204621677, "grad_norm": 0.3028272009952808, "learning_rate": 3.253829358630571e-06, "loss": 0.3398, "step": 4427 }, { "epoch": 1.95794788633767, "grad_norm": 0.38871200988575616, "learning_rate": 3.251419417985068e-06, "loss": 0.3417, "step": 4428 }, { "epoch": 1.9583901522131721, "grad_norm": 0.29674504148548964, "learning_rate": 3.249009940071578e-06, "loss": 0.3323, "step": 4429 }, { "epoch": 1.9588324180886743, "grad_norm": 0.3477784613441662, "learning_rate": 3.2466009255277297e-06, "loss": 0.3469, "step": 4430 }, { "epoch": 1.9592746839641766, "grad_norm": 0.30567804244769714, "learning_rate": 3.244192374991027e-06, "loss": 0.3505, "step": 4431 }, { "epoch": 1.9597169498396787, "grad_norm": 0.3411808193274508, "learning_rate": 3.241784289098851e-06, "loss": 0.3458, "step": 4432 }, { "epoch": 1.9601592157151808, "grad_norm": 0.29149977727500676, "learning_rate": 3.239376668488462e-06, "loss": 0.3291, "step": 4433 }, { "epoch": 1.960601481590683, "grad_norm": 0.3098286373203515, "learning_rate": 3.2369695137969957e-06, "loss": 0.325, "step": 4434 }, { "epoch": 1.961043747466185, "grad_norm": 0.29894704961863416, "learning_rate": 3.2345628256614646e-06, "loss": 0.3229, "step": 4435 }, { "epoch": 1.9614860133416872, "grad_norm": 0.2936884934166096, "learning_rate": 3.2321566047187564e-06, "loss": 0.3007, "step": 4436 }, { "epoch": 1.9619282792171893, "grad_norm": 0.3025946780003902, "learning_rate": 3.229750851605634e-06, "loss": 0.3059, "step": 4437 }, { "epoch": 1.9623705450926916, "grad_norm": 0.3577013243809982, "learning_rate": 3.227345566958746e-06, "loss": 0.3181, "step": 4438 }, { "epoch": 1.9628128109681937, "grad_norm": 0.33933322865294535, "learning_rate": 3.2249407514146048e-06, "loss": 0.3451, "step": 4439 }, { "epoch": 1.9632550768436958, "grad_norm": 0.3262397325788082, "learning_rate": 3.2225364056096044e-06, "loss": 0.3408, "step": 4440 }, { "epoch": 1.9636973427191982, "grad_norm": 0.28027600206398956, "learning_rate": 3.2201325301800148e-06, "loss": 0.294, "step": 4441 }, { "epoch": 1.9641396085947003, "grad_norm": 0.268645447601374, "learning_rate": 3.217729125761979e-06, "loss": 0.3354, "step": 4442 }, { "epoch": 1.9645818744702024, "grad_norm": 0.3427188067819958, "learning_rate": 3.2153261929915182e-06, "loss": 0.2925, "step": 4443 }, { "epoch": 1.9650241403457045, "grad_norm": 0.2942926040793734, "learning_rate": 3.2129237325045283e-06, "loss": 0.3232, "step": 4444 }, { "epoch": 1.9654664062212066, "grad_norm": 0.337208381940518, "learning_rate": 3.2105217449367787e-06, "loss": 0.3215, "step": 4445 }, { "epoch": 1.9659086720967087, "grad_norm": 0.3451615768336217, "learning_rate": 3.2081202309239136e-06, "loss": 0.3244, "step": 4446 }, { "epoch": 1.9663509379722108, "grad_norm": 0.3273989307313262, "learning_rate": 3.205719191101455e-06, "loss": 0.322, "step": 4447 }, { "epoch": 1.9667932038477132, "grad_norm": 0.3693575116472824, "learning_rate": 3.203318626104793e-06, "loss": 0.2927, "step": 4448 }, { "epoch": 1.9672354697232153, "grad_norm": 0.32308499578501765, "learning_rate": 3.2009185365692002e-06, "loss": 0.3457, "step": 4449 }, { "epoch": 1.9676777355987174, "grad_norm": 0.3090857043717236, "learning_rate": 3.1985189231298185e-06, "loss": 0.3287, "step": 4450 }, { "epoch": 1.9681200014742197, "grad_norm": 0.3456768768258506, "learning_rate": 3.196119786421664e-06, "loss": 0.3223, "step": 4451 }, { "epoch": 1.9685622673497218, "grad_norm": 0.27359187503604326, "learning_rate": 3.1937211270796276e-06, "loss": 0.3086, "step": 4452 }, { "epoch": 1.969004533225224, "grad_norm": 0.3359609956519503, "learning_rate": 3.191322945738471e-06, "loss": 0.321, "step": 4453 }, { "epoch": 1.969446799100726, "grad_norm": 0.28798598247518725, "learning_rate": 3.1889252430328368e-06, "loss": 0.3787, "step": 4454 }, { "epoch": 1.9698890649762282, "grad_norm": 0.352256153663426, "learning_rate": 3.1865280195972328e-06, "loss": 0.3504, "step": 4455 }, { "epoch": 1.9703313308517303, "grad_norm": 0.22450502909000208, "learning_rate": 3.1841312760660437e-06, "loss": 0.3202, "step": 4456 }, { "epoch": 1.9707735967272324, "grad_norm": 0.22906174636277632, "learning_rate": 3.1817350130735274e-06, "loss": 0.3293, "step": 4457 }, { "epoch": 1.9712158626027345, "grad_norm": 0.3229152962533508, "learning_rate": 3.179339231253813e-06, "loss": 0.316, "step": 4458 }, { "epoch": 1.9716581284782368, "grad_norm": 0.2640070365393751, "learning_rate": 3.1769439312409045e-06, "loss": 0.3381, "step": 4459 }, { "epoch": 1.972100394353739, "grad_norm": 0.312828081998184, "learning_rate": 3.1745491136686775e-06, "loss": 0.3511, "step": 4460 }, { "epoch": 1.9725426602292413, "grad_norm": 0.30931690685568397, "learning_rate": 3.1721547791708786e-06, "loss": 0.2899, "step": 4461 }, { "epoch": 1.9729849261047434, "grad_norm": 0.27616493212218235, "learning_rate": 3.169760928381127e-06, "loss": 0.3455, "step": 4462 }, { "epoch": 1.9734271919802455, "grad_norm": 0.3731270621230948, "learning_rate": 3.167367561932917e-06, "loss": 0.3474, "step": 4463 }, { "epoch": 1.9738694578557476, "grad_norm": 0.3879385708000914, "learning_rate": 3.164974680459607e-06, "loss": 0.3282, "step": 4464 }, { "epoch": 1.9743117237312497, "grad_norm": 0.24646663326377422, "learning_rate": 3.1625822845944385e-06, "loss": 0.29, "step": 4465 }, { "epoch": 1.9747539896067519, "grad_norm": 0.3569853314782431, "learning_rate": 3.160190374970516e-06, "loss": 0.3603, "step": 4466 }, { "epoch": 1.975196255482254, "grad_norm": 0.30001364489329774, "learning_rate": 3.1577989522208156e-06, "loss": 0.3354, "step": 4467 }, { "epoch": 1.975638521357756, "grad_norm": 0.2786032816229432, "learning_rate": 3.1554080169781885e-06, "loss": 0.3358, "step": 4468 }, { "epoch": 1.9760807872332584, "grad_norm": 0.30501787245048295, "learning_rate": 3.153017569875353e-06, "loss": 0.3617, "step": 4469 }, { "epoch": 1.9765230531087605, "grad_norm": 0.28070177691933157, "learning_rate": 3.1506276115449025e-06, "loss": 0.3254, "step": 4470 }, { "epoch": 1.9769653189842629, "grad_norm": 0.3106982513095064, "learning_rate": 3.1482381426192974e-06, "loss": 0.3126, "step": 4471 }, { "epoch": 1.977407584859765, "grad_norm": 0.3385229005390539, "learning_rate": 3.1458491637308706e-06, "loss": 0.3117, "step": 4472 }, { "epoch": 1.977849850735267, "grad_norm": 0.261369836502131, "learning_rate": 3.1434606755118237e-06, "loss": 0.3136, "step": 4473 }, { "epoch": 1.9782921166107692, "grad_norm": 0.29849491443839876, "learning_rate": 3.1410726785942293e-06, "loss": 0.3279, "step": 4474 }, { "epoch": 1.9787343824862713, "grad_norm": 0.3136212723856803, "learning_rate": 3.138685173610031e-06, "loss": 0.3389, "step": 4475 }, { "epoch": 1.9791766483617734, "grad_norm": 0.3470595157343114, "learning_rate": 3.1362981611910414e-06, "loss": 0.3027, "step": 4476 }, { "epoch": 1.9796189142372755, "grad_norm": 0.3141498164096647, "learning_rate": 3.133911641968942e-06, "loss": 0.2865, "step": 4477 }, { "epoch": 1.9800611801127777, "grad_norm": 0.31584591444824556, "learning_rate": 3.131525616575285e-06, "loss": 0.3253, "step": 4478 }, { "epoch": 1.98050344598828, "grad_norm": 0.27441874045939396, "learning_rate": 3.1291400856414892e-06, "loss": 0.343, "step": 4479 }, { "epoch": 1.980945711863782, "grad_norm": 0.35240402397734955, "learning_rate": 3.1267550497988446e-06, "loss": 0.3202, "step": 4480 }, { "epoch": 1.9813879777392842, "grad_norm": 0.25765347642024866, "learning_rate": 3.124370509678514e-06, "loss": 0.3176, "step": 4481 }, { "epoch": 1.9818302436147865, "grad_norm": 0.3753429101399068, "learning_rate": 3.121986465911524e-06, "loss": 0.3372, "step": 4482 }, { "epoch": 1.9822725094902887, "grad_norm": 0.32812919960623993, "learning_rate": 3.119602919128769e-06, "loss": 0.3295, "step": 4483 }, { "epoch": 1.9827147753657908, "grad_norm": 0.3132505207813021, "learning_rate": 3.117219869961015e-06, "loss": 0.3449, "step": 4484 }, { "epoch": 1.9831570412412929, "grad_norm": 0.3146677353339341, "learning_rate": 3.1148373190388947e-06, "loss": 0.3177, "step": 4485 }, { "epoch": 1.983599307116795, "grad_norm": 0.2786856400958907, "learning_rate": 3.112455266992911e-06, "loss": 0.3001, "step": 4486 }, { "epoch": 1.984041572992297, "grad_norm": 0.3339253835018693, "learning_rate": 3.1100737144534326e-06, "loss": 0.3422, "step": 4487 }, { "epoch": 1.9844838388677992, "grad_norm": 0.25108742424416297, "learning_rate": 3.107692662050697e-06, "loss": 0.3142, "step": 4488 }, { "epoch": 1.9849261047433016, "grad_norm": 0.38195980217221076, "learning_rate": 3.105312110414809e-06, "loss": 0.3303, "step": 4489 }, { "epoch": 1.9853683706188037, "grad_norm": 0.2948144058137821, "learning_rate": 3.102932060175739e-06, "loss": 0.3298, "step": 4490 }, { "epoch": 1.9858106364943058, "grad_norm": 0.3597209815233857, "learning_rate": 3.10055251196333e-06, "loss": 0.3745, "step": 4491 }, { "epoch": 1.9862529023698081, "grad_norm": 0.2800037719612802, "learning_rate": 3.0981734664072873e-06, "loss": 0.2964, "step": 4492 }, { "epoch": 1.9866951682453102, "grad_norm": 0.33231588207570867, "learning_rate": 3.0957949241371845e-06, "loss": 0.3088, "step": 4493 }, { "epoch": 1.9871374341208123, "grad_norm": 0.35009598878789816, "learning_rate": 3.0934168857824637e-06, "loss": 0.322, "step": 4494 }, { "epoch": 1.9875796999963145, "grad_norm": 0.3026318239916002, "learning_rate": 3.091039351972428e-06, "loss": 0.3225, "step": 4495 }, { "epoch": 1.9880219658718166, "grad_norm": 0.34844799118806036, "learning_rate": 3.0886623233362555e-06, "loss": 0.3448, "step": 4496 }, { "epoch": 1.9884642317473187, "grad_norm": 0.6993386451845096, "learning_rate": 3.0862858005029865e-06, "loss": 0.3989, "step": 4497 }, { "epoch": 1.9889064976228208, "grad_norm": 0.3273893636531369, "learning_rate": 3.0839097841015243e-06, "loss": 0.3082, "step": 4498 }, { "epoch": 1.9893487634983231, "grad_norm": 0.3383680419111177, "learning_rate": 3.0815342747606424e-06, "loss": 0.3016, "step": 4499 }, { "epoch": 1.9897910293738252, "grad_norm": 0.38276723322029776, "learning_rate": 3.0791592731089785e-06, "loss": 0.32, "step": 4500 }, { "epoch": 1.9902332952493273, "grad_norm": 0.35725366106842454, "learning_rate": 3.0767847797750363e-06, "loss": 0.3402, "step": 4501 }, { "epoch": 1.9906755611248297, "grad_norm": 0.32955084817327523, "learning_rate": 3.0744107953871847e-06, "loss": 0.2918, "step": 4502 }, { "epoch": 1.9911178270003318, "grad_norm": 0.27956180936756403, "learning_rate": 3.0720373205736597e-06, "loss": 0.318, "step": 4503 }, { "epoch": 1.991560092875834, "grad_norm": 0.34649897748131586, "learning_rate": 3.0696643559625598e-06, "loss": 0.3458, "step": 4504 }, { "epoch": 1.992002358751336, "grad_norm": 0.33319382281061466, "learning_rate": 3.0672919021818494e-06, "loss": 0.3256, "step": 4505 }, { "epoch": 1.9924446246268381, "grad_norm": 0.38399091135815494, "learning_rate": 3.064919959859357e-06, "loss": 0.3422, "step": 4506 }, { "epoch": 1.9928868905023402, "grad_norm": 0.3575498923337647, "learning_rate": 3.0625485296227798e-06, "loss": 0.3451, "step": 4507 }, { "epoch": 1.9933291563778424, "grad_norm": 0.37754497926379393, "learning_rate": 3.060177612099673e-06, "loss": 0.3733, "step": 4508 }, { "epoch": 1.9937714222533445, "grad_norm": 0.32997611214716493, "learning_rate": 3.0578072079174626e-06, "loss": 0.2978, "step": 4509 }, { "epoch": 1.9942136881288468, "grad_norm": 0.2831839726340566, "learning_rate": 3.0554373177034322e-06, "loss": 0.315, "step": 4510 }, { "epoch": 1.994655954004349, "grad_norm": 0.2759836309454834, "learning_rate": 3.053067942084733e-06, "loss": 0.3168, "step": 4511 }, { "epoch": 1.9950982198798513, "grad_norm": 0.31849571191670245, "learning_rate": 3.0506990816883835e-06, "loss": 0.3294, "step": 4512 }, { "epoch": 1.9955404857553534, "grad_norm": 0.39626879717422914, "learning_rate": 3.0483307371412606e-06, "loss": 0.3256, "step": 4513 }, { "epoch": 1.9959827516308555, "grad_norm": 0.4186981390552353, "learning_rate": 3.045962909070105e-06, "loss": 0.3194, "step": 4514 }, { "epoch": 1.9964250175063576, "grad_norm": 0.2981650226760145, "learning_rate": 3.043595598101522e-06, "loss": 0.3313, "step": 4515 }, { "epoch": 1.9968672833818597, "grad_norm": 0.34228138721478, "learning_rate": 3.041228804861981e-06, "loss": 0.3172, "step": 4516 }, { "epoch": 1.9973095492573618, "grad_norm": 0.27614270754465564, "learning_rate": 3.038862529977812e-06, "loss": 0.3221, "step": 4517 }, { "epoch": 1.997751815132864, "grad_norm": 0.34984295176933816, "learning_rate": 3.036496774075211e-06, "loss": 0.331, "step": 4518 }, { "epoch": 1.998194081008366, "grad_norm": 0.2711960051746663, "learning_rate": 3.034131537780235e-06, "loss": 0.3376, "step": 4519 }, { "epoch": 1.9986363468838684, "grad_norm": 0.3206719684950383, "learning_rate": 3.0317668217188025e-06, "loss": 0.3617, "step": 4520 }, { "epoch": 1.9990786127593705, "grad_norm": 0.30673384101740214, "learning_rate": 3.0294026265166953e-06, "loss": 0.3202, "step": 4521 }, { "epoch": 1.9995208786348728, "grad_norm": 0.30738249555513714, "learning_rate": 3.027038952799557e-06, "loss": 0.331, "step": 4522 }, { "epoch": 1.999963144510375, "grad_norm": 0.3274317236382903, "learning_rate": 3.0246758011928946e-06, "loss": 0.3592, "step": 4523 }, { "epoch": 2.0, "grad_norm": 0.3274317236382903, "learning_rate": 3.0223131723220756e-06, "loss": 0.4089, "step": 4524 }, { "epoch": 2.000442265875502, "grad_norm": 2.2032757293574874, "learning_rate": 3.019951066812329e-06, "loss": 0.291, "step": 4525 }, { "epoch": 2.0008845317510042, "grad_norm": 0.4085822914645364, "learning_rate": 3.0175894852887468e-06, "loss": 0.2872, "step": 4526 }, { "epoch": 2.0013267976265063, "grad_norm": 0.36812068656215413, "learning_rate": 3.0152284283762768e-06, "loss": 0.263, "step": 4527 }, { "epoch": 2.0017690635020084, "grad_norm": 0.3579594619300345, "learning_rate": 3.0128678966997394e-06, "loss": 0.2934, "step": 4528 }, { "epoch": 2.0022113293775106, "grad_norm": 0.29006309456791024, "learning_rate": 3.010507890883804e-06, "loss": 0.2478, "step": 4529 }, { "epoch": 2.002653595253013, "grad_norm": 0.38370388893305823, "learning_rate": 3.008148411553008e-06, "loss": 0.2988, "step": 4530 }, { "epoch": 2.0030958611285152, "grad_norm": 0.38275073802559784, "learning_rate": 3.005789459331746e-06, "loss": 0.2723, "step": 4531 }, { "epoch": 2.0035381270040173, "grad_norm": 0.3625624382360611, "learning_rate": 3.003431034844275e-06, "loss": 0.254, "step": 4532 }, { "epoch": 2.0039803928795195, "grad_norm": 0.3527728793870892, "learning_rate": 3.001073138714713e-06, "loss": 0.2768, "step": 4533 }, { "epoch": 2.0044226587550216, "grad_norm": 0.268951437315243, "learning_rate": 2.9987157715670355e-06, "loss": 0.2761, "step": 4534 }, { "epoch": 2.0048649246305237, "grad_norm": 0.3343330322797753, "learning_rate": 2.99635893402508e-06, "loss": 0.277, "step": 4535 }, { "epoch": 2.005307190506026, "grad_norm": 0.2733596334771239, "learning_rate": 2.9940026267125432e-06, "loss": 0.2783, "step": 4536 }, { "epoch": 2.005749456381528, "grad_norm": 0.3369420020704987, "learning_rate": 2.991646850252983e-06, "loss": 0.2744, "step": 4537 }, { "epoch": 2.00619172225703, "grad_norm": 0.32192827425260534, "learning_rate": 2.9892916052698116e-06, "loss": 0.2415, "step": 4538 }, { "epoch": 2.006633988132532, "grad_norm": 0.28628158685270527, "learning_rate": 2.9869368923863086e-06, "loss": 0.2436, "step": 4539 }, { "epoch": 2.0070762540080347, "grad_norm": 0.36973292364215044, "learning_rate": 2.984582712225608e-06, "loss": 0.287, "step": 4540 }, { "epoch": 2.007518519883537, "grad_norm": 0.3154017804393871, "learning_rate": 2.9822290654107007e-06, "loss": 0.2572, "step": 4541 }, { "epoch": 2.007960785759039, "grad_norm": 0.3748301626722012, "learning_rate": 2.9798759525644412e-06, "loss": 0.2826, "step": 4542 }, { "epoch": 2.008403051634541, "grad_norm": 0.3412069882983331, "learning_rate": 2.9775233743095378e-06, "loss": 0.2885, "step": 4543 }, { "epoch": 2.008845317510043, "grad_norm": 0.29654288914427523, "learning_rate": 2.9751713312685664e-06, "loss": 0.2782, "step": 4544 }, { "epoch": 2.0092875833855453, "grad_norm": 0.3607832863253973, "learning_rate": 2.9728198240639506e-06, "loss": 0.2882, "step": 4545 }, { "epoch": 2.0097298492610474, "grad_norm": 0.40619905180692334, "learning_rate": 2.9704688533179763e-06, "loss": 0.2679, "step": 4546 }, { "epoch": 2.0101721151365495, "grad_norm": 0.3809465469142293, "learning_rate": 2.9681184196527897e-06, "loss": 0.3075, "step": 4547 }, { "epoch": 2.0106143810120516, "grad_norm": 0.35963687253613624, "learning_rate": 2.965768523690391e-06, "loss": 0.277, "step": 4548 }, { "epoch": 2.0110566468875537, "grad_norm": 0.3184051118598365, "learning_rate": 2.9634191660526424e-06, "loss": 0.272, "step": 4549 }, { "epoch": 2.0114989127630563, "grad_norm": 0.42648482637677376, "learning_rate": 2.961070347361259e-06, "loss": 0.2955, "step": 4550 }, { "epoch": 2.0119411786385584, "grad_norm": 0.37483573662637276, "learning_rate": 2.958722068237818e-06, "loss": 0.2481, "step": 4551 }, { "epoch": 2.0123834445140605, "grad_norm": 0.32838457188261944, "learning_rate": 2.9563743293037495e-06, "loss": 0.2735, "step": 4552 }, { "epoch": 2.0128257103895626, "grad_norm": 0.29368923063836816, "learning_rate": 2.954027131180344e-06, "loss": 0.2626, "step": 4553 }, { "epoch": 2.0132679762650647, "grad_norm": 0.43212065596532934, "learning_rate": 2.951680474488744e-06, "loss": 0.2601, "step": 4554 }, { "epoch": 2.013710242140567, "grad_norm": 0.378827255657009, "learning_rate": 2.949334359849957e-06, "loss": 0.2682, "step": 4555 }, { "epoch": 2.014152508016069, "grad_norm": 0.31903877871689584, "learning_rate": 2.9469887878848403e-06, "loss": 0.2608, "step": 4556 }, { "epoch": 2.014594773891571, "grad_norm": 0.32883413196027655, "learning_rate": 2.9446437592141085e-06, "loss": 0.2811, "step": 4557 }, { "epoch": 2.015037039767073, "grad_norm": 0.3932866699751054, "learning_rate": 2.9422992744583332e-06, "loss": 0.2804, "step": 4558 }, { "epoch": 2.0154793056425753, "grad_norm": 0.3932811882091221, "learning_rate": 2.939955334237943e-06, "loss": 0.2787, "step": 4559 }, { "epoch": 2.015921571518078, "grad_norm": 0.3005837971370114, "learning_rate": 2.9376119391732227e-06, "loss": 0.3173, "step": 4560 }, { "epoch": 2.01636383739358, "grad_norm": 0.42638701444819527, "learning_rate": 2.935269089884312e-06, "loss": 0.2594, "step": 4561 }, { "epoch": 2.016806103269082, "grad_norm": 0.34491538255597454, "learning_rate": 2.9329267869912033e-06, "loss": 0.261, "step": 4562 }, { "epoch": 2.017248369144584, "grad_norm": 0.3733563905206527, "learning_rate": 2.9305850311137497e-06, "loss": 0.2775, "step": 4563 }, { "epoch": 2.0176906350200863, "grad_norm": 0.2624738087766978, "learning_rate": 2.9282438228716538e-06, "loss": 0.253, "step": 4564 }, { "epoch": 2.0181329008955884, "grad_norm": 0.3000709003686318, "learning_rate": 2.9259031628844803e-06, "loss": 0.2575, "step": 4565 }, { "epoch": 2.0185751667710905, "grad_norm": 0.28523200659961684, "learning_rate": 2.9235630517716434e-06, "loss": 0.2574, "step": 4566 }, { "epoch": 2.0190174326465926, "grad_norm": 0.3195562290772677, "learning_rate": 2.9212234901524134e-06, "loss": 0.2598, "step": 4567 }, { "epoch": 2.0194596985220947, "grad_norm": 0.32506796490420004, "learning_rate": 2.918884478645916e-06, "loss": 0.292, "step": 4568 }, { "epoch": 2.019901964397597, "grad_norm": 0.2592588140896866, "learning_rate": 2.9165460178711267e-06, "loss": 0.2808, "step": 4569 }, { "epoch": 2.0203442302730994, "grad_norm": 0.3722856785868823, "learning_rate": 2.9142081084468854e-06, "loss": 0.2639, "step": 4570 }, { "epoch": 2.0207864961486015, "grad_norm": 0.3483476322731864, "learning_rate": 2.9118707509918764e-06, "loss": 0.2736, "step": 4571 }, { "epoch": 2.0212287620241036, "grad_norm": 0.3494948301460053, "learning_rate": 2.909533946124643e-06, "loss": 0.2727, "step": 4572 }, { "epoch": 2.0216710278996057, "grad_norm": 0.4907380201686091, "learning_rate": 2.9071976944635792e-06, "loss": 0.3197, "step": 4573 }, { "epoch": 2.022113293775108, "grad_norm": 0.35325931009612155, "learning_rate": 2.9048619966269366e-06, "loss": 0.2605, "step": 4574 }, { "epoch": 2.02255555965061, "grad_norm": 0.32245932503540964, "learning_rate": 2.9025268532328126e-06, "loss": 0.2578, "step": 4575 }, { "epoch": 2.022997825526112, "grad_norm": 0.4530838910957023, "learning_rate": 2.9001922648991683e-06, "loss": 0.2826, "step": 4576 }, { "epoch": 2.023440091401614, "grad_norm": 0.29970562266140277, "learning_rate": 2.8978582322438107e-06, "loss": 0.2684, "step": 4577 }, { "epoch": 2.0238823572771163, "grad_norm": 0.3982397225423445, "learning_rate": 2.8955247558844017e-06, "loss": 0.2905, "step": 4578 }, { "epoch": 2.0243246231526184, "grad_norm": 0.39070922144887366, "learning_rate": 2.8931918364384566e-06, "loss": 0.2583, "step": 4579 }, { "epoch": 2.0247668890281205, "grad_norm": 0.27391789148405427, "learning_rate": 2.8908594745233396e-06, "loss": 0.2472, "step": 4580 }, { "epoch": 2.025209154903623, "grad_norm": 0.4068575311352622, "learning_rate": 2.888527670756278e-06, "loss": 0.274, "step": 4581 }, { "epoch": 2.025651420779125, "grad_norm": 0.37159769380020397, "learning_rate": 2.8861964257543363e-06, "loss": 0.2862, "step": 4582 }, { "epoch": 2.0260936866546273, "grad_norm": 0.31832429142827207, "learning_rate": 2.8838657401344418e-06, "loss": 0.2714, "step": 4583 }, { "epoch": 2.0265359525301294, "grad_norm": 0.371750162407223, "learning_rate": 2.8815356145133707e-06, "loss": 0.2811, "step": 4584 }, { "epoch": 2.0269782184056315, "grad_norm": 0.33310786713021484, "learning_rate": 2.8792060495077477e-06, "loss": 0.2848, "step": 4585 }, { "epoch": 2.0274204842811336, "grad_norm": 0.28325275708248465, "learning_rate": 2.876877045734058e-06, "loss": 0.2639, "step": 4586 }, { "epoch": 2.0278627501566358, "grad_norm": 0.3277276010484155, "learning_rate": 2.87454860380863e-06, "loss": 0.282, "step": 4587 }, { "epoch": 2.028305016032138, "grad_norm": 0.3406638403833372, "learning_rate": 2.8722207243476456e-06, "loss": 0.2569, "step": 4588 }, { "epoch": 2.02874728190764, "grad_norm": 0.3744331608663219, "learning_rate": 2.8698934079671394e-06, "loss": 0.2763, "step": 4589 }, { "epoch": 2.029189547783142, "grad_norm": 0.3164643494853745, "learning_rate": 2.8675666552829972e-06, "loss": 0.2807, "step": 4590 }, { "epoch": 2.0296318136586446, "grad_norm": 0.32577831053322714, "learning_rate": 2.865240466910949e-06, "loss": 0.264, "step": 4591 }, { "epoch": 2.0300740795341468, "grad_norm": 0.2909414601006975, "learning_rate": 2.8629148434665867e-06, "loss": 0.2617, "step": 4592 }, { "epoch": 2.030516345409649, "grad_norm": 0.36066159412824167, "learning_rate": 2.860589785565344e-06, "loss": 0.2855, "step": 4593 }, { "epoch": 2.030958611285151, "grad_norm": 0.32384242879071273, "learning_rate": 2.8582652938225097e-06, "loss": 0.2511, "step": 4594 }, { "epoch": 2.031400877160653, "grad_norm": 0.30119026839448276, "learning_rate": 2.855941368853219e-06, "loss": 0.2853, "step": 4595 }, { "epoch": 2.031843143036155, "grad_norm": 0.3002655543168251, "learning_rate": 2.85361801127246e-06, "loss": 0.2749, "step": 4596 }, { "epoch": 2.0322854089116573, "grad_norm": 0.3653583146577153, "learning_rate": 2.8512952216950685e-06, "loss": 0.2755, "step": 4597 }, { "epoch": 2.0327276747871594, "grad_norm": 0.29906304790238675, "learning_rate": 2.8489730007357325e-06, "loss": 0.2466, "step": 4598 }, { "epoch": 2.0331699406626615, "grad_norm": 0.31904960489987755, "learning_rate": 2.8466513490089877e-06, "loss": 0.2606, "step": 4599 }, { "epoch": 2.0336122065381637, "grad_norm": 0.3708340859246959, "learning_rate": 2.8443302671292193e-06, "loss": 0.2654, "step": 4600 }, { "epoch": 2.034054472413666, "grad_norm": 0.37329283119738915, "learning_rate": 2.8420097557106606e-06, "loss": 0.2543, "step": 4601 }, { "epoch": 2.0344967382891683, "grad_norm": 0.328154547094084, "learning_rate": 2.8396898153673985e-06, "loss": 0.2696, "step": 4602 }, { "epoch": 2.0349390041646704, "grad_norm": 0.3150099874749454, "learning_rate": 2.8373704467133647e-06, "loss": 0.2357, "step": 4603 }, { "epoch": 2.0353812700401726, "grad_norm": 0.320693922446943, "learning_rate": 2.8350516503623394e-06, "loss": 0.2958, "step": 4604 }, { "epoch": 2.0358235359156747, "grad_norm": 0.28849132301235053, "learning_rate": 2.832733426927955e-06, "loss": 0.2631, "step": 4605 }, { "epoch": 2.0362658017911768, "grad_norm": 0.32386587914557524, "learning_rate": 2.830415777023684e-06, "loss": 0.2801, "step": 4606 }, { "epoch": 2.036708067666679, "grad_norm": 0.3097278036487964, "learning_rate": 2.8280987012628592e-06, "loss": 0.2521, "step": 4607 }, { "epoch": 2.037150333542181, "grad_norm": 0.32565913443775657, "learning_rate": 2.8257822002586516e-06, "loss": 0.2864, "step": 4608 }, { "epoch": 2.037592599417683, "grad_norm": 0.3420010740471346, "learning_rate": 2.8234662746240854e-06, "loss": 0.2543, "step": 4609 }, { "epoch": 2.0380348652931852, "grad_norm": 0.35965176701414203, "learning_rate": 2.8211509249720305e-06, "loss": 0.2712, "step": 4610 }, { "epoch": 2.038477131168688, "grad_norm": 0.35643452880620724, "learning_rate": 2.8188361519152034e-06, "loss": 0.2572, "step": 4611 }, { "epoch": 2.03891939704419, "grad_norm": 0.3507828628353485, "learning_rate": 2.8165219560661707e-06, "loss": 0.2425, "step": 4612 }, { "epoch": 2.039361662919692, "grad_norm": 0.3297926713023743, "learning_rate": 2.814208338037344e-06, "loss": 0.3006, "step": 4613 }, { "epoch": 2.039803928795194, "grad_norm": 0.351054774056054, "learning_rate": 2.811895298440984e-06, "loss": 0.2713, "step": 4614 }, { "epoch": 2.0402461946706962, "grad_norm": 0.34336981270025346, "learning_rate": 2.8095828378891953e-06, "loss": 0.2739, "step": 4615 }, { "epoch": 2.0406884605461983, "grad_norm": 0.2968130160741918, "learning_rate": 2.807270956993933e-06, "loss": 0.2767, "step": 4616 }, { "epoch": 2.0411307264217005, "grad_norm": 0.3500029697854131, "learning_rate": 2.8049596563669936e-06, "loss": 0.2896, "step": 4617 }, { "epoch": 2.0415729922972026, "grad_norm": 0.2933748020534632, "learning_rate": 2.8026489366200282e-06, "loss": 0.2654, "step": 4618 }, { "epoch": 2.0420152581727047, "grad_norm": 0.33345038172755004, "learning_rate": 2.800338798364527e-06, "loss": 0.2549, "step": 4619 }, { "epoch": 2.042457524048207, "grad_norm": 0.30113812002428975, "learning_rate": 2.7980292422118282e-06, "loss": 0.2759, "step": 4620 }, { "epoch": 2.0428997899237094, "grad_norm": 0.3198302802272132, "learning_rate": 2.7957202687731196e-06, "loss": 0.281, "step": 4621 }, { "epoch": 2.0433420557992115, "grad_norm": 0.2747602162528814, "learning_rate": 2.7934118786594244e-06, "loss": 0.2451, "step": 4622 }, { "epoch": 2.0437843216747136, "grad_norm": 0.3673118893821289, "learning_rate": 2.7911040724816253e-06, "loss": 0.2619, "step": 4623 }, { "epoch": 2.0442265875502157, "grad_norm": 0.377154773106409, "learning_rate": 2.7887968508504415e-06, "loss": 0.2732, "step": 4624 }, { "epoch": 2.044668853425718, "grad_norm": 0.3732726976182839, "learning_rate": 2.7864902143764407e-06, "loss": 0.2706, "step": 4625 }, { "epoch": 2.04511111930122, "grad_norm": 0.40251836881278186, "learning_rate": 2.7841841636700338e-06, "loss": 0.2725, "step": 4626 }, { "epoch": 2.045553385176722, "grad_norm": 0.3744062348756464, "learning_rate": 2.7818786993414783e-06, "loss": 0.2747, "step": 4627 }, { "epoch": 2.045995651052224, "grad_norm": 0.28914981423728087, "learning_rate": 2.779573822000876e-06, "loss": 0.2575, "step": 4628 }, { "epoch": 2.0464379169277263, "grad_norm": 0.38014696276288296, "learning_rate": 2.777269532258173e-06, "loss": 0.2467, "step": 4629 }, { "epoch": 2.0468801828032284, "grad_norm": 0.36883198974614784, "learning_rate": 2.7749658307231608e-06, "loss": 0.2555, "step": 4630 }, { "epoch": 2.047322448678731, "grad_norm": 0.3073979049626988, "learning_rate": 2.772662718005474e-06, "loss": 0.2934, "step": 4631 }, { "epoch": 2.047764714554233, "grad_norm": 0.2607531753902034, "learning_rate": 2.770360194714593e-06, "loss": 0.2585, "step": 4632 }, { "epoch": 2.048206980429735, "grad_norm": 0.31672026502314227, "learning_rate": 2.7680582614598384e-06, "loss": 0.2483, "step": 4633 }, { "epoch": 2.0486492463052373, "grad_norm": 0.3495143082490336, "learning_rate": 2.7657569188503817e-06, "loss": 0.2514, "step": 4634 }, { "epoch": 2.0490915121807394, "grad_norm": 0.49538343569833715, "learning_rate": 2.7634561674952333e-06, "loss": 0.2546, "step": 4635 }, { "epoch": 2.0495337780562415, "grad_norm": 0.45355370284663554, "learning_rate": 2.7611560080032475e-06, "loss": 0.2709, "step": 4636 }, { "epoch": 2.0499760439317436, "grad_norm": 0.3051428071038488, "learning_rate": 2.7588564409831197e-06, "loss": 0.2743, "step": 4637 }, { "epoch": 2.0504183098072457, "grad_norm": 0.32556995456152865, "learning_rate": 2.7565574670433914e-06, "loss": 0.2543, "step": 4638 }, { "epoch": 2.050860575682748, "grad_norm": 0.3126874344240025, "learning_rate": 2.75425908679245e-06, "loss": 0.2599, "step": 4639 }, { "epoch": 2.05130284155825, "grad_norm": 0.3746984580204987, "learning_rate": 2.751961300838521e-06, "loss": 0.3165, "step": 4640 }, { "epoch": 2.051745107433752, "grad_norm": 0.3725430385736785, "learning_rate": 2.749664109789674e-06, "loss": 0.2694, "step": 4641 }, { "epoch": 2.0521873733092546, "grad_norm": 0.30845645509799674, "learning_rate": 2.7473675142538213e-06, "loss": 0.2549, "step": 4642 }, { "epoch": 2.0526296391847567, "grad_norm": 0.32879700234010145, "learning_rate": 2.745071514838718e-06, "loss": 0.2835, "step": 4643 }, { "epoch": 2.053071905060259, "grad_norm": 0.3663986322707419, "learning_rate": 2.7427761121519615e-06, "loss": 0.2939, "step": 4644 }, { "epoch": 2.053514170935761, "grad_norm": 0.3171857059508269, "learning_rate": 2.7404813068009904e-06, "loss": 0.2996, "step": 4645 }, { "epoch": 2.053956436811263, "grad_norm": 0.3317395769238811, "learning_rate": 2.7381870993930858e-06, "loss": 0.2766, "step": 4646 }, { "epoch": 2.054398702686765, "grad_norm": 0.36546099364315004, "learning_rate": 2.735893490535371e-06, "loss": 0.2501, "step": 4647 }, { "epoch": 2.0548409685622673, "grad_norm": 0.31344910582398183, "learning_rate": 2.7336004808348094e-06, "loss": 0.2638, "step": 4648 }, { "epoch": 2.0552832344377694, "grad_norm": 0.3266151572509574, "learning_rate": 2.731308070898206e-06, "loss": 0.2426, "step": 4649 }, { "epoch": 2.0557255003132715, "grad_norm": 0.3037578270081548, "learning_rate": 2.729016261332212e-06, "loss": 0.2698, "step": 4650 }, { "epoch": 2.0561677661887736, "grad_norm": 0.3719036849393139, "learning_rate": 2.7267250527433124e-06, "loss": 0.2912, "step": 4651 }, { "epoch": 2.056610032064276, "grad_norm": 0.3856729847044402, "learning_rate": 2.7244344457378403e-06, "loss": 0.2756, "step": 4652 }, { "epoch": 2.0570522979397783, "grad_norm": 0.32687724954409975, "learning_rate": 2.72214444092196e-06, "loss": 0.2872, "step": 4653 }, { "epoch": 2.0574945638152804, "grad_norm": 0.3553770312587507, "learning_rate": 2.7198550389016837e-06, "loss": 0.2661, "step": 4654 }, { "epoch": 2.0579368296907825, "grad_norm": 0.37452386622232037, "learning_rate": 2.7175662402828658e-06, "loss": 0.3048, "step": 4655 }, { "epoch": 2.0583790955662846, "grad_norm": 0.3154783303871395, "learning_rate": 2.7152780456711956e-06, "loss": 0.2547, "step": 4656 }, { "epoch": 2.0588213614417867, "grad_norm": 0.31289013903128987, "learning_rate": 2.7129904556722052e-06, "loss": 0.2499, "step": 4657 }, { "epoch": 2.059263627317289, "grad_norm": 0.4245010064232995, "learning_rate": 2.710703470891266e-06, "loss": 0.3152, "step": 4658 }, { "epoch": 2.059705893192791, "grad_norm": 0.33528421293664507, "learning_rate": 2.7084170919335896e-06, "loss": 0.2801, "step": 4659 }, { "epoch": 2.060148159068293, "grad_norm": 0.3098106474120013, "learning_rate": 2.706131319404227e-06, "loss": 0.2765, "step": 4660 }, { "epoch": 2.060590424943795, "grad_norm": 0.3233613523158064, "learning_rate": 2.7038461539080696e-06, "loss": 0.2412, "step": 4661 }, { "epoch": 2.0610326908192977, "grad_norm": 0.29603121129140214, "learning_rate": 2.7015615960498463e-06, "loss": 0.2716, "step": 4662 }, { "epoch": 2.0614749566948, "grad_norm": 0.30039197819241625, "learning_rate": 2.6992776464341274e-06, "loss": 0.2612, "step": 4663 }, { "epoch": 2.061917222570302, "grad_norm": 0.3148030019695123, "learning_rate": 2.696994305665321e-06, "loss": 0.2522, "step": 4664 }, { "epoch": 2.062359488445804, "grad_norm": 0.29643050968851486, "learning_rate": 2.6947115743476715e-06, "loss": 0.2702, "step": 4665 }, { "epoch": 2.062801754321306, "grad_norm": 0.3144344356950858, "learning_rate": 2.6924294530852702e-06, "loss": 0.2539, "step": 4666 }, { "epoch": 2.0632440201968083, "grad_norm": 0.3309026907789394, "learning_rate": 2.6901479424820396e-06, "loss": 0.2798, "step": 4667 }, { "epoch": 2.0636862860723104, "grad_norm": 0.3570498565553771, "learning_rate": 2.6878670431417405e-06, "loss": 0.2488, "step": 4668 }, { "epoch": 2.0641285519478125, "grad_norm": 0.31624007639293183, "learning_rate": 2.685586755667975e-06, "loss": 0.2798, "step": 4669 }, { "epoch": 2.0645708178233146, "grad_norm": 0.363076234142838, "learning_rate": 2.6833070806641798e-06, "loss": 0.292, "step": 4670 }, { "epoch": 2.0650130836988168, "grad_norm": 0.33348239961384046, "learning_rate": 2.681028018733637e-06, "loss": 0.2562, "step": 4671 }, { "epoch": 2.065455349574319, "grad_norm": 0.33832258358263156, "learning_rate": 2.6787495704794587e-06, "loss": 0.2984, "step": 4672 }, { "epoch": 2.0658976154498214, "grad_norm": 0.3449373324623513, "learning_rate": 2.676471736504598e-06, "loss": 0.2423, "step": 4673 }, { "epoch": 2.0663398813253235, "grad_norm": 0.3976316384387357, "learning_rate": 2.6741945174118444e-06, "loss": 0.2848, "step": 4674 }, { "epoch": 2.0667821472008256, "grad_norm": 0.3866593187726686, "learning_rate": 2.6719179138038252e-06, "loss": 0.2978, "step": 4675 }, { "epoch": 2.0672244130763278, "grad_norm": 0.4307730429565172, "learning_rate": 2.6696419262830052e-06, "loss": 0.2601, "step": 4676 }, { "epoch": 2.06766667895183, "grad_norm": 0.3398458578230752, "learning_rate": 2.667366555451685e-06, "loss": 0.2588, "step": 4677 }, { "epoch": 2.068108944827332, "grad_norm": 0.3172274142518461, "learning_rate": 2.665091801912002e-06, "loss": 0.2468, "step": 4678 }, { "epoch": 2.068551210702834, "grad_norm": 0.3532495396546883, "learning_rate": 2.662817666265932e-06, "loss": 0.2657, "step": 4679 }, { "epoch": 2.068993476578336, "grad_norm": 0.33760865430316034, "learning_rate": 2.660544149115284e-06, "loss": 0.2507, "step": 4680 }, { "epoch": 2.0694357424538383, "grad_norm": 0.36234581859402537, "learning_rate": 2.6582712510617093e-06, "loss": 0.2722, "step": 4681 }, { "epoch": 2.069878008329341, "grad_norm": 0.3358205867742649, "learning_rate": 2.6559989727066903e-06, "loss": 0.2542, "step": 4682 }, { "epoch": 2.070320274204843, "grad_norm": 0.3371865782635844, "learning_rate": 2.6537273146515474e-06, "loss": 0.2407, "step": 4683 }, { "epoch": 2.070762540080345, "grad_norm": 0.3755982206687375, "learning_rate": 2.651456277497433e-06, "loss": 0.2707, "step": 4684 }, { "epoch": 2.071204805955847, "grad_norm": 0.4245284525587184, "learning_rate": 2.6491858618453404e-06, "loss": 0.2865, "step": 4685 }, { "epoch": 2.0716470718313493, "grad_norm": 0.36190296984273834, "learning_rate": 2.6469160682960938e-06, "loss": 0.2844, "step": 4686 }, { "epoch": 2.0720893377068514, "grad_norm": 0.3200674112614676, "learning_rate": 2.6446468974503596e-06, "loss": 0.2689, "step": 4687 }, { "epoch": 2.0725316035823536, "grad_norm": 0.3501593368530092, "learning_rate": 2.6423783499086334e-06, "loss": 0.2559, "step": 4688 }, { "epoch": 2.0729738694578557, "grad_norm": 0.3385233051168512, "learning_rate": 2.6401104262712473e-06, "loss": 0.266, "step": 4689 }, { "epoch": 2.073416135333358, "grad_norm": 0.3272971866228646, "learning_rate": 2.637843127138369e-06, "loss": 0.271, "step": 4690 }, { "epoch": 2.07385840120886, "grad_norm": 0.34002206016824266, "learning_rate": 2.63557645311e-06, "loss": 0.2854, "step": 4691 }, { "epoch": 2.074300667084362, "grad_norm": 0.39166416704306983, "learning_rate": 2.633310404785977e-06, "loss": 0.2749, "step": 4692 }, { "epoch": 2.0747429329598646, "grad_norm": 0.39695311475192474, "learning_rate": 2.6310449827659713e-06, "loss": 0.3017, "step": 4693 }, { "epoch": 2.0751851988353667, "grad_norm": 0.3424182135151136, "learning_rate": 2.628780187649488e-06, "loss": 0.2865, "step": 4694 }, { "epoch": 2.075627464710869, "grad_norm": 0.3382586536011371, "learning_rate": 2.626516020035866e-06, "loss": 0.2839, "step": 4695 }, { "epoch": 2.076069730586371, "grad_norm": 0.3700004020023534, "learning_rate": 2.6242524805242787e-06, "loss": 0.2812, "step": 4696 }, { "epoch": 2.076511996461873, "grad_norm": 0.3249204238572804, "learning_rate": 2.6219895697137343e-06, "loss": 0.2715, "step": 4697 }, { "epoch": 2.076954262337375, "grad_norm": 0.3996068415981633, "learning_rate": 2.619727288203075e-06, "loss": 0.2821, "step": 4698 }, { "epoch": 2.0773965282128772, "grad_norm": 0.45970263687392704, "learning_rate": 2.6174656365909713e-06, "loss": 0.2768, "step": 4699 }, { "epoch": 2.0778387940883793, "grad_norm": 0.3090783353341797, "learning_rate": 2.6152046154759316e-06, "loss": 0.3017, "step": 4700 }, { "epoch": 2.0782810599638815, "grad_norm": 0.4017116066488864, "learning_rate": 2.6129442254562977e-06, "loss": 0.2852, "step": 4701 }, { "epoch": 2.0787233258393836, "grad_norm": 0.32137013478351545, "learning_rate": 2.6106844671302406e-06, "loss": 0.2517, "step": 4702 }, { "epoch": 2.079165591714886, "grad_norm": 0.27911116264972013, "learning_rate": 2.6084253410957713e-06, "loss": 0.2827, "step": 4703 }, { "epoch": 2.0796078575903882, "grad_norm": 0.374726255877934, "learning_rate": 2.606166847950726e-06, "loss": 0.2716, "step": 4704 }, { "epoch": 2.0800501234658904, "grad_norm": 0.3073678759186685, "learning_rate": 2.6039089882927775e-06, "loss": 0.2613, "step": 4705 }, { "epoch": 2.0804923893413925, "grad_norm": 0.3586578517421295, "learning_rate": 2.6016517627194293e-06, "loss": 0.2707, "step": 4706 }, { "epoch": 2.0809346552168946, "grad_norm": 0.5175297078025161, "learning_rate": 2.5993951718280175e-06, "loss": 0.2753, "step": 4707 }, { "epoch": 2.0813769210923967, "grad_norm": 0.2886883493389791, "learning_rate": 2.5971392162157105e-06, "loss": 0.2443, "step": 4708 }, { "epoch": 2.081819186967899, "grad_norm": 0.3159914429544373, "learning_rate": 2.5948838964795087e-06, "loss": 0.2632, "step": 4709 }, { "epoch": 2.082261452843401, "grad_norm": 0.3339611726442776, "learning_rate": 2.5926292132162432e-06, "loss": 0.2635, "step": 4710 }, { "epoch": 2.082703718718903, "grad_norm": 0.3255681255814294, "learning_rate": 2.5903751670225785e-06, "loss": 0.3025, "step": 4711 }, { "epoch": 2.083145984594405, "grad_norm": 0.331005434034775, "learning_rate": 2.588121758495007e-06, "loss": 0.2823, "step": 4712 }, { "epoch": 2.0835882504699077, "grad_norm": 0.3126973411349585, "learning_rate": 2.585868988229859e-06, "loss": 0.2918, "step": 4713 }, { "epoch": 2.08403051634541, "grad_norm": 0.3052712535372023, "learning_rate": 2.583616856823291e-06, "loss": 0.2416, "step": 4714 }, { "epoch": 2.084472782220912, "grad_norm": 0.4120234047287234, "learning_rate": 2.581365364871288e-06, "loss": 0.2975, "step": 4715 }, { "epoch": 2.084915048096414, "grad_norm": 0.3220994242745793, "learning_rate": 2.5791145129696703e-06, "loss": 0.2808, "step": 4716 }, { "epoch": 2.085357313971916, "grad_norm": 0.3379559901876017, "learning_rate": 2.576864301714086e-06, "loss": 0.2766, "step": 4717 }, { "epoch": 2.0857995798474183, "grad_norm": 0.3138767000099584, "learning_rate": 2.574614731700019e-06, "loss": 0.2629, "step": 4718 }, { "epoch": 2.0862418457229204, "grad_norm": 0.31741079575656855, "learning_rate": 2.572365803522777e-06, "loss": 0.2673, "step": 4719 }, { "epoch": 2.0866841115984225, "grad_norm": 0.3426437561549992, "learning_rate": 2.570117517777502e-06, "loss": 0.2726, "step": 4720 }, { "epoch": 2.0871263774739246, "grad_norm": 0.40696267600060143, "learning_rate": 2.5678698750591624e-06, "loss": 0.273, "step": 4721 }, { "epoch": 2.0875686433494267, "grad_norm": 0.3745415062184489, "learning_rate": 2.565622875962559e-06, "loss": 0.2511, "step": 4722 }, { "epoch": 2.0880109092249293, "grad_norm": 0.3524091417446308, "learning_rate": 2.5633765210823237e-06, "loss": 0.266, "step": 4723 }, { "epoch": 2.0884531751004314, "grad_norm": 0.4137980677862906, "learning_rate": 2.561130811012913e-06, "loss": 0.2379, "step": 4724 }, { "epoch": 2.0888954409759335, "grad_norm": 0.5300029417452087, "learning_rate": 2.5588857463486184e-06, "loss": 0.2681, "step": 4725 }, { "epoch": 2.0893377068514356, "grad_norm": 0.36505409916506737, "learning_rate": 2.5566413276835568e-06, "loss": 0.2911, "step": 4726 }, { "epoch": 2.0897799727269377, "grad_norm": 0.28639874149820505, "learning_rate": 2.5543975556116744e-06, "loss": 0.2553, "step": 4727 }, { "epoch": 2.09022223860244, "grad_norm": 0.3035082599826608, "learning_rate": 2.552154430726747e-06, "loss": 0.2527, "step": 4728 }, { "epoch": 2.090664504477942, "grad_norm": 0.2830533012703125, "learning_rate": 2.549911953622384e-06, "loss": 0.2394, "step": 4729 }, { "epoch": 2.091106770353444, "grad_norm": 0.35991737297875714, "learning_rate": 2.547670124892013e-06, "loss": 0.2642, "step": 4730 }, { "epoch": 2.091549036228946, "grad_norm": 0.27776852082074305, "learning_rate": 2.545428945128899e-06, "loss": 0.2413, "step": 4731 }, { "epoch": 2.0919913021044483, "grad_norm": 0.35591625170199775, "learning_rate": 2.54318841492613e-06, "loss": 0.2637, "step": 4732 }, { "epoch": 2.092433567979951, "grad_norm": 0.31599400689189866, "learning_rate": 2.5409485348766234e-06, "loss": 0.2617, "step": 4733 }, { "epoch": 2.092875833855453, "grad_norm": 0.4004020156385748, "learning_rate": 2.538709305573128e-06, "loss": 0.3013, "step": 4734 }, { "epoch": 2.093318099730955, "grad_norm": 0.42367704779473986, "learning_rate": 2.5364707276082175e-06, "loss": 0.2565, "step": 4735 }, { "epoch": 2.093760365606457, "grad_norm": 0.30874953850688164, "learning_rate": 2.534232801574291e-06, "loss": 0.2939, "step": 4736 }, { "epoch": 2.0942026314819593, "grad_norm": 0.3509690872889521, "learning_rate": 2.53199552806358e-06, "loss": 0.2502, "step": 4737 }, { "epoch": 2.0946448973574614, "grad_norm": 0.31131703166085495, "learning_rate": 2.5297589076681405e-06, "loss": 0.2565, "step": 4738 }, { "epoch": 2.0950871632329635, "grad_norm": 0.37344975756369153, "learning_rate": 2.5275229409798517e-06, "loss": 0.3191, "step": 4739 }, { "epoch": 2.0955294291084656, "grad_norm": 0.3778675682239416, "learning_rate": 2.5252876285904283e-06, "loss": 0.2502, "step": 4740 }, { "epoch": 2.0959716949839677, "grad_norm": 0.3947265529922578, "learning_rate": 2.5230529710914074e-06, "loss": 0.2879, "step": 4741 }, { "epoch": 2.09641396085947, "grad_norm": 0.31815281596491807, "learning_rate": 2.5208189690741515e-06, "loss": 0.2696, "step": 4742 }, { "epoch": 2.096856226734972, "grad_norm": 0.346159077527981, "learning_rate": 2.5185856231298522e-06, "loss": 0.2553, "step": 4743 }, { "epoch": 2.0972984926104745, "grad_norm": 0.42069530560510743, "learning_rate": 2.516352933849523e-06, "loss": 0.2969, "step": 4744 }, { "epoch": 2.0977407584859766, "grad_norm": 0.4313461936581601, "learning_rate": 2.514120901824015e-06, "loss": 0.2768, "step": 4745 }, { "epoch": 2.0981830243614787, "grad_norm": 0.39152962917415685, "learning_rate": 2.5118895276439893e-06, "loss": 0.3109, "step": 4746 }, { "epoch": 2.098625290236981, "grad_norm": 0.33775623818149114, "learning_rate": 2.509658811899944e-06, "loss": 0.2569, "step": 4747 }, { "epoch": 2.099067556112483, "grad_norm": 0.28286226661572217, "learning_rate": 2.5074287551822007e-06, "loss": 0.2673, "step": 4748 }, { "epoch": 2.099509821987985, "grad_norm": 0.341306161753182, "learning_rate": 2.5051993580809033e-06, "loss": 0.2843, "step": 4749 }, { "epoch": 2.099952087863487, "grad_norm": 0.3369462449846202, "learning_rate": 2.502970621186027e-06, "loss": 0.2498, "step": 4750 }, { "epoch": 2.1003943537389893, "grad_norm": 0.3322491533861887, "learning_rate": 2.5007425450873677e-06, "loss": 0.2611, "step": 4751 }, { "epoch": 2.1008366196144914, "grad_norm": 0.3430604150419479, "learning_rate": 2.4985151303745477e-06, "loss": 0.2659, "step": 4752 }, { "epoch": 2.1012788854899935, "grad_norm": 0.2898135680743732, "learning_rate": 2.4962883776370135e-06, "loss": 0.2751, "step": 4753 }, { "epoch": 2.101721151365496, "grad_norm": 0.36529019628643705, "learning_rate": 2.494062287464038e-06, "loss": 0.2707, "step": 4754 }, { "epoch": 2.102163417240998, "grad_norm": 0.3282989220077298, "learning_rate": 2.491836860444718e-06, "loss": 0.2516, "step": 4755 }, { "epoch": 2.1026056831165003, "grad_norm": 0.6857888886900916, "learning_rate": 2.4896120971679743e-06, "loss": 0.3167, "step": 4756 }, { "epoch": 2.1030479489920024, "grad_norm": 0.3906573239767201, "learning_rate": 2.4873879982225536e-06, "loss": 0.2411, "step": 4757 }, { "epoch": 2.1034902148675045, "grad_norm": 0.3283641073508249, "learning_rate": 2.4851645641970247e-06, "loss": 0.2839, "step": 4758 }, { "epoch": 2.1039324807430066, "grad_norm": 0.3180238116489658, "learning_rate": 2.4829417956797813e-06, "loss": 0.2802, "step": 4759 }, { "epoch": 2.1043747466185088, "grad_norm": 0.35728977094131903, "learning_rate": 2.4807196932590417e-06, "loss": 0.2289, "step": 4760 }, { "epoch": 2.104817012494011, "grad_norm": 0.3309350452214471, "learning_rate": 2.4784982575228476e-06, "loss": 0.2513, "step": 4761 }, { "epoch": 2.105259278369513, "grad_norm": 0.31968194927149163, "learning_rate": 2.4762774890590635e-06, "loss": 0.2675, "step": 4762 }, { "epoch": 2.105701544245015, "grad_norm": 0.3550842344146324, "learning_rate": 2.4740573884553782e-06, "loss": 0.274, "step": 4763 }, { "epoch": 2.1061438101205177, "grad_norm": 0.3740789810078847, "learning_rate": 2.471837956299303e-06, "loss": 0.2649, "step": 4764 }, { "epoch": 2.1065860759960198, "grad_norm": 0.3012778963675684, "learning_rate": 2.469619193178172e-06, "loss": 0.2828, "step": 4765 }, { "epoch": 2.107028341871522, "grad_norm": 0.2765240043800179, "learning_rate": 2.4674010996791457e-06, "loss": 0.2571, "step": 4766 }, { "epoch": 2.107470607747024, "grad_norm": 0.38104738387134524, "learning_rate": 2.4651836763892034e-06, "loss": 0.2371, "step": 4767 }, { "epoch": 2.107912873622526, "grad_norm": 0.4920025049665832, "learning_rate": 2.462966923895148e-06, "loss": 0.2688, "step": 4768 }, { "epoch": 2.108355139498028, "grad_norm": 0.32207067024581904, "learning_rate": 2.4607508427836067e-06, "loss": 0.2527, "step": 4769 }, { "epoch": 2.1087974053735303, "grad_norm": 0.3419445732958144, "learning_rate": 2.4585354336410226e-06, "loss": 0.2733, "step": 4770 }, { "epoch": 2.1092396712490324, "grad_norm": 0.40145537487388067, "learning_rate": 2.456320697053672e-06, "loss": 0.2943, "step": 4771 }, { "epoch": 2.1096819371245346, "grad_norm": 0.2988665883366237, "learning_rate": 2.4541066336076434e-06, "loss": 0.2647, "step": 4772 }, { "epoch": 2.1101242030000367, "grad_norm": 0.38149074738311206, "learning_rate": 2.451893243888853e-06, "loss": 0.2485, "step": 4773 }, { "epoch": 2.1105664688755392, "grad_norm": 0.48840793210247246, "learning_rate": 2.4496805284830355e-06, "loss": 0.2571, "step": 4774 }, { "epoch": 2.1110087347510413, "grad_norm": 0.30031183117859545, "learning_rate": 2.447468487975749e-06, "loss": 0.2708, "step": 4775 }, { "epoch": 2.1114510006265435, "grad_norm": 0.3687054135160842, "learning_rate": 2.445257122952371e-06, "loss": 0.253, "step": 4776 }, { "epoch": 2.1118932665020456, "grad_norm": 0.3493295790429714, "learning_rate": 2.4430464339981025e-06, "loss": 0.2856, "step": 4777 }, { "epoch": 2.1123355323775477, "grad_norm": 0.32280362199088647, "learning_rate": 2.4408364216979648e-06, "loss": 0.2708, "step": 4778 }, { "epoch": 2.11277779825305, "grad_norm": 0.3115148970468465, "learning_rate": 2.438627086636799e-06, "loss": 0.3046, "step": 4779 }, { "epoch": 2.113220064128552, "grad_norm": 0.36697699956736407, "learning_rate": 2.436418429399269e-06, "loss": 0.2892, "step": 4780 }, { "epoch": 2.113662330004054, "grad_norm": 0.3850896828584372, "learning_rate": 2.4342104505698553e-06, "loss": 0.2426, "step": 4781 }, { "epoch": 2.114104595879556, "grad_norm": 0.2884168818948159, "learning_rate": 2.432003150732867e-06, "loss": 0.2398, "step": 4782 }, { "epoch": 2.1145468617550582, "grad_norm": 0.30582565186734995, "learning_rate": 2.4297965304724245e-06, "loss": 0.2603, "step": 4783 }, { "epoch": 2.114989127630561, "grad_norm": 0.35992857586083915, "learning_rate": 2.4275905903724745e-06, "loss": 0.2409, "step": 4784 }, { "epoch": 2.115431393506063, "grad_norm": 0.32790817063162464, "learning_rate": 2.4253853310167814e-06, "loss": 0.2735, "step": 4785 }, { "epoch": 2.115873659381565, "grad_norm": 0.36305723874895074, "learning_rate": 2.4231807529889245e-06, "loss": 0.2569, "step": 4786 }, { "epoch": 2.116315925257067, "grad_norm": 0.31810134458824624, "learning_rate": 2.4209768568723136e-06, "loss": 0.2783, "step": 4787 }, { "epoch": 2.1167581911325692, "grad_norm": 0.3583267711590064, "learning_rate": 2.41877364325017e-06, "loss": 0.2869, "step": 4788 }, { "epoch": 2.1172004570080714, "grad_norm": 0.3416973668139652, "learning_rate": 2.416571112705537e-06, "loss": 0.2432, "step": 4789 }, { "epoch": 2.1176427228835735, "grad_norm": 0.4223866157940136, "learning_rate": 2.414369265821276e-06, "loss": 0.2584, "step": 4790 }, { "epoch": 2.1180849887590756, "grad_norm": 0.291672792024237, "learning_rate": 2.4121681031800686e-06, "loss": 0.2717, "step": 4791 }, { "epoch": 2.1185272546345777, "grad_norm": 0.30593083490186823, "learning_rate": 2.4099676253644156e-06, "loss": 0.2361, "step": 4792 }, { "epoch": 2.11896952051008, "grad_norm": 0.37861727787364596, "learning_rate": 2.4077678329566346e-06, "loss": 0.2647, "step": 4793 }, { "epoch": 2.119411786385582, "grad_norm": 0.3493804185209996, "learning_rate": 2.4055687265388643e-06, "loss": 0.2669, "step": 4794 }, { "epoch": 2.1198540522610845, "grad_norm": 0.3028309996823545, "learning_rate": 2.4033703066930604e-06, "loss": 0.2547, "step": 4795 }, { "epoch": 2.1202963181365866, "grad_norm": 0.37207194725499676, "learning_rate": 2.401172574000996e-06, "loss": 0.2565, "step": 4796 }, { "epoch": 2.1207385840120887, "grad_norm": 0.2979681840531237, "learning_rate": 2.3989755290442633e-06, "loss": 0.2507, "step": 4797 }, { "epoch": 2.121180849887591, "grad_norm": 0.40076295135343415, "learning_rate": 2.396779172404276e-06, "loss": 0.2625, "step": 4798 }, { "epoch": 2.121623115763093, "grad_norm": 0.37100037198646646, "learning_rate": 2.3945835046622605e-06, "loss": 0.2805, "step": 4799 }, { "epoch": 2.122065381638595, "grad_norm": 0.3232720289230613, "learning_rate": 2.392388526399264e-06, "loss": 0.2785, "step": 4800 }, { "epoch": 2.122507647514097, "grad_norm": 0.38678365260162095, "learning_rate": 2.3901942381961475e-06, "loss": 0.2908, "step": 4801 }, { "epoch": 2.1229499133895993, "grad_norm": 0.29286584683048295, "learning_rate": 2.388000640633591e-06, "loss": 0.2776, "step": 4802 }, { "epoch": 2.1233921792651014, "grad_norm": 0.40186480565692173, "learning_rate": 2.385807734292097e-06, "loss": 0.2995, "step": 4803 }, { "epoch": 2.1238344451406035, "grad_norm": 0.3533622510147097, "learning_rate": 2.3836155197519785e-06, "loss": 0.2729, "step": 4804 }, { "epoch": 2.124276711016106, "grad_norm": 0.35102251065662154, "learning_rate": 2.3814239975933676e-06, "loss": 0.268, "step": 4805 }, { "epoch": 2.124718976891608, "grad_norm": 0.37173269804147496, "learning_rate": 2.3792331683962137e-06, "loss": 0.2961, "step": 4806 }, { "epoch": 2.1251612427671103, "grad_norm": 0.37941024539718277, "learning_rate": 2.3770430327402823e-06, "loss": 0.2634, "step": 4807 }, { "epoch": 2.1256035086426124, "grad_norm": 0.29270172041249853, "learning_rate": 2.3748535912051546e-06, "loss": 0.2763, "step": 4808 }, { "epoch": 2.1260457745181145, "grad_norm": 0.35506138636245865, "learning_rate": 2.3726648443702293e-06, "loss": 0.2554, "step": 4809 }, { "epoch": 2.1264880403936166, "grad_norm": 0.27905802837159843, "learning_rate": 2.370476792814722e-06, "loss": 0.2552, "step": 4810 }, { "epoch": 2.1269303062691187, "grad_norm": 0.3594597168137774, "learning_rate": 2.3682894371176617e-06, "loss": 0.2747, "step": 4811 }, { "epoch": 2.127372572144621, "grad_norm": 0.36781595480997276, "learning_rate": 2.366102777857896e-06, "loss": 0.2828, "step": 4812 }, { "epoch": 2.127814838020123, "grad_norm": 0.2961230083940767, "learning_rate": 2.363916815614084e-06, "loss": 0.254, "step": 4813 }, { "epoch": 2.128257103895625, "grad_norm": 0.31633456959953676, "learning_rate": 2.3617315509647083e-06, "loss": 0.2637, "step": 4814 }, { "epoch": 2.1286993697711276, "grad_norm": 0.2765836992152342, "learning_rate": 2.3595469844880585e-06, "loss": 0.2651, "step": 4815 }, { "epoch": 2.1291416356466297, "grad_norm": 0.3892291373905264, "learning_rate": 2.357363116762246e-06, "loss": 0.2702, "step": 4816 }, { "epoch": 2.129583901522132, "grad_norm": 0.4055723589971986, "learning_rate": 2.3551799483651894e-06, "loss": 0.2662, "step": 4817 }, { "epoch": 2.130026167397634, "grad_norm": 0.35874556270839886, "learning_rate": 2.352997479874628e-06, "loss": 0.2623, "step": 4818 }, { "epoch": 2.130468433273136, "grad_norm": 0.3438125650222666, "learning_rate": 2.350815711868118e-06, "loss": 0.2873, "step": 4819 }, { "epoch": 2.130910699148638, "grad_norm": 0.28507776515271394, "learning_rate": 2.3486346449230247e-06, "loss": 0.2581, "step": 4820 }, { "epoch": 2.1313529650241403, "grad_norm": 0.38343870086014475, "learning_rate": 2.3464542796165313e-06, "loss": 0.2367, "step": 4821 }, { "epoch": 2.1317952308996424, "grad_norm": 0.3482897357673417, "learning_rate": 2.3442746165256335e-06, "loss": 0.2339, "step": 4822 }, { "epoch": 2.1322374967751445, "grad_norm": 0.3760742533282648, "learning_rate": 2.3420956562271424e-06, "loss": 0.2755, "step": 4823 }, { "epoch": 2.1326797626506466, "grad_norm": 0.278544641720074, "learning_rate": 2.3399173992976825e-06, "loss": 0.2685, "step": 4824 }, { "epoch": 2.1331220285261487, "grad_norm": 0.3890586764087674, "learning_rate": 2.337739846313692e-06, "loss": 0.2337, "step": 4825 }, { "epoch": 2.1335642944016513, "grad_norm": 0.30753635213167796, "learning_rate": 2.335562997851424e-06, "loss": 0.2587, "step": 4826 }, { "epoch": 2.1340065602771534, "grad_norm": 0.3814812003275398, "learning_rate": 2.333386854486943e-06, "loss": 0.261, "step": 4827 }, { "epoch": 2.1344488261526555, "grad_norm": 0.31658324042400543, "learning_rate": 2.33121141679613e-06, "loss": 0.2906, "step": 4828 }, { "epoch": 2.1348910920281576, "grad_norm": 0.2918931907783346, "learning_rate": 2.3290366853546735e-06, "loss": 0.2455, "step": 4829 }, { "epoch": 2.1353333579036597, "grad_norm": 0.34575177110219274, "learning_rate": 2.3268626607380844e-06, "loss": 0.2502, "step": 4830 }, { "epoch": 2.135775623779162, "grad_norm": 0.469642622570989, "learning_rate": 2.3246893435216803e-06, "loss": 0.291, "step": 4831 }, { "epoch": 2.136217889654664, "grad_norm": 0.33155824836327663, "learning_rate": 2.322516734280589e-06, "loss": 0.267, "step": 4832 }, { "epoch": 2.136660155530166, "grad_norm": 0.32720650081085967, "learning_rate": 2.3203448335897572e-06, "loss": 0.2458, "step": 4833 }, { "epoch": 2.137102421405668, "grad_norm": 0.31815899782478874, "learning_rate": 2.318173642023939e-06, "loss": 0.25, "step": 4834 }, { "epoch": 2.1375446872811708, "grad_norm": 0.329085203925662, "learning_rate": 2.316003160157706e-06, "loss": 0.265, "step": 4835 }, { "epoch": 2.137986953156673, "grad_norm": 0.270727840489417, "learning_rate": 2.313833388565438e-06, "loss": 0.2508, "step": 4836 }, { "epoch": 2.138429219032175, "grad_norm": 0.31804451402540757, "learning_rate": 2.3116643278213286e-06, "loss": 0.2932, "step": 4837 }, { "epoch": 2.138871484907677, "grad_norm": 0.6720499380760638, "learning_rate": 2.3094959784993815e-06, "loss": 0.2968, "step": 4838 }, { "epoch": 2.139313750783179, "grad_norm": 0.4155216565922103, "learning_rate": 2.3073283411734145e-06, "loss": 0.2695, "step": 4839 }, { "epoch": 2.1397560166586813, "grad_norm": 0.3634323904144035, "learning_rate": 2.305161416417055e-06, "loss": 0.2883, "step": 4840 }, { "epoch": 2.1401982825341834, "grad_norm": 0.46091837966209076, "learning_rate": 2.3029952048037434e-06, "loss": 0.2669, "step": 4841 }, { "epoch": 2.1406405484096855, "grad_norm": 0.323381274836839, "learning_rate": 2.3008297069067297e-06, "loss": 0.2523, "step": 4842 }, { "epoch": 2.1410828142851877, "grad_norm": 0.3627170560310965, "learning_rate": 2.298664923299077e-06, "loss": 0.2326, "step": 4843 }, { "epoch": 2.1415250801606898, "grad_norm": 0.3741338253385395, "learning_rate": 2.296500854553656e-06, "loss": 0.254, "step": 4844 }, { "epoch": 2.141967346036192, "grad_norm": 0.30559894023398465, "learning_rate": 2.294337501243154e-06, "loss": 0.2787, "step": 4845 }, { "epoch": 2.1424096119116944, "grad_norm": 0.35075627769107426, "learning_rate": 2.292174863940064e-06, "loss": 0.2734, "step": 4846 }, { "epoch": 2.1428518777871965, "grad_norm": 0.24969950375122407, "learning_rate": 2.2900129432166936e-06, "loss": 0.2367, "step": 4847 }, { "epoch": 2.1432941436626987, "grad_norm": 0.28348321277618294, "learning_rate": 2.2878517396451533e-06, "loss": 0.2734, "step": 4848 }, { "epoch": 2.1437364095382008, "grad_norm": 0.34836501536621106, "learning_rate": 2.285691253797371e-06, "loss": 0.2819, "step": 4849 }, { "epoch": 2.144178675413703, "grad_norm": 0.31862142211659783, "learning_rate": 2.283531486245081e-06, "loss": 0.2579, "step": 4850 }, { "epoch": 2.144620941289205, "grad_norm": 0.3112415594505847, "learning_rate": 2.2813724375598318e-06, "loss": 0.2625, "step": 4851 }, { "epoch": 2.145063207164707, "grad_norm": 0.29078360626382344, "learning_rate": 2.2792141083129775e-06, "loss": 0.2954, "step": 4852 }, { "epoch": 2.145505473040209, "grad_norm": 0.39854838042856583, "learning_rate": 2.2770564990756833e-06, "loss": 0.2775, "step": 4853 }, { "epoch": 2.1459477389157113, "grad_norm": 0.2787327905450492, "learning_rate": 2.274899610418923e-06, "loss": 0.2765, "step": 4854 }, { "epoch": 2.1463900047912134, "grad_norm": 0.34954907361629917, "learning_rate": 2.2727434429134815e-06, "loss": 0.2771, "step": 4855 }, { "epoch": 2.146832270666716, "grad_norm": 0.3901619635719988, "learning_rate": 2.2705879971299504e-06, "loss": 0.2604, "step": 4856 }, { "epoch": 2.147274536542218, "grad_norm": 0.26985445050602724, "learning_rate": 2.2684332736387328e-06, "loss": 0.2393, "step": 4857 }, { "epoch": 2.1477168024177202, "grad_norm": 0.3333561756504513, "learning_rate": 2.2662792730100385e-06, "loss": 0.2812, "step": 4858 }, { "epoch": 2.1481590682932223, "grad_norm": 0.37913753865774946, "learning_rate": 2.2641259958138877e-06, "loss": 0.268, "step": 4859 }, { "epoch": 2.1486013341687245, "grad_norm": 0.31661192862979687, "learning_rate": 2.261973442620106e-06, "loss": 0.277, "step": 4860 }, { "epoch": 2.1490436000442266, "grad_norm": 0.34834441741253336, "learning_rate": 2.259821613998335e-06, "loss": 0.245, "step": 4861 }, { "epoch": 2.1494858659197287, "grad_norm": 0.35312336348891066, "learning_rate": 2.2576705105180174e-06, "loss": 0.2387, "step": 4862 }, { "epoch": 2.149928131795231, "grad_norm": 0.3695898259952621, "learning_rate": 2.255520132748404e-06, "loss": 0.2429, "step": 4863 }, { "epoch": 2.150370397670733, "grad_norm": 0.2824484878749177, "learning_rate": 2.2533704812585567e-06, "loss": 0.2453, "step": 4864 }, { "epoch": 2.150812663546235, "grad_norm": 0.36385017962955546, "learning_rate": 2.251221556617344e-06, "loss": 0.2573, "step": 4865 }, { "epoch": 2.1512549294217376, "grad_norm": 0.292574324291189, "learning_rate": 2.2490733593934407e-06, "loss": 0.2736, "step": 4866 }, { "epoch": 2.1516971952972397, "grad_norm": 0.28204793366286435, "learning_rate": 2.2469258901553333e-06, "loss": 0.2839, "step": 4867 }, { "epoch": 2.152139461172742, "grad_norm": 0.3592076706696665, "learning_rate": 2.244779149471312e-06, "loss": 0.3059, "step": 4868 }, { "epoch": 2.152581727048244, "grad_norm": 0.36223092943384677, "learning_rate": 2.242633137909475e-06, "loss": 0.2496, "step": 4869 }, { "epoch": 2.153023992923746, "grad_norm": 0.34693902235669244, "learning_rate": 2.2404878560377272e-06, "loss": 0.2277, "step": 4870 }, { "epoch": 2.153466258799248, "grad_norm": 0.27734154290071517, "learning_rate": 2.238343304423782e-06, "loss": 0.245, "step": 4871 }, { "epoch": 2.1539085246747502, "grad_norm": 0.3246488709534017, "learning_rate": 2.2361994836351568e-06, "loss": 0.259, "step": 4872 }, { "epoch": 2.1543507905502524, "grad_norm": 0.3580765893470807, "learning_rate": 2.234056394239178e-06, "loss": 0.2714, "step": 4873 }, { "epoch": 2.1547930564257545, "grad_norm": 0.3776506352947795, "learning_rate": 2.2319140368029773e-06, "loss": 0.2575, "step": 4874 }, { "epoch": 2.1552353223012566, "grad_norm": 0.33829612904790973, "learning_rate": 2.2297724118934933e-06, "loss": 0.2625, "step": 4875 }, { "epoch": 2.1556775881767587, "grad_norm": 0.36302738462126255, "learning_rate": 2.227631520077469e-06, "loss": 0.2434, "step": 4876 }, { "epoch": 2.1561198540522613, "grad_norm": 0.388310196773375, "learning_rate": 2.225491361921458e-06, "loss": 0.281, "step": 4877 }, { "epoch": 2.1565621199277634, "grad_norm": 0.383554939606563, "learning_rate": 2.2233519379918165e-06, "loss": 0.2868, "step": 4878 }, { "epoch": 2.1570043858032655, "grad_norm": 0.335999138692224, "learning_rate": 2.2212132488547035e-06, "loss": 0.2314, "step": 4879 }, { "epoch": 2.1574466516787676, "grad_norm": 0.32857457360573983, "learning_rate": 2.219075295076088e-06, "loss": 0.267, "step": 4880 }, { "epoch": 2.1578889175542697, "grad_norm": 0.39224469068138007, "learning_rate": 2.2169380772217412e-06, "loss": 0.2947, "step": 4881 }, { "epoch": 2.158331183429772, "grad_norm": 0.29533265773217, "learning_rate": 2.214801595857245e-06, "loss": 0.271, "step": 4882 }, { "epoch": 2.158773449305274, "grad_norm": 0.34048701896849914, "learning_rate": 2.212665851547981e-06, "loss": 0.26, "step": 4883 }, { "epoch": 2.159215715180776, "grad_norm": 0.3692706510405416, "learning_rate": 2.210530844859138e-06, "loss": 0.2772, "step": 4884 }, { "epoch": 2.159657981056278, "grad_norm": 0.3575284328539077, "learning_rate": 2.208396576355708e-06, "loss": 0.2915, "step": 4885 }, { "epoch": 2.1601002469317807, "grad_norm": 0.2970520411121747, "learning_rate": 2.2062630466024894e-06, "loss": 0.2789, "step": 4886 }, { "epoch": 2.160542512807283, "grad_norm": 0.31474019070431586, "learning_rate": 2.204130256164084e-06, "loss": 0.2538, "step": 4887 }, { "epoch": 2.160984778682785, "grad_norm": 0.36946202376846965, "learning_rate": 2.201998205604899e-06, "loss": 0.263, "step": 4888 }, { "epoch": 2.161427044558287, "grad_norm": 0.3664020761256548, "learning_rate": 2.199866895489145e-06, "loss": 0.2594, "step": 4889 }, { "epoch": 2.161869310433789, "grad_norm": 0.4083013836635064, "learning_rate": 2.1977363263808366e-06, "loss": 0.2809, "step": 4890 }, { "epoch": 2.1623115763092913, "grad_norm": 0.3887511685799158, "learning_rate": 2.195606498843793e-06, "loss": 0.289, "step": 4891 }, { "epoch": 2.1627538421847934, "grad_norm": 0.31034552190280756, "learning_rate": 2.193477413441634e-06, "loss": 0.2805, "step": 4892 }, { "epoch": 2.1631961080602955, "grad_norm": 0.35583603500858524, "learning_rate": 2.191349070737793e-06, "loss": 0.2418, "step": 4893 }, { "epoch": 2.1636383739357976, "grad_norm": 0.375563674967093, "learning_rate": 2.189221471295492e-06, "loss": 0.2634, "step": 4894 }, { "epoch": 2.1640806398112997, "grad_norm": 0.33137496083446605, "learning_rate": 2.187094615677767e-06, "loss": 0.2589, "step": 4895 }, { "epoch": 2.164522905686802, "grad_norm": 0.3438503204144167, "learning_rate": 2.184968504447453e-06, "loss": 0.3065, "step": 4896 }, { "epoch": 2.1649651715623044, "grad_norm": 0.38800748147890385, "learning_rate": 2.1828431381671883e-06, "loss": 0.3205, "step": 4897 }, { "epoch": 2.1654074374378065, "grad_norm": 0.3778712750610269, "learning_rate": 2.1807185173994183e-06, "loss": 0.2705, "step": 4898 }, { "epoch": 2.1658497033133086, "grad_norm": 0.4000875764271323, "learning_rate": 2.1785946427063857e-06, "loss": 0.2682, "step": 4899 }, { "epoch": 2.1662919691888107, "grad_norm": 0.31652509780071797, "learning_rate": 2.1764715146501376e-06, "loss": 0.2618, "step": 4900 }, { "epoch": 2.166734235064313, "grad_norm": 0.39375443092308615, "learning_rate": 2.1743491337925233e-06, "loss": 0.2892, "step": 4901 }, { "epoch": 2.167176500939815, "grad_norm": 0.3780354267387007, "learning_rate": 2.172227500695197e-06, "loss": 0.2427, "step": 4902 }, { "epoch": 2.167618766815317, "grad_norm": 0.428453623491721, "learning_rate": 2.1701066159196065e-06, "loss": 0.294, "step": 4903 }, { "epoch": 2.168061032690819, "grad_norm": 0.3810813061291085, "learning_rate": 2.1679864800270143e-06, "loss": 0.2703, "step": 4904 }, { "epoch": 2.1685032985663213, "grad_norm": 0.32932461081153935, "learning_rate": 2.165867093578476e-06, "loss": 0.2371, "step": 4905 }, { "epoch": 2.1689455644418234, "grad_norm": 0.2948538653065867, "learning_rate": 2.1637484571348507e-06, "loss": 0.2764, "step": 4906 }, { "epoch": 2.169387830317326, "grad_norm": 0.311960366764367, "learning_rate": 2.161630571256799e-06, "loss": 0.2677, "step": 4907 }, { "epoch": 2.169830096192828, "grad_norm": 0.5926364351574327, "learning_rate": 2.159513436504783e-06, "loss": 0.2691, "step": 4908 }, { "epoch": 2.17027236206833, "grad_norm": 0.3668267729988962, "learning_rate": 2.1573970534390698e-06, "loss": 0.2542, "step": 4909 }, { "epoch": 2.1707146279438323, "grad_norm": 0.33800572762658804, "learning_rate": 2.1552814226197203e-06, "loss": 0.274, "step": 4910 }, { "epoch": 2.1711568938193344, "grad_norm": 0.3880194481648189, "learning_rate": 2.1531665446066007e-06, "loss": 0.2688, "step": 4911 }, { "epoch": 2.1715991596948365, "grad_norm": 0.32461618214696386, "learning_rate": 2.1510524199593787e-06, "loss": 0.2653, "step": 4912 }, { "epoch": 2.1720414255703386, "grad_norm": 0.3020802287333837, "learning_rate": 2.1489390492375177e-06, "loss": 0.2676, "step": 4913 }, { "epoch": 2.1724836914458407, "grad_norm": 0.3124795634114043, "learning_rate": 2.14682643300029e-06, "loss": 0.2945, "step": 4914 }, { "epoch": 2.172925957321343, "grad_norm": 0.35269030849759864, "learning_rate": 2.144714571806761e-06, "loss": 0.2513, "step": 4915 }, { "epoch": 2.173368223196845, "grad_norm": 0.3366891081358319, "learning_rate": 2.1426034662158e-06, "loss": 0.2738, "step": 4916 }, { "epoch": 2.1738104890723475, "grad_norm": 0.3342649587398713, "learning_rate": 2.140493116786074e-06, "loss": 0.2676, "step": 4917 }, { "epoch": 2.1742527549478496, "grad_norm": 0.3248867253059273, "learning_rate": 2.1383835240760514e-06, "loss": 0.2557, "step": 4918 }, { "epoch": 2.1746950208233518, "grad_norm": 0.3300782243110706, "learning_rate": 2.136274688643999e-06, "loss": 0.2783, "step": 4919 }, { "epoch": 2.175137286698854, "grad_norm": 0.2993619607585854, "learning_rate": 2.134166611047985e-06, "loss": 0.2737, "step": 4920 }, { "epoch": 2.175579552574356, "grad_norm": 0.38028361307824726, "learning_rate": 2.1320592918458764e-06, "loss": 0.2561, "step": 4921 }, { "epoch": 2.176021818449858, "grad_norm": 0.34841491998544766, "learning_rate": 2.1299527315953385e-06, "loss": 0.2421, "step": 4922 }, { "epoch": 2.17646408432536, "grad_norm": 0.2992386659680743, "learning_rate": 2.1278469308538367e-06, "loss": 0.2657, "step": 4923 }, { "epoch": 2.1769063502008623, "grad_norm": 0.4519814722253754, "learning_rate": 2.125741890178636e-06, "loss": 0.2804, "step": 4924 }, { "epoch": 2.1773486160763644, "grad_norm": 0.3076733623476652, "learning_rate": 2.1236376101267984e-06, "loss": 0.2707, "step": 4925 }, { "epoch": 2.1777908819518665, "grad_norm": 0.32615938636460157, "learning_rate": 2.121534091255186e-06, "loss": 0.2373, "step": 4926 }, { "epoch": 2.1782331478273687, "grad_norm": 0.3021247491741697, "learning_rate": 2.11943133412046e-06, "loss": 0.2323, "step": 4927 }, { "epoch": 2.178675413702871, "grad_norm": 0.3431902331878005, "learning_rate": 2.117329339279078e-06, "loss": 0.2845, "step": 4928 }, { "epoch": 2.1791176795783733, "grad_norm": 0.31967314492006244, "learning_rate": 2.115228107287296e-06, "loss": 0.2824, "step": 4929 }, { "epoch": 2.1795599454538754, "grad_norm": 0.30856592076221445, "learning_rate": 2.113127638701173e-06, "loss": 0.2733, "step": 4930 }, { "epoch": 2.1800022113293775, "grad_norm": 0.3022539452592239, "learning_rate": 2.111027934076559e-06, "loss": 0.273, "step": 4931 }, { "epoch": 2.1804444772048797, "grad_norm": 0.32644563166180796, "learning_rate": 2.108928993969107e-06, "loss": 0.2748, "step": 4932 }, { "epoch": 2.1808867430803818, "grad_norm": 0.35200336600208004, "learning_rate": 2.106830818934267e-06, "loss": 0.2381, "step": 4933 }, { "epoch": 2.181329008955884, "grad_norm": 0.3689482135253797, "learning_rate": 2.104733409527278e-06, "loss": 0.2451, "step": 4934 }, { "epoch": 2.181771274831386, "grad_norm": 0.3360230858432469, "learning_rate": 2.102636766303191e-06, "loss": 0.276, "step": 4935 }, { "epoch": 2.182213540706888, "grad_norm": 0.413248847304461, "learning_rate": 2.100540889816844e-06, "loss": 0.2727, "step": 4936 }, { "epoch": 2.1826558065823907, "grad_norm": 0.3919298217863727, "learning_rate": 2.098445780622875e-06, "loss": 0.2589, "step": 4937 }, { "epoch": 2.183098072457893, "grad_norm": 0.34341980328744287, "learning_rate": 2.09635143927572e-06, "loss": 0.266, "step": 4938 }, { "epoch": 2.183540338333395, "grad_norm": 0.38020288907888955, "learning_rate": 2.09425786632961e-06, "loss": 0.2807, "step": 4939 }, { "epoch": 2.183982604208897, "grad_norm": 0.2814042196490318, "learning_rate": 2.0921650623385724e-06, "loss": 0.2777, "step": 4940 }, { "epoch": 2.184424870084399, "grad_norm": 0.3087071568155149, "learning_rate": 2.0900730278564335e-06, "loss": 0.2563, "step": 4941 }, { "epoch": 2.1848671359599012, "grad_norm": 0.3752614927199149, "learning_rate": 2.0879817634368142e-06, "loss": 0.2728, "step": 4942 }, { "epoch": 2.1853094018354033, "grad_norm": 0.38773379249157713, "learning_rate": 2.085891269633131e-06, "loss": 0.2642, "step": 4943 }, { "epoch": 2.1857516677109055, "grad_norm": 0.32317506463246254, "learning_rate": 2.083801546998599e-06, "loss": 0.2714, "step": 4944 }, { "epoch": 2.1861939335864076, "grad_norm": 0.33215197294886883, "learning_rate": 2.0817125960862244e-06, "loss": 0.2376, "step": 4945 }, { "epoch": 2.1866361994619097, "grad_norm": 0.3083051104477078, "learning_rate": 2.079624417448817e-06, "loss": 0.2711, "step": 4946 }, { "epoch": 2.187078465337412, "grad_norm": 0.28464840314497725, "learning_rate": 2.0775370116389747e-06, "loss": 0.2845, "step": 4947 }, { "epoch": 2.1875207312129143, "grad_norm": 0.26558291093633174, "learning_rate": 2.0754503792090947e-06, "loss": 0.2695, "step": 4948 }, { "epoch": 2.1879629970884165, "grad_norm": 0.38106518076000523, "learning_rate": 2.0733645207113704e-06, "loss": 0.2953, "step": 4949 }, { "epoch": 2.1884052629639186, "grad_norm": 0.3462013856707333, "learning_rate": 2.0712794366977835e-06, "loss": 0.2497, "step": 4950 }, { "epoch": 2.1888475288394207, "grad_norm": 0.32451072161708505, "learning_rate": 2.069195127720121e-06, "loss": 0.2816, "step": 4951 }, { "epoch": 2.189289794714923, "grad_norm": 0.3754166328726283, "learning_rate": 2.0671115943299574e-06, "loss": 0.2727, "step": 4952 }, { "epoch": 2.189732060590425, "grad_norm": 0.27124020013692807, "learning_rate": 2.0650288370786652e-06, "loss": 0.2666, "step": 4953 }, { "epoch": 2.190174326465927, "grad_norm": 0.3361205039249788, "learning_rate": 2.06294685651741e-06, "loss": 0.2601, "step": 4954 }, { "epoch": 2.190616592341429, "grad_norm": 0.31841708864726054, "learning_rate": 2.0608656531971523e-06, "loss": 0.2736, "step": 4955 }, { "epoch": 2.1910588582169312, "grad_norm": 0.3789813749834722, "learning_rate": 2.0587852276686478e-06, "loss": 0.2546, "step": 4956 }, { "epoch": 2.1915011240924334, "grad_norm": 0.3195893790915116, "learning_rate": 2.056705580482444e-06, "loss": 0.2867, "step": 4957 }, { "epoch": 2.191943389967936, "grad_norm": 0.3411195034955162, "learning_rate": 2.0546267121888863e-06, "loss": 0.3042, "step": 4958 }, { "epoch": 2.192385655843438, "grad_norm": 0.2760299432195682, "learning_rate": 2.0525486233381103e-06, "loss": 0.2567, "step": 4959 }, { "epoch": 2.19282792171894, "grad_norm": 0.34385641454301347, "learning_rate": 2.050471314480047e-06, "loss": 0.2379, "step": 4960 }, { "epoch": 2.1932701875944423, "grad_norm": 0.35902892018801735, "learning_rate": 2.048394786164419e-06, "loss": 0.2976, "step": 4961 }, { "epoch": 2.1937124534699444, "grad_norm": 0.3961667855280278, "learning_rate": 2.0463190389407474e-06, "loss": 0.2407, "step": 4962 }, { "epoch": 2.1941547193454465, "grad_norm": 0.27567630148175676, "learning_rate": 2.0442440733583415e-06, "loss": 0.2739, "step": 4963 }, { "epoch": 2.1945969852209486, "grad_norm": 0.34889017579942677, "learning_rate": 2.042169889966308e-06, "loss": 0.2957, "step": 4964 }, { "epoch": 2.1950392510964507, "grad_norm": 0.3741695387972245, "learning_rate": 2.04009648931354e-06, "loss": 0.2759, "step": 4965 }, { "epoch": 2.195481516971953, "grad_norm": 0.3796504524357924, "learning_rate": 2.038023871948728e-06, "loss": 0.2682, "step": 4966 }, { "epoch": 2.195923782847455, "grad_norm": 0.40639592875667535, "learning_rate": 2.0359520384203574e-06, "loss": 0.2999, "step": 4967 }, { "epoch": 2.1963660487229575, "grad_norm": 0.3198929123459795, "learning_rate": 2.033880989276703e-06, "loss": 0.2779, "step": 4968 }, { "epoch": 2.1968083145984596, "grad_norm": 0.3356379601764344, "learning_rate": 2.031810725065832e-06, "loss": 0.2836, "step": 4969 }, { "epoch": 2.1972505804739617, "grad_norm": 0.35314123629974425, "learning_rate": 2.0297412463356048e-06, "loss": 0.25, "step": 4970 }, { "epoch": 2.197692846349464, "grad_norm": 0.3945319641629262, "learning_rate": 2.0276725536336733e-06, "loss": 0.2824, "step": 4971 }, { "epoch": 2.198135112224966, "grad_norm": 0.2889872321334298, "learning_rate": 2.025604647507482e-06, "loss": 0.2533, "step": 4972 }, { "epoch": 2.198577378100468, "grad_norm": 0.6847363530296379, "learning_rate": 2.023537528504267e-06, "loss": 0.3061, "step": 4973 }, { "epoch": 2.19901964397597, "grad_norm": 0.47121028401889714, "learning_rate": 2.0214711971710554e-06, "loss": 0.2632, "step": 4974 }, { "epoch": 2.1994619098514723, "grad_norm": 0.377151283770984, "learning_rate": 2.0194056540546676e-06, "loss": 0.252, "step": 4975 }, { "epoch": 2.1999041757269744, "grad_norm": 0.42843924052383114, "learning_rate": 2.0173408997017134e-06, "loss": 0.246, "step": 4976 }, { "epoch": 2.2003464416024765, "grad_norm": 0.3392422936957994, "learning_rate": 2.015276934658594e-06, "loss": 0.2707, "step": 4977 }, { "epoch": 2.2007887074779786, "grad_norm": 0.45895704994653985, "learning_rate": 2.0132137594715052e-06, "loss": 0.2806, "step": 4978 }, { "epoch": 2.201230973353481, "grad_norm": 0.2826858397224389, "learning_rate": 2.011151374686429e-06, "loss": 0.2552, "step": 4979 }, { "epoch": 2.2016732392289833, "grad_norm": 0.3525533710828656, "learning_rate": 2.009089780849144e-06, "loss": 0.2365, "step": 4980 }, { "epoch": 2.2021155051044854, "grad_norm": 0.33939876047784767, "learning_rate": 2.0070289785052106e-06, "loss": 0.2635, "step": 4981 }, { "epoch": 2.2025577709799875, "grad_norm": 0.3270246316583852, "learning_rate": 2.0049689681999853e-06, "loss": 0.2511, "step": 4982 }, { "epoch": 2.2030000368554896, "grad_norm": 0.3266376858414491, "learning_rate": 2.0029097504786176e-06, "loss": 0.2468, "step": 4983 }, { "epoch": 2.2034423027309917, "grad_norm": 0.35589347024247925, "learning_rate": 2.000851325886044e-06, "loss": 0.239, "step": 4984 }, { "epoch": 2.203884568606494, "grad_norm": 0.33011316369813115, "learning_rate": 1.998793694966991e-06, "loss": 0.2727, "step": 4985 }, { "epoch": 2.204326834481996, "grad_norm": 0.32180738037209394, "learning_rate": 1.996736858265975e-06, "loss": 0.2487, "step": 4986 }, { "epoch": 2.204769100357498, "grad_norm": 0.34681050211391784, "learning_rate": 1.994680816327303e-06, "loss": 0.254, "step": 4987 }, { "epoch": 2.2052113662330006, "grad_norm": 0.33903348857683274, "learning_rate": 1.9926255696950715e-06, "loss": 0.2505, "step": 4988 }, { "epoch": 2.2056536321085027, "grad_norm": 0.3338666123350076, "learning_rate": 1.990571118913166e-06, "loss": 0.2907, "step": 4989 }, { "epoch": 2.206095897984005, "grad_norm": 0.37172568635954284, "learning_rate": 1.9885174645252624e-06, "loss": 0.2541, "step": 4990 }, { "epoch": 2.206538163859507, "grad_norm": 0.40547860292234306, "learning_rate": 1.9864646070748256e-06, "loss": 0.2384, "step": 4991 }, { "epoch": 2.206980429735009, "grad_norm": 0.4164273920323257, "learning_rate": 1.984412547105107e-06, "loss": 0.28, "step": 4992 }, { "epoch": 2.207422695610511, "grad_norm": 0.2720662666684577, "learning_rate": 1.982361285159152e-06, "loss": 0.2609, "step": 4993 }, { "epoch": 2.2078649614860133, "grad_norm": 0.4189736706457445, "learning_rate": 1.9803108217797918e-06, "loss": 0.2734, "step": 4994 }, { "epoch": 2.2083072273615154, "grad_norm": 0.3866661655877981, "learning_rate": 1.9782611575096472e-06, "loss": 0.2715, "step": 4995 }, { "epoch": 2.2087494932370175, "grad_norm": 0.37532468249839035, "learning_rate": 1.976212292891124e-06, "loss": 0.2822, "step": 4996 }, { "epoch": 2.2091917591125196, "grad_norm": 0.33840104156288014, "learning_rate": 1.9741642284664203e-06, "loss": 0.2743, "step": 4997 }, { "epoch": 2.2096340249880217, "grad_norm": 0.3438370372336445, "learning_rate": 1.9721169647775206e-06, "loss": 0.2578, "step": 4998 }, { "epoch": 2.2100762908635243, "grad_norm": 0.382489914028594, "learning_rate": 1.9700705023662015e-06, "loss": 0.2876, "step": 4999 }, { "epoch": 2.2105185567390264, "grad_norm": 0.3268466240973372, "learning_rate": 1.9680248417740224e-06, "loss": 0.3226, "step": 5000 }, { "epoch": 2.2109608226145285, "grad_norm": 0.3697790353325218, "learning_rate": 1.9659799835423326e-06, "loss": 0.2799, "step": 5001 }, { "epoch": 2.2114030884900306, "grad_norm": 0.3374475663994866, "learning_rate": 1.9639359282122684e-06, "loss": 0.2398, "step": 5002 }, { "epoch": 2.2118453543655328, "grad_norm": 0.3266152700484507, "learning_rate": 1.9618926763247553e-06, "loss": 0.2627, "step": 5003 }, { "epoch": 2.212287620241035, "grad_norm": 0.31647832039017, "learning_rate": 1.959850228420504e-06, "loss": 0.2715, "step": 5004 }, { "epoch": 2.212729886116537, "grad_norm": 0.3963856095855509, "learning_rate": 1.9578085850400137e-06, "loss": 0.2712, "step": 5005 }, { "epoch": 2.213172151992039, "grad_norm": 0.43450227864967206, "learning_rate": 1.9557677467235708e-06, "loss": 0.2742, "step": 5006 }, { "epoch": 2.213614417867541, "grad_norm": 0.2696799812524099, "learning_rate": 1.953727714011248e-06, "loss": 0.2423, "step": 5007 }, { "epoch": 2.2140566837430433, "grad_norm": 0.4149804372120516, "learning_rate": 1.9516884874429044e-06, "loss": 0.2414, "step": 5008 }, { "epoch": 2.214498949618546, "grad_norm": 0.3225272754230801, "learning_rate": 1.949650067558188e-06, "loss": 0.2387, "step": 5009 }, { "epoch": 2.214941215494048, "grad_norm": 0.3938852343501493, "learning_rate": 1.947612454896532e-06, "loss": 0.3037, "step": 5010 }, { "epoch": 2.21538348136955, "grad_norm": 0.36141888148879275, "learning_rate": 1.945575649997157e-06, "loss": 0.256, "step": 5011 }, { "epoch": 2.215825747245052, "grad_norm": 0.31629274913788835, "learning_rate": 1.9435396533990646e-06, "loss": 0.2543, "step": 5012 }, { "epoch": 2.2162680131205543, "grad_norm": 0.3395768532096082, "learning_rate": 1.9415044656410493e-06, "loss": 0.2759, "step": 5013 }, { "epoch": 2.2167102789960564, "grad_norm": 0.3386997293377115, "learning_rate": 1.9394700872616856e-06, "loss": 0.2993, "step": 5014 }, { "epoch": 2.2171525448715586, "grad_norm": 0.2958038867573915, "learning_rate": 1.937436518799342e-06, "loss": 0.2783, "step": 5015 }, { "epoch": 2.2175948107470607, "grad_norm": 0.3153538185922784, "learning_rate": 1.9354037607921653e-06, "loss": 0.2684, "step": 5016 }, { "epoch": 2.2180370766225628, "grad_norm": 0.2881201571550892, "learning_rate": 1.9333718137780905e-06, "loss": 0.2919, "step": 5017 }, { "epoch": 2.218479342498065, "grad_norm": 0.3267570976378012, "learning_rate": 1.931340678294837e-06, "loss": 0.2563, "step": 5018 }, { "epoch": 2.2189216083735674, "grad_norm": 0.3887849872881839, "learning_rate": 1.929310354879911e-06, "loss": 0.2958, "step": 5019 }, { "epoch": 2.2193638742490696, "grad_norm": 0.3770826178945667, "learning_rate": 1.9272808440706024e-06, "loss": 0.3005, "step": 5020 }, { "epoch": 2.2198061401245717, "grad_norm": 0.392754350852765, "learning_rate": 1.925252146403987e-06, "loss": 0.2623, "step": 5021 }, { "epoch": 2.220248406000074, "grad_norm": 0.2939450218632894, "learning_rate": 1.9232242624169246e-06, "loss": 0.3228, "step": 5022 }, { "epoch": 2.220690671875576, "grad_norm": 0.47133664234896094, "learning_rate": 1.92119719264606e-06, "loss": 0.2582, "step": 5023 }, { "epoch": 2.221132937751078, "grad_norm": 0.29262096139851795, "learning_rate": 1.9191709376278218e-06, "loss": 0.2758, "step": 5024 }, { "epoch": 2.22157520362658, "grad_norm": 0.3575932605446019, "learning_rate": 1.9171454978984263e-06, "loss": 0.291, "step": 5025 }, { "epoch": 2.2220174695020822, "grad_norm": 0.3094550409002769, "learning_rate": 1.915120873993872e-06, "loss": 0.2839, "step": 5026 }, { "epoch": 2.2224597353775843, "grad_norm": 0.35020056858192344, "learning_rate": 1.9130970664499377e-06, "loss": 0.2654, "step": 5027 }, { "epoch": 2.2229020012530865, "grad_norm": 0.37589179293188557, "learning_rate": 1.9110740758021905e-06, "loss": 0.2661, "step": 5028 }, { "epoch": 2.2233442671285886, "grad_norm": 0.30523880292089317, "learning_rate": 1.90905190258598e-06, "loss": 0.2733, "step": 5029 }, { "epoch": 2.223786533004091, "grad_norm": 0.3996586284061671, "learning_rate": 1.907030547336442e-06, "loss": 0.2859, "step": 5030 }, { "epoch": 2.2242287988795932, "grad_norm": 0.2998021136106064, "learning_rate": 1.9050100105884927e-06, "loss": 0.3006, "step": 5031 }, { "epoch": 2.2246710647550954, "grad_norm": 0.4596236808732868, "learning_rate": 1.9029902928768324e-06, "loss": 0.2939, "step": 5032 }, { "epoch": 2.2251133306305975, "grad_norm": 0.4049773709847376, "learning_rate": 1.9009713947359453e-06, "loss": 0.2785, "step": 5033 }, { "epoch": 2.2255555965060996, "grad_norm": 0.3667509617965911, "learning_rate": 1.8989533167000973e-06, "loss": 0.2172, "step": 5034 }, { "epoch": 2.2259978623816017, "grad_norm": 0.333378555967773, "learning_rate": 1.8969360593033397e-06, "loss": 0.2568, "step": 5035 }, { "epoch": 2.226440128257104, "grad_norm": 0.3526289917865817, "learning_rate": 1.8949196230795042e-06, "loss": 0.2803, "step": 5036 }, { "epoch": 2.226882394132606, "grad_norm": 0.3537446959958034, "learning_rate": 1.892904008562207e-06, "loss": 0.2572, "step": 5037 }, { "epoch": 2.227324660008108, "grad_norm": 0.32595251849790713, "learning_rate": 1.8908892162848451e-06, "loss": 0.2476, "step": 5038 }, { "epoch": 2.2277669258836106, "grad_norm": 0.3034918268981851, "learning_rate": 1.8888752467806004e-06, "loss": 0.252, "step": 5039 }, { "epoch": 2.2282091917591127, "grad_norm": 0.507935665593914, "learning_rate": 1.8868621005824322e-06, "loss": 0.2496, "step": 5040 }, { "epoch": 2.228651457634615, "grad_norm": 0.32912820343832017, "learning_rate": 1.8848497782230895e-06, "loss": 0.2737, "step": 5041 }, { "epoch": 2.229093723510117, "grad_norm": 0.31920558468628335, "learning_rate": 1.8828382802350992e-06, "loss": 0.2634, "step": 5042 }, { "epoch": 2.229535989385619, "grad_norm": 0.32281861029499836, "learning_rate": 1.8808276071507663e-06, "loss": 0.2867, "step": 5043 }, { "epoch": 2.229978255261121, "grad_norm": 0.35108589410213925, "learning_rate": 1.8788177595021834e-06, "loss": 0.2432, "step": 5044 }, { "epoch": 2.2304205211366233, "grad_norm": 0.27587148709805853, "learning_rate": 1.8768087378212196e-06, "loss": 0.2618, "step": 5045 }, { "epoch": 2.2308627870121254, "grad_norm": 0.28064380811614587, "learning_rate": 1.8748005426395322e-06, "loss": 0.2868, "step": 5046 }, { "epoch": 2.2313050528876275, "grad_norm": 0.45346098667731805, "learning_rate": 1.872793174488554e-06, "loss": 0.2579, "step": 5047 }, { "epoch": 2.2317473187631296, "grad_norm": 0.3342343129107772, "learning_rate": 1.8707866338995012e-06, "loss": 0.2758, "step": 5048 }, { "epoch": 2.2321895846386317, "grad_norm": 0.37492811455712216, "learning_rate": 1.8687809214033697e-06, "loss": 0.2493, "step": 5049 }, { "epoch": 2.2326318505141343, "grad_norm": 0.3274475432898788, "learning_rate": 1.866776037530938e-06, "loss": 0.2724, "step": 5050 }, { "epoch": 2.2330741163896364, "grad_norm": 0.3129580769227382, "learning_rate": 1.864771982812763e-06, "loss": 0.2656, "step": 5051 }, { "epoch": 2.2335163822651385, "grad_norm": 0.3093188042857144, "learning_rate": 1.8627687577791852e-06, "loss": 0.2779, "step": 5052 }, { "epoch": 2.2339586481406406, "grad_norm": 0.400354456900956, "learning_rate": 1.860766362960323e-06, "loss": 0.2657, "step": 5053 }, { "epoch": 2.2344009140161427, "grad_norm": 0.3793546334263574, "learning_rate": 1.8587647988860764e-06, "loss": 0.2704, "step": 5054 }, { "epoch": 2.234843179891645, "grad_norm": 0.2865570689092265, "learning_rate": 1.8567640660861247e-06, "loss": 0.2874, "step": 5055 }, { "epoch": 2.235285445767147, "grad_norm": 0.44702697188849994, "learning_rate": 1.8547641650899267e-06, "loss": 0.2883, "step": 5056 }, { "epoch": 2.235727711642649, "grad_norm": 0.3424552719591585, "learning_rate": 1.8527650964267264e-06, "loss": 0.2758, "step": 5057 }, { "epoch": 2.236169977518151, "grad_norm": 0.34533298099568976, "learning_rate": 1.8507668606255385e-06, "loss": 0.2595, "step": 5058 }, { "epoch": 2.2366122433936537, "grad_norm": 0.2897999175554871, "learning_rate": 1.8487694582151643e-06, "loss": 0.265, "step": 5059 }, { "epoch": 2.237054509269156, "grad_norm": 0.4016772545589521, "learning_rate": 1.846772889724181e-06, "loss": 0.2647, "step": 5060 }, { "epoch": 2.237496775144658, "grad_norm": 0.2624795823562207, "learning_rate": 1.8447771556809458e-06, "loss": 0.2728, "step": 5061 }, { "epoch": 2.23793904102016, "grad_norm": 0.33889092278867744, "learning_rate": 1.8427822566135983e-06, "loss": 0.2668, "step": 5062 }, { "epoch": 2.238381306895662, "grad_norm": 0.3622175314584564, "learning_rate": 1.8407881930500532e-06, "loss": 0.2729, "step": 5063 }, { "epoch": 2.2388235727711643, "grad_norm": 0.30370315270054254, "learning_rate": 1.8387949655180055e-06, "loss": 0.2783, "step": 5064 }, { "epoch": 2.2392658386466664, "grad_norm": 0.35322194475549884, "learning_rate": 1.836802574544928e-06, "loss": 0.2907, "step": 5065 }, { "epoch": 2.2397081045221685, "grad_norm": 0.3036189198184608, "learning_rate": 1.8348110206580745e-06, "loss": 0.2503, "step": 5066 }, { "epoch": 2.2401503703976706, "grad_norm": 0.4271294162012232, "learning_rate": 1.832820304384474e-06, "loss": 0.2492, "step": 5067 }, { "epoch": 2.2405926362731727, "grad_norm": 0.3134471305059257, "learning_rate": 1.8308304262509368e-06, "loss": 0.2546, "step": 5068 }, { "epoch": 2.241034902148675, "grad_norm": 0.3220807068712218, "learning_rate": 1.8288413867840505e-06, "loss": 0.2295, "step": 5069 }, { "epoch": 2.2414771680241774, "grad_norm": 0.33266181670195705, "learning_rate": 1.826853186510179e-06, "loss": 0.2944, "step": 5070 }, { "epoch": 2.2419194338996795, "grad_norm": 0.27964719486924017, "learning_rate": 1.8248658259554668e-06, "loss": 0.2742, "step": 5071 }, { "epoch": 2.2423616997751816, "grad_norm": 0.29352007765751253, "learning_rate": 1.8228793056458327e-06, "loss": 0.2488, "step": 5072 }, { "epoch": 2.2428039656506837, "grad_norm": 0.3043846396391323, "learning_rate": 1.8208936261069809e-06, "loss": 0.2773, "step": 5073 }, { "epoch": 2.243246231526186, "grad_norm": 0.28710230113428087, "learning_rate": 1.818908787864382e-06, "loss": 0.2528, "step": 5074 }, { "epoch": 2.243688497401688, "grad_norm": 0.37174244356111996, "learning_rate": 1.8169247914432915e-06, "loss": 0.2579, "step": 5075 }, { "epoch": 2.24413076327719, "grad_norm": 0.30563296376835797, "learning_rate": 1.8149416373687407e-06, "loss": 0.2887, "step": 5076 }, { "epoch": 2.244573029152692, "grad_norm": 0.3509887658466663, "learning_rate": 1.812959326165535e-06, "loss": 0.2697, "step": 5077 }, { "epoch": 2.2450152950281943, "grad_norm": 0.325192140178732, "learning_rate": 1.8109778583582626e-06, "loss": 0.244, "step": 5078 }, { "epoch": 2.2454575609036964, "grad_norm": 0.30631851323612147, "learning_rate": 1.8089972344712841e-06, "loss": 0.2609, "step": 5079 }, { "epoch": 2.2458998267791985, "grad_norm": 0.293703592785355, "learning_rate": 1.8070174550287378e-06, "loss": 0.2593, "step": 5080 }, { "epoch": 2.246342092654701, "grad_norm": 0.3160121657426917, "learning_rate": 1.8050385205545383e-06, "loss": 0.2645, "step": 5081 }, { "epoch": 2.246784358530203, "grad_norm": 0.29887878518351074, "learning_rate": 1.8030604315723765e-06, "loss": 0.2632, "step": 5082 }, { "epoch": 2.2472266244057053, "grad_norm": 0.3104659417546676, "learning_rate": 1.80108318860572e-06, "loss": 0.2644, "step": 5083 }, { "epoch": 2.2476688902812074, "grad_norm": 0.3044390044217166, "learning_rate": 1.7991067921778133e-06, "loss": 0.2744, "step": 5084 }, { "epoch": 2.2481111561567095, "grad_norm": 0.32759216592699547, "learning_rate": 1.797131242811675e-06, "loss": 0.2924, "step": 5085 }, { "epoch": 2.2485534220322116, "grad_norm": 0.2796104170101047, "learning_rate": 1.7951565410301004e-06, "loss": 0.283, "step": 5086 }, { "epoch": 2.2489956879077138, "grad_norm": 0.3278391291418322, "learning_rate": 1.7931826873556613e-06, "loss": 0.2709, "step": 5087 }, { "epoch": 2.249437953783216, "grad_norm": 0.3115288198981472, "learning_rate": 1.7912096823107045e-06, "loss": 0.2896, "step": 5088 }, { "epoch": 2.249880219658718, "grad_norm": 0.32429430329197195, "learning_rate": 1.7892375264173517e-06, "loss": 0.2682, "step": 5089 }, { "epoch": 2.2503224855342205, "grad_norm": 0.3679295980762954, "learning_rate": 1.7872662201975e-06, "loss": 0.264, "step": 5090 }, { "epoch": 2.2507647514097227, "grad_norm": 0.3346030434284065, "learning_rate": 1.7852957641728224e-06, "loss": 0.2586, "step": 5091 }, { "epoch": 2.2512070172852248, "grad_norm": 0.4134766021427156, "learning_rate": 1.7833261588647666e-06, "loss": 0.2507, "step": 5092 }, { "epoch": 2.251649283160727, "grad_norm": 0.35558543915276264, "learning_rate": 1.7813574047945532e-06, "loss": 0.2479, "step": 5093 }, { "epoch": 2.252091549036229, "grad_norm": 0.31390514682147386, "learning_rate": 1.7793895024831826e-06, "loss": 0.2416, "step": 5094 }, { "epoch": 2.252533814911731, "grad_norm": 0.4674769722968112, "learning_rate": 1.777422452451425e-06, "loss": 0.2391, "step": 5095 }, { "epoch": 2.252976080787233, "grad_norm": 0.48775544156177075, "learning_rate": 1.7754562552198269e-06, "loss": 0.2713, "step": 5096 }, { "epoch": 2.2534183466627353, "grad_norm": 0.3159496976944407, "learning_rate": 1.77349091130871e-06, "loss": 0.268, "step": 5097 }, { "epoch": 2.2538606125382374, "grad_norm": 0.3680414867383307, "learning_rate": 1.7715264212381638e-06, "loss": 0.273, "step": 5098 }, { "epoch": 2.2543028784137396, "grad_norm": 0.4473596536297606, "learning_rate": 1.7695627855280627e-06, "loss": 0.3012, "step": 5099 }, { "epoch": 2.2547451442892417, "grad_norm": 0.41164913417826765, "learning_rate": 1.767600004698048e-06, "loss": 0.3023, "step": 5100 }, { "epoch": 2.255187410164744, "grad_norm": 0.4185740716406089, "learning_rate": 1.7656380792675354e-06, "loss": 0.2486, "step": 5101 }, { "epoch": 2.2556296760402463, "grad_norm": 0.34522605102815623, "learning_rate": 1.7636770097557155e-06, "loss": 0.2608, "step": 5102 }, { "epoch": 2.2560719419157484, "grad_norm": 0.2932823396088964, "learning_rate": 1.7617167966815495e-06, "loss": 0.2482, "step": 5103 }, { "epoch": 2.2565142077912506, "grad_norm": 0.2911115816315477, "learning_rate": 1.7597574405637801e-06, "loss": 0.2642, "step": 5104 }, { "epoch": 2.2569564736667527, "grad_norm": 0.3017616528893616, "learning_rate": 1.7577989419209124e-06, "loss": 0.2544, "step": 5105 }, { "epoch": 2.257398739542255, "grad_norm": 0.3939109718894928, "learning_rate": 1.7558413012712312e-06, "loss": 0.2881, "step": 5106 }, { "epoch": 2.257841005417757, "grad_norm": 0.2650173514756612, "learning_rate": 1.7538845191327924e-06, "loss": 0.2793, "step": 5107 }, { "epoch": 2.258283271293259, "grad_norm": 0.345837091858566, "learning_rate": 1.751928596023425e-06, "loss": 0.2743, "step": 5108 }, { "epoch": 2.258725537168761, "grad_norm": 0.3981234251416509, "learning_rate": 1.7499735324607293e-06, "loss": 0.2831, "step": 5109 }, { "epoch": 2.2591678030442637, "grad_norm": 0.8786466106594897, "learning_rate": 1.7480193289620827e-06, "loss": 0.2769, "step": 5110 }, { "epoch": 2.259610068919766, "grad_norm": 0.3722611468734663, "learning_rate": 1.7460659860446306e-06, "loss": 0.2298, "step": 5111 }, { "epoch": 2.260052334795268, "grad_norm": 0.3242883749420354, "learning_rate": 1.7441135042252915e-06, "loss": 0.2479, "step": 5112 }, { "epoch": 2.26049460067077, "grad_norm": 0.3495676719110167, "learning_rate": 1.7421618840207576e-06, "loss": 0.2721, "step": 5113 }, { "epoch": 2.260936866546272, "grad_norm": 0.35447527182265587, "learning_rate": 1.740211125947488e-06, "loss": 0.2455, "step": 5114 }, { "epoch": 2.2613791324217742, "grad_norm": 0.31534797827079314, "learning_rate": 1.7382612305217212e-06, "loss": 0.2412, "step": 5115 }, { "epoch": 2.2618213982972764, "grad_norm": 0.30914173783465004, "learning_rate": 1.7363121982594627e-06, "loss": 0.2747, "step": 5116 }, { "epoch": 2.2622636641727785, "grad_norm": 0.394933681274775, "learning_rate": 1.7343640296764913e-06, "loss": 0.2876, "step": 5117 }, { "epoch": 2.2627059300482806, "grad_norm": 0.3336424273640027, "learning_rate": 1.7324167252883556e-06, "loss": 0.248, "step": 5118 }, { "epoch": 2.2631481959237827, "grad_norm": 0.3435536513523599, "learning_rate": 1.7304702856103767e-06, "loss": 0.2557, "step": 5119 }, { "epoch": 2.263590461799285, "grad_norm": 0.3831373189024649, "learning_rate": 1.728524711157647e-06, "loss": 0.2658, "step": 5120 }, { "epoch": 2.2640327276747874, "grad_norm": 0.3404741989972247, "learning_rate": 1.7265800024450292e-06, "loss": 0.2969, "step": 5121 }, { "epoch": 2.2644749935502895, "grad_norm": 0.40896379043633296, "learning_rate": 1.7246361599871576e-06, "loss": 0.282, "step": 5122 }, { "epoch": 2.2649172594257916, "grad_norm": 0.33114526926461946, "learning_rate": 1.7226931842984367e-06, "loss": 0.3033, "step": 5123 }, { "epoch": 2.2653595253012937, "grad_norm": 0.3386514414421415, "learning_rate": 1.7207510758930423e-06, "loss": 0.2763, "step": 5124 }, { "epoch": 2.265801791176796, "grad_norm": 0.2831304655726724, "learning_rate": 1.7188098352849175e-06, "loss": 0.2595, "step": 5125 }, { "epoch": 2.266244057052298, "grad_norm": 0.3582196182305486, "learning_rate": 1.716869462987783e-06, "loss": 0.2846, "step": 5126 }, { "epoch": 2.2666863229278, "grad_norm": 0.31931592856458374, "learning_rate": 1.7149299595151231e-06, "loss": 0.2526, "step": 5127 }, { "epoch": 2.267128588803302, "grad_norm": 0.33349414053337556, "learning_rate": 1.7129913253801962e-06, "loss": 0.2488, "step": 5128 }, { "epoch": 2.2675708546788043, "grad_norm": 0.3218861668942548, "learning_rate": 1.711053561096025e-06, "loss": 0.2492, "step": 5129 }, { "epoch": 2.2680131205543064, "grad_norm": 0.3182352507902117, "learning_rate": 1.7091166671754056e-06, "loss": 0.2987, "step": 5130 }, { "epoch": 2.2684553864298085, "grad_norm": 0.3161446378483786, "learning_rate": 1.707180644130908e-06, "loss": 0.2691, "step": 5131 }, { "epoch": 2.268897652305311, "grad_norm": 0.35193723954651146, "learning_rate": 1.7052454924748658e-06, "loss": 0.2436, "step": 5132 }, { "epoch": 2.269339918180813, "grad_norm": 0.2856133738617285, "learning_rate": 1.7033112127193828e-06, "loss": 0.246, "step": 5133 }, { "epoch": 2.2697821840563153, "grad_norm": 0.31689535880566144, "learning_rate": 1.7013778053763346e-06, "loss": 0.2651, "step": 5134 }, { "epoch": 2.2702244499318174, "grad_norm": 0.34613003968691386, "learning_rate": 1.699445270957364e-06, "loss": 0.263, "step": 5135 }, { "epoch": 2.2706667158073195, "grad_norm": 0.42034697836320134, "learning_rate": 1.6975136099738827e-06, "loss": 0.2599, "step": 5136 }, { "epoch": 2.2711089816828216, "grad_norm": 0.328996855617159, "learning_rate": 1.6955828229370724e-06, "loss": 0.2426, "step": 5137 }, { "epoch": 2.2715512475583237, "grad_norm": 0.6515661973759084, "learning_rate": 1.6936529103578831e-06, "loss": 0.3444, "step": 5138 }, { "epoch": 2.271993513433826, "grad_norm": 0.4416870638344207, "learning_rate": 1.6917238727470325e-06, "loss": 0.2502, "step": 5139 }, { "epoch": 2.272435779309328, "grad_norm": 0.32659123784589167, "learning_rate": 1.6897957106150065e-06, "loss": 0.2786, "step": 5140 }, { "epoch": 2.2728780451848305, "grad_norm": 0.42192496256367684, "learning_rate": 1.6878684244720633e-06, "loss": 0.2873, "step": 5141 }, { "epoch": 2.2733203110603326, "grad_norm": 0.3330163216345388, "learning_rate": 1.6859420148282245e-06, "loss": 0.2857, "step": 5142 }, { "epoch": 2.2737625769358347, "grad_norm": 0.32331594841707223, "learning_rate": 1.6840164821932819e-06, "loss": 0.2592, "step": 5143 }, { "epoch": 2.274204842811337, "grad_norm": 0.29615449372855057, "learning_rate": 1.682091827076796e-06, "loss": 0.277, "step": 5144 }, { "epoch": 2.274647108686839, "grad_norm": 0.5494164734373872, "learning_rate": 1.6801680499880913e-06, "loss": 0.2746, "step": 5145 }, { "epoch": 2.275089374562341, "grad_norm": 0.3246639038016573, "learning_rate": 1.6782451514362625e-06, "loss": 0.252, "step": 5146 }, { "epoch": 2.275531640437843, "grad_norm": 0.3602502544122202, "learning_rate": 1.6763231319301742e-06, "loss": 0.2825, "step": 5147 }, { "epoch": 2.2759739063133453, "grad_norm": 0.3361748867226237, "learning_rate": 1.674401991978456e-06, "loss": 0.2865, "step": 5148 }, { "epoch": 2.2764161721888474, "grad_norm": 0.32037778152642477, "learning_rate": 1.6724817320895037e-06, "loss": 0.2688, "step": 5149 }, { "epoch": 2.2768584380643495, "grad_norm": 0.29653005164794516, "learning_rate": 1.6705623527714816e-06, "loss": 0.2527, "step": 5150 }, { "epoch": 2.2773007039398516, "grad_norm": 0.3065049047520487, "learning_rate": 1.6686438545323202e-06, "loss": 0.2614, "step": 5151 }, { "epoch": 2.277742969815354, "grad_norm": 0.3258011884870884, "learning_rate": 1.6667262378797183e-06, "loss": 0.2615, "step": 5152 }, { "epoch": 2.2781852356908563, "grad_norm": 0.336322846411305, "learning_rate": 1.6648095033211398e-06, "loss": 0.2663, "step": 5153 }, { "epoch": 2.2786275015663584, "grad_norm": 0.30348258024200997, "learning_rate": 1.6628936513638156e-06, "loss": 0.2851, "step": 5154 }, { "epoch": 2.2790697674418605, "grad_norm": 0.36226036912898163, "learning_rate": 1.6609786825147439e-06, "loss": 0.2817, "step": 5155 }, { "epoch": 2.2795120333173626, "grad_norm": 0.3092489611211026, "learning_rate": 1.6590645972806863e-06, "loss": 0.2818, "step": 5156 }, { "epoch": 2.2799542991928647, "grad_norm": 0.42568839973199546, "learning_rate": 1.6571513961681761e-06, "loss": 0.2801, "step": 5157 }, { "epoch": 2.280396565068367, "grad_norm": 0.42190488801949266, "learning_rate": 1.6552390796835082e-06, "loss": 0.2554, "step": 5158 }, { "epoch": 2.280838830943869, "grad_norm": 0.335159343458966, "learning_rate": 1.653327648332746e-06, "loss": 0.2785, "step": 5159 }, { "epoch": 2.281281096819371, "grad_norm": 0.3591219568911471, "learning_rate": 1.6514171026217141e-06, "loss": 0.2724, "step": 5160 }, { "epoch": 2.2817233626948736, "grad_norm": 0.3264561151015458, "learning_rate": 1.6495074430560065e-06, "loss": 0.2634, "step": 5161 }, { "epoch": 2.2821656285703757, "grad_norm": 0.28678529620427934, "learning_rate": 1.6475986701409814e-06, "loss": 0.2607, "step": 5162 }, { "epoch": 2.282607894445878, "grad_norm": 0.3458548029424356, "learning_rate": 1.6456907843817654e-06, "loss": 0.3037, "step": 5163 }, { "epoch": 2.28305016032138, "grad_norm": 0.38479284287002113, "learning_rate": 1.6437837862832469e-06, "loss": 0.2768, "step": 5164 }, { "epoch": 2.283492426196882, "grad_norm": 0.3165450860738522, "learning_rate": 1.6418776763500804e-06, "loss": 0.266, "step": 5165 }, { "epoch": 2.283934692072384, "grad_norm": 0.4068081127927527, "learning_rate": 1.6399724550866852e-06, "loss": 0.2557, "step": 5166 }, { "epoch": 2.2843769579478863, "grad_norm": 0.449259350193444, "learning_rate": 1.6380681229972456e-06, "loss": 0.2801, "step": 5167 }, { "epoch": 2.2848192238233884, "grad_norm": 0.32449399906587173, "learning_rate": 1.6361646805857107e-06, "loss": 0.2746, "step": 5168 }, { "epoch": 2.2852614896988905, "grad_norm": 0.32776035224713457, "learning_rate": 1.6342621283557937e-06, "loss": 0.2786, "step": 5169 }, { "epoch": 2.2857037555743926, "grad_norm": 0.29465802018600473, "learning_rate": 1.6323604668109726e-06, "loss": 0.2849, "step": 5170 }, { "epoch": 2.2861460214498948, "grad_norm": 0.4163002171439696, "learning_rate": 1.6304596964544906e-06, "loss": 0.272, "step": 5171 }, { "epoch": 2.2865882873253973, "grad_norm": 0.40852130273011095, "learning_rate": 1.6285598177893513e-06, "loss": 0.2802, "step": 5172 }, { "epoch": 2.2870305532008994, "grad_norm": 0.37758925857811404, "learning_rate": 1.6266608313183297e-06, "loss": 0.2479, "step": 5173 }, { "epoch": 2.2874728190764015, "grad_norm": 0.32842216161469723, "learning_rate": 1.6247627375439566e-06, "loss": 0.2673, "step": 5174 }, { "epoch": 2.2879150849519037, "grad_norm": 0.3239733829894798, "learning_rate": 1.6228655369685342e-06, "loss": 0.2713, "step": 5175 }, { "epoch": 2.2883573508274058, "grad_norm": 0.289888141079775, "learning_rate": 1.6209692300941193e-06, "loss": 0.252, "step": 5176 }, { "epoch": 2.288799616702908, "grad_norm": 0.34439941213562547, "learning_rate": 1.6190738174225374e-06, "loss": 0.249, "step": 5177 }, { "epoch": 2.28924188257841, "grad_norm": 0.36548801841149525, "learning_rate": 1.617179299455381e-06, "loss": 0.2357, "step": 5178 }, { "epoch": 2.289684148453912, "grad_norm": 0.28092060590644113, "learning_rate": 1.6152856766939994e-06, "loss": 0.2851, "step": 5179 }, { "epoch": 2.290126414329414, "grad_norm": 0.3287267058623433, "learning_rate": 1.6133929496395084e-06, "loss": 0.2643, "step": 5180 }, { "epoch": 2.2905686802049163, "grad_norm": 0.31621120599388863, "learning_rate": 1.6115011187927853e-06, "loss": 0.2358, "step": 5181 }, { "epoch": 2.2910109460804184, "grad_norm": 0.3078713054396145, "learning_rate": 1.6096101846544716e-06, "loss": 0.2411, "step": 5182 }, { "epoch": 2.291453211955921, "grad_norm": 0.383074381437392, "learning_rate": 1.6077201477249698e-06, "loss": 0.2572, "step": 5183 }, { "epoch": 2.291895477831423, "grad_norm": 0.38879222733650864, "learning_rate": 1.6058310085044464e-06, "loss": 0.2849, "step": 5184 }, { "epoch": 2.2923377437069252, "grad_norm": 0.35405214794620676, "learning_rate": 1.6039427674928305e-06, "loss": 0.2606, "step": 5185 }, { "epoch": 2.2927800095824273, "grad_norm": 0.32024937537026055, "learning_rate": 1.6020554251898119e-06, "loss": 0.2672, "step": 5186 }, { "epoch": 2.2932222754579294, "grad_norm": 0.3549389244476473, "learning_rate": 1.6001689820948434e-06, "loss": 0.243, "step": 5187 }, { "epoch": 2.2936645413334316, "grad_norm": 0.35734267621579385, "learning_rate": 1.5982834387071394e-06, "loss": 0.2384, "step": 5188 }, { "epoch": 2.2941068072089337, "grad_norm": 0.3147962911118802, "learning_rate": 1.5963987955256793e-06, "loss": 0.2676, "step": 5189 }, { "epoch": 2.294549073084436, "grad_norm": 0.4828276975711497, "learning_rate": 1.594515053049202e-06, "loss": 0.2969, "step": 5190 }, { "epoch": 2.294991338959938, "grad_norm": 0.33360846757795753, "learning_rate": 1.5926322117762043e-06, "loss": 0.2776, "step": 5191 }, { "epoch": 2.2954336048354405, "grad_norm": 0.34276488320570886, "learning_rate": 1.5907502722049501e-06, "loss": 0.256, "step": 5192 }, { "epoch": 2.2958758707109426, "grad_norm": 0.3663813778701289, "learning_rate": 1.5888692348334607e-06, "loss": 0.2757, "step": 5193 }, { "epoch": 2.2963181365864447, "grad_norm": 0.3407635434826641, "learning_rate": 1.586989100159524e-06, "loss": 0.264, "step": 5194 }, { "epoch": 2.296760402461947, "grad_norm": 0.3417331869186116, "learning_rate": 1.5851098686806832e-06, "loss": 0.2522, "step": 5195 }, { "epoch": 2.297202668337449, "grad_norm": 0.31976814978975054, "learning_rate": 1.583231540894245e-06, "loss": 0.2821, "step": 5196 }, { "epoch": 2.297644934212951, "grad_norm": 0.40568091897231207, "learning_rate": 1.581354117297278e-06, "loss": 0.2755, "step": 5197 }, { "epoch": 2.298087200088453, "grad_norm": 0.2925559819901946, "learning_rate": 1.579477598386609e-06, "loss": 0.2518, "step": 5198 }, { "epoch": 2.2985294659639552, "grad_norm": 0.38488302629944554, "learning_rate": 1.577601984658827e-06, "loss": 0.2446, "step": 5199 }, { "epoch": 2.2989717318394574, "grad_norm": 0.31705669702684125, "learning_rate": 1.5757272766102816e-06, "loss": 0.2871, "step": 5200 }, { "epoch": 2.2994139977149595, "grad_norm": 0.3270060731966094, "learning_rate": 1.573853474737082e-06, "loss": 0.2476, "step": 5201 }, { "epoch": 2.2998562635904616, "grad_norm": 0.39953183445281376, "learning_rate": 1.5719805795350984e-06, "loss": 0.3009, "step": 5202 }, { "epoch": 2.300298529465964, "grad_norm": 0.30587662611520156, "learning_rate": 1.5701085914999597e-06, "loss": 0.2503, "step": 5203 }, { "epoch": 2.3007407953414662, "grad_norm": 0.335518959137294, "learning_rate": 1.5682375111270542e-06, "loss": 0.2672, "step": 5204 }, { "epoch": 2.3011830612169684, "grad_norm": 0.3039518214687439, "learning_rate": 1.5663673389115348e-06, "loss": 0.2445, "step": 5205 }, { "epoch": 2.3016253270924705, "grad_norm": 0.3145542377201535, "learning_rate": 1.5644980753483109e-06, "loss": 0.2812, "step": 5206 }, { "epoch": 2.3020675929679726, "grad_norm": 0.38237440036485404, "learning_rate": 1.5626297209320478e-06, "loss": 0.2687, "step": 5207 }, { "epoch": 2.3025098588434747, "grad_norm": 0.33145758951878407, "learning_rate": 1.5607622761571751e-06, "loss": 0.2958, "step": 5208 }, { "epoch": 2.302952124718977, "grad_norm": 0.4622788627256233, "learning_rate": 1.5588957415178784e-06, "loss": 0.2778, "step": 5209 }, { "epoch": 2.303394390594479, "grad_norm": 0.3581013917847626, "learning_rate": 1.5570301175081077e-06, "loss": 0.2838, "step": 5210 }, { "epoch": 2.303836656469981, "grad_norm": 0.4606600764462968, "learning_rate": 1.555165404621567e-06, "loss": 0.2591, "step": 5211 }, { "epoch": 2.3042789223454836, "grad_norm": 0.33761356318035673, "learning_rate": 1.5533016033517211e-06, "loss": 0.2742, "step": 5212 }, { "epoch": 2.3047211882209857, "grad_norm": 0.32348084963475904, "learning_rate": 1.5514387141917925e-06, "loss": 0.2546, "step": 5213 }, { "epoch": 2.305163454096488, "grad_norm": 0.33590036847774557, "learning_rate": 1.549576737634763e-06, "loss": 0.2585, "step": 5214 }, { "epoch": 2.30560571997199, "grad_norm": 0.3342626827554177, "learning_rate": 1.547715674173374e-06, "loss": 0.2869, "step": 5215 }, { "epoch": 2.306047985847492, "grad_norm": 0.3145173050695412, "learning_rate": 1.5458555243001234e-06, "loss": 0.2823, "step": 5216 }, { "epoch": 2.306490251722994, "grad_norm": 0.3064938920165464, "learning_rate": 1.5439962885072673e-06, "loss": 0.2324, "step": 5217 }, { "epoch": 2.3069325175984963, "grad_norm": 0.368247255184518, "learning_rate": 1.5421379672868225e-06, "loss": 0.2688, "step": 5218 }, { "epoch": 2.3073747834739984, "grad_norm": 0.35921924475262673, "learning_rate": 1.5402805611305604e-06, "loss": 0.2562, "step": 5219 }, { "epoch": 2.3078170493495005, "grad_norm": 0.3372318386077768, "learning_rate": 1.5384240705300112e-06, "loss": 0.2583, "step": 5220 }, { "epoch": 2.3082593152250026, "grad_norm": 0.39069945022912583, "learning_rate": 1.5365684959764687e-06, "loss": 0.242, "step": 5221 }, { "epoch": 2.3087015811005047, "grad_norm": 0.35786741903793085, "learning_rate": 1.5347138379609726e-06, "loss": 0.2535, "step": 5222 }, { "epoch": 2.3091438469760073, "grad_norm": 0.3705430879772759, "learning_rate": 1.5328600969743296e-06, "loss": 0.2824, "step": 5223 }, { "epoch": 2.3095861128515094, "grad_norm": 0.32247989074263766, "learning_rate": 1.5310072735070996e-06, "loss": 0.2691, "step": 5224 }, { "epoch": 2.3100283787270115, "grad_norm": 0.3041574389106831, "learning_rate": 1.5291553680496001e-06, "loss": 0.2572, "step": 5225 }, { "epoch": 2.3104706446025136, "grad_norm": 0.334503688492561, "learning_rate": 1.5273043810919087e-06, "loss": 0.2624, "step": 5226 }, { "epoch": 2.3109129104780157, "grad_norm": 0.5793355416984003, "learning_rate": 1.5254543131238564e-06, "loss": 0.2606, "step": 5227 }, { "epoch": 2.311355176353518, "grad_norm": 0.3775409956677012, "learning_rate": 1.5236051646350314e-06, "loss": 0.267, "step": 5228 }, { "epoch": 2.31179744222902, "grad_norm": 0.37367603385555853, "learning_rate": 1.5217569361147804e-06, "loss": 0.2642, "step": 5229 }, { "epoch": 2.312239708104522, "grad_norm": 0.29735461615276715, "learning_rate": 1.5199096280522046e-06, "loss": 0.2912, "step": 5230 }, { "epoch": 2.312681973980024, "grad_norm": 0.33448914721853723, "learning_rate": 1.5180632409361628e-06, "loss": 0.269, "step": 5231 }, { "epoch": 2.3131242398555263, "grad_norm": 0.3707611751515117, "learning_rate": 1.5162177752552699e-06, "loss": 0.2695, "step": 5232 }, { "epoch": 2.3135665057310284, "grad_norm": 0.27178990053070246, "learning_rate": 1.514373231497897e-06, "loss": 0.261, "step": 5233 }, { "epoch": 2.314008771606531, "grad_norm": 0.36236239694625094, "learning_rate": 1.5125296101521713e-06, "loss": 0.2449, "step": 5234 }, { "epoch": 2.314451037482033, "grad_norm": 0.32759969620030227, "learning_rate": 1.5106869117059753e-06, "loss": 0.2747, "step": 5235 }, { "epoch": 2.314893303357535, "grad_norm": 0.3574324268915163, "learning_rate": 1.5088451366469465e-06, "loss": 0.2454, "step": 5236 }, { "epoch": 2.3153355692330373, "grad_norm": 0.27771524075221515, "learning_rate": 1.5070042854624833e-06, "loss": 0.2543, "step": 5237 }, { "epoch": 2.3157778351085394, "grad_norm": 0.36241130490214035, "learning_rate": 1.505164358639732e-06, "loss": 0.3018, "step": 5238 }, { "epoch": 2.3162201009840415, "grad_norm": 0.35185359064172106, "learning_rate": 1.503325356665598e-06, "loss": 0.2767, "step": 5239 }, { "epoch": 2.3166623668595436, "grad_norm": 0.343953124343219, "learning_rate": 1.5014872800267427e-06, "loss": 0.308, "step": 5240 }, { "epoch": 2.3171046327350457, "grad_norm": 0.40526991980031696, "learning_rate": 1.4996501292095806e-06, "loss": 0.2823, "step": 5241 }, { "epoch": 2.317546898610548, "grad_norm": 0.450748240333717, "learning_rate": 1.4978139047002844e-06, "loss": 0.2776, "step": 5242 }, { "epoch": 2.3179891644860504, "grad_norm": 0.3284080976434108, "learning_rate": 1.4959786069847781e-06, "loss": 0.2517, "step": 5243 }, { "epoch": 2.3184314303615525, "grad_norm": 0.3183563294993092, "learning_rate": 1.4941442365487424e-06, "loss": 0.2907, "step": 5244 }, { "epoch": 2.3188736962370546, "grad_norm": 0.4907794848714519, "learning_rate": 1.4923107938776127e-06, "loss": 0.2893, "step": 5245 }, { "epoch": 2.3193159621125568, "grad_norm": 0.3616905410838819, "learning_rate": 1.490478279456577e-06, "loss": 0.2633, "step": 5246 }, { "epoch": 2.319758227988059, "grad_norm": 0.42239762928290453, "learning_rate": 1.4886466937705796e-06, "loss": 0.2767, "step": 5247 }, { "epoch": 2.320200493863561, "grad_norm": 0.3897519485016764, "learning_rate": 1.4868160373043183e-06, "loss": 0.2666, "step": 5248 }, { "epoch": 2.320642759739063, "grad_norm": 0.31066389017369855, "learning_rate": 1.4849863105422452e-06, "loss": 0.2728, "step": 5249 }, { "epoch": 2.321085025614565, "grad_norm": 0.31471973437781764, "learning_rate": 1.4831575139685667e-06, "loss": 0.2864, "step": 5250 }, { "epoch": 2.3215272914900673, "grad_norm": 0.27462238176411735, "learning_rate": 1.4813296480672401e-06, "loss": 0.3029, "step": 5251 }, { "epoch": 2.3219695573655694, "grad_norm": 0.3408021813460061, "learning_rate": 1.479502713321984e-06, "loss": 0.266, "step": 5252 }, { "epoch": 2.3224118232410715, "grad_norm": 0.31173983543750744, "learning_rate": 1.4776767102162616e-06, "loss": 0.2572, "step": 5253 }, { "epoch": 2.322854089116574, "grad_norm": 0.3942254528955115, "learning_rate": 1.475851639233294e-06, "loss": 0.2681, "step": 5254 }, { "epoch": 2.323296354992076, "grad_norm": 0.346007742006841, "learning_rate": 1.4740275008560551e-06, "loss": 0.2826, "step": 5255 }, { "epoch": 2.3237386208675783, "grad_norm": 0.33113699896396104, "learning_rate": 1.4722042955672727e-06, "loss": 0.2405, "step": 5256 }, { "epoch": 2.3241808867430804, "grad_norm": 0.3611381108529766, "learning_rate": 1.4703820238494248e-06, "loss": 0.2739, "step": 5257 }, { "epoch": 2.3246231526185825, "grad_norm": 1.0925447503711456, "learning_rate": 1.4685606861847484e-06, "loss": 0.3152, "step": 5258 }, { "epoch": 2.3250654184940847, "grad_norm": 0.37842344775604725, "learning_rate": 1.466740283055227e-06, "loss": 0.2831, "step": 5259 }, { "epoch": 2.3255076843695868, "grad_norm": 0.3516041885002731, "learning_rate": 1.4649208149425993e-06, "loss": 0.2578, "step": 5260 }, { "epoch": 2.325949950245089, "grad_norm": 0.3308443139352456, "learning_rate": 1.4631022823283585e-06, "loss": 0.2653, "step": 5261 }, { "epoch": 2.326392216120591, "grad_norm": 0.42089007798654327, "learning_rate": 1.4612846856937434e-06, "loss": 0.2622, "step": 5262 }, { "epoch": 2.3268344819960936, "grad_norm": 0.27807495370774465, "learning_rate": 1.4594680255197552e-06, "loss": 0.263, "step": 5263 }, { "epoch": 2.3272767478715957, "grad_norm": 0.3033993922216572, "learning_rate": 1.457652302287139e-06, "loss": 0.2592, "step": 5264 }, { "epoch": 2.3277190137470978, "grad_norm": 0.29496541249728925, "learning_rate": 1.4558375164763961e-06, "loss": 0.2395, "step": 5265 }, { "epoch": 2.3281612796226, "grad_norm": 0.29834088294929, "learning_rate": 1.4540236685677783e-06, "loss": 0.2497, "step": 5266 }, { "epoch": 2.328603545498102, "grad_norm": 0.34453432916914717, "learning_rate": 1.4522107590412882e-06, "loss": 0.2751, "step": 5267 }, { "epoch": 2.329045811373604, "grad_norm": 0.314284975905969, "learning_rate": 1.4503987883766857e-06, "loss": 0.2738, "step": 5268 }, { "epoch": 2.3294880772491062, "grad_norm": 0.3136025432863969, "learning_rate": 1.448587757053473e-06, "loss": 0.2755, "step": 5269 }, { "epoch": 2.3299303431246083, "grad_norm": 0.37842662045696873, "learning_rate": 1.446777665550911e-06, "loss": 0.248, "step": 5270 }, { "epoch": 2.3303726090001105, "grad_norm": 0.3811610471948137, "learning_rate": 1.4449685143480086e-06, "loss": 0.2673, "step": 5271 }, { "epoch": 2.3308148748756126, "grad_norm": 0.29658288772661756, "learning_rate": 1.4431603039235275e-06, "loss": 0.2872, "step": 5272 }, { "epoch": 2.3312571407511147, "grad_norm": 0.4294148111669911, "learning_rate": 1.441353034755978e-06, "loss": 0.2698, "step": 5273 }, { "epoch": 2.3316994066266172, "grad_norm": 0.37139544811800207, "learning_rate": 1.4395467073236257e-06, "loss": 0.2773, "step": 5274 }, { "epoch": 2.3321416725021193, "grad_norm": 0.32170702188736694, "learning_rate": 1.4377413221044838e-06, "loss": 0.26, "step": 5275 }, { "epoch": 2.3325839383776215, "grad_norm": 0.31519191673430397, "learning_rate": 1.4359368795763156e-06, "loss": 0.2935, "step": 5276 }, { "epoch": 2.3330262042531236, "grad_norm": 0.4290692744868372, "learning_rate": 1.4341333802166385e-06, "loss": 0.2616, "step": 5277 }, { "epoch": 2.3334684701286257, "grad_norm": 0.2871068360781467, "learning_rate": 1.4323308245027124e-06, "loss": 0.2579, "step": 5278 }, { "epoch": 2.333910736004128, "grad_norm": 0.45275127391053044, "learning_rate": 1.4305292129115578e-06, "loss": 0.2428, "step": 5279 }, { "epoch": 2.33435300187963, "grad_norm": 0.33521432954795854, "learning_rate": 1.428728545919939e-06, "loss": 0.25, "step": 5280 }, { "epoch": 2.334795267755132, "grad_norm": 0.38752301170255987, "learning_rate": 1.4269288240043722e-06, "loss": 0.25, "step": 5281 }, { "epoch": 2.335237533630634, "grad_norm": 0.3379795769467037, "learning_rate": 1.4251300476411224e-06, "loss": 0.2706, "step": 5282 }, { "epoch": 2.3356797995061362, "grad_norm": 0.3607144939746316, "learning_rate": 1.4233322173062052e-06, "loss": 0.2782, "step": 5283 }, { "epoch": 2.3361220653816384, "grad_norm": 0.3526773969603063, "learning_rate": 1.4215353334753863e-06, "loss": 0.2703, "step": 5284 }, { "epoch": 2.336564331257141, "grad_norm": 0.27771163335717414, "learning_rate": 1.4197393966241795e-06, "loss": 0.2354, "step": 5285 }, { "epoch": 2.337006597132643, "grad_norm": 0.42901311392759833, "learning_rate": 1.4179444072278498e-06, "loss": 0.2461, "step": 5286 }, { "epoch": 2.337448863008145, "grad_norm": 0.328130006393471, "learning_rate": 1.4161503657614095e-06, "loss": 0.2709, "step": 5287 }, { "epoch": 2.3378911288836473, "grad_norm": 0.2808574508253801, "learning_rate": 1.4143572726996219e-06, "loss": 0.2944, "step": 5288 }, { "epoch": 2.3383333947591494, "grad_norm": 0.30220268394814925, "learning_rate": 1.412565128516996e-06, "loss": 0.2939, "step": 5289 }, { "epoch": 2.3387756606346515, "grad_norm": 0.38058566731461746, "learning_rate": 1.4107739336877962e-06, "loss": 0.2576, "step": 5290 }, { "epoch": 2.3392179265101536, "grad_norm": 0.38497944606899615, "learning_rate": 1.4089836886860297e-06, "loss": 0.2491, "step": 5291 }, { "epoch": 2.3396601923856557, "grad_norm": 0.25739626066399635, "learning_rate": 1.407194393985456e-06, "loss": 0.2438, "step": 5292 }, { "epoch": 2.340102458261158, "grad_norm": 0.3252265690345078, "learning_rate": 1.4054060500595784e-06, "loss": 0.2409, "step": 5293 }, { "epoch": 2.3405447241366604, "grad_norm": 0.36151621566896225, "learning_rate": 1.4036186573816513e-06, "loss": 0.2404, "step": 5294 }, { "epoch": 2.3409869900121625, "grad_norm": 0.41817647649211076, "learning_rate": 1.4018322164246806e-06, "loss": 0.2896, "step": 5295 }, { "epoch": 2.3414292558876646, "grad_norm": 0.39623609320867825, "learning_rate": 1.4000467276614166e-06, "loss": 0.2422, "step": 5296 }, { "epoch": 2.3418715217631667, "grad_norm": 0.3800960341700916, "learning_rate": 1.3982621915643573e-06, "loss": 0.2722, "step": 5297 }, { "epoch": 2.342313787638669, "grad_norm": 0.33941644238884167, "learning_rate": 1.3964786086057507e-06, "loss": 0.2725, "step": 5298 }, { "epoch": 2.342756053514171, "grad_norm": 0.45824014934475266, "learning_rate": 1.3946959792575915e-06, "loss": 0.2727, "step": 5299 }, { "epoch": 2.343198319389673, "grad_norm": 0.3705811621577748, "learning_rate": 1.392914303991622e-06, "loss": 0.2555, "step": 5300 }, { "epoch": 2.343640585265175, "grad_norm": 0.3326380796737886, "learning_rate": 1.391133583279332e-06, "loss": 0.2678, "step": 5301 }, { "epoch": 2.3440828511406773, "grad_norm": 0.48299444970358285, "learning_rate": 1.3893538175919586e-06, "loss": 0.2452, "step": 5302 }, { "epoch": 2.3445251170161794, "grad_norm": 0.32887431337293593, "learning_rate": 1.3875750074004873e-06, "loss": 0.2734, "step": 5303 }, { "epoch": 2.3449673828916815, "grad_norm": 0.41947074031016285, "learning_rate": 1.3857971531756476e-06, "loss": 0.2191, "step": 5304 }, { "epoch": 2.345409648767184, "grad_norm": 0.33769749496243534, "learning_rate": 1.3840202553879217e-06, "loss": 0.2485, "step": 5305 }, { "epoch": 2.345851914642686, "grad_norm": 0.2813187198002011, "learning_rate": 1.3822443145075331e-06, "loss": 0.2645, "step": 5306 }, { "epoch": 2.3462941805181883, "grad_norm": 0.3138743151064865, "learning_rate": 1.3804693310044554e-06, "loss": 0.2561, "step": 5307 }, { "epoch": 2.3467364463936904, "grad_norm": 0.39585181863636804, "learning_rate": 1.3786953053484082e-06, "loss": 0.2707, "step": 5308 }, { "epoch": 2.3471787122691925, "grad_norm": 0.3149486252103403, "learning_rate": 1.3769222380088543e-06, "loss": 0.2651, "step": 5309 }, { "epoch": 2.3476209781446946, "grad_norm": 0.3816285944624502, "learning_rate": 1.3751501294550062e-06, "loss": 0.2674, "step": 5310 }, { "epoch": 2.3480632440201967, "grad_norm": 0.37743454983810315, "learning_rate": 1.3733789801558245e-06, "loss": 0.2914, "step": 5311 }, { "epoch": 2.348505509895699, "grad_norm": 0.5098360089088355, "learning_rate": 1.371608790580013e-06, "loss": 0.2866, "step": 5312 }, { "epoch": 2.348947775771201, "grad_norm": 0.33403728421147977, "learning_rate": 1.3698395611960212e-06, "loss": 0.2187, "step": 5313 }, { "epoch": 2.3493900416467035, "grad_norm": 0.35193928162917154, "learning_rate": 1.3680712924720463e-06, "loss": 0.2887, "step": 5314 }, { "epoch": 2.3498323075222056, "grad_norm": 0.3126532539061418, "learning_rate": 1.36630398487603e-06, "loss": 0.2761, "step": 5315 }, { "epoch": 2.3502745733977077, "grad_norm": 0.3595269690257131, "learning_rate": 1.3645376388756603e-06, "loss": 0.311, "step": 5316 }, { "epoch": 2.35071683927321, "grad_norm": 0.3368719216800279, "learning_rate": 1.3627722549383705e-06, "loss": 0.2911, "step": 5317 }, { "epoch": 2.351159105148712, "grad_norm": 0.35634504876254514, "learning_rate": 1.3610078335313388e-06, "loss": 0.2738, "step": 5318 }, { "epoch": 2.351601371024214, "grad_norm": 0.31494538941774103, "learning_rate": 1.3592443751214895e-06, "loss": 0.2717, "step": 5319 }, { "epoch": 2.352043636899716, "grad_norm": 0.35711820996377647, "learning_rate": 1.3574818801754913e-06, "loss": 0.2822, "step": 5320 }, { "epoch": 2.3524859027752183, "grad_norm": 0.308215521342811, "learning_rate": 1.3557203491597597e-06, "loss": 0.272, "step": 5321 }, { "epoch": 2.3529281686507204, "grad_norm": 0.2928090238464702, "learning_rate": 1.3539597825404537e-06, "loss": 0.2822, "step": 5322 }, { "epoch": 2.3533704345262225, "grad_norm": 0.38004130187710755, "learning_rate": 1.3522001807834783e-06, "loss": 0.2941, "step": 5323 }, { "epoch": 2.3538127004017246, "grad_norm": 0.33096862781083997, "learning_rate": 1.3504415443544788e-06, "loss": 0.2691, "step": 5324 }, { "epoch": 2.354254966277227, "grad_norm": 0.3497706386019166, "learning_rate": 1.3486838737188501e-06, "loss": 0.2833, "step": 5325 }, { "epoch": 2.3546972321527293, "grad_norm": 0.4245942339915072, "learning_rate": 1.346927169341728e-06, "loss": 0.2778, "step": 5326 }, { "epoch": 2.3551394980282314, "grad_norm": 0.34797689796522924, "learning_rate": 1.345171431687997e-06, "loss": 0.2745, "step": 5327 }, { "epoch": 2.3555817639037335, "grad_norm": 0.4282687584989007, "learning_rate": 1.3434166612222827e-06, "loss": 0.2566, "step": 5328 }, { "epoch": 2.3560240297792356, "grad_norm": 0.29585747871623375, "learning_rate": 1.3416628584089547e-06, "loss": 0.252, "step": 5329 }, { "epoch": 2.3564662956547378, "grad_norm": 0.41361134574556774, "learning_rate": 1.3399100237121266e-06, "loss": 0.2439, "step": 5330 }, { "epoch": 2.35690856153024, "grad_norm": 0.384868770426134, "learning_rate": 1.3381581575956565e-06, "loss": 0.2642, "step": 5331 }, { "epoch": 2.357350827405742, "grad_norm": 0.25569989582353336, "learning_rate": 1.3364072605231454e-06, "loss": 0.286, "step": 5332 }, { "epoch": 2.357793093281244, "grad_norm": 0.311436537547127, "learning_rate": 1.3346573329579393e-06, "loss": 0.2639, "step": 5333 }, { "epoch": 2.3582353591567466, "grad_norm": 0.4249858942173235, "learning_rate": 1.3329083753631255e-06, "loss": 0.2754, "step": 5334 }, { "epoch": 2.3586776250322483, "grad_norm": 0.2735405887919205, "learning_rate": 1.3311603882015372e-06, "loss": 0.2506, "step": 5335 }, { "epoch": 2.359119890907751, "grad_norm": 0.35296879335849785, "learning_rate": 1.3294133719357466e-06, "loss": 0.2569, "step": 5336 }, { "epoch": 2.359562156783253, "grad_norm": 0.3909883636045889, "learning_rate": 1.3276673270280749e-06, "loss": 0.2473, "step": 5337 }, { "epoch": 2.360004422658755, "grad_norm": 0.3071267843931882, "learning_rate": 1.3259222539405825e-06, "loss": 0.3026, "step": 5338 }, { "epoch": 2.360446688534257, "grad_norm": 0.3372985408638367, "learning_rate": 1.3241781531350738e-06, "loss": 0.253, "step": 5339 }, { "epoch": 2.3608889544097593, "grad_norm": 0.322403353510536, "learning_rate": 1.3224350250730927e-06, "loss": 0.2768, "step": 5340 }, { "epoch": 2.3613312202852614, "grad_norm": 0.4330284770847923, "learning_rate": 1.3206928702159283e-06, "loss": 0.2946, "step": 5341 }, { "epoch": 2.3617734861607635, "grad_norm": 0.41968936498779075, "learning_rate": 1.3189516890246152e-06, "loss": 0.2651, "step": 5342 }, { "epoch": 2.3622157520362657, "grad_norm": 0.2737547523038903, "learning_rate": 1.3172114819599252e-06, "loss": 0.2591, "step": 5343 }, { "epoch": 2.3626580179117678, "grad_norm": 0.3588471963238358, "learning_rate": 1.3154722494823757e-06, "loss": 0.2613, "step": 5344 }, { "epoch": 2.3631002837872703, "grad_norm": 0.4056233056871189, "learning_rate": 1.3137339920522235e-06, "loss": 0.266, "step": 5345 }, { "epoch": 2.3635425496627724, "grad_norm": 0.34047216418606135, "learning_rate": 1.3119967101294694e-06, "loss": 0.2765, "step": 5346 }, { "epoch": 2.3639848155382746, "grad_norm": 0.30644596722253636, "learning_rate": 1.3102604041738553e-06, "loss": 0.2696, "step": 5347 }, { "epoch": 2.3644270814137767, "grad_norm": 0.29510254226433014, "learning_rate": 1.3085250746448652e-06, "loss": 0.2748, "step": 5348 }, { "epoch": 2.364869347289279, "grad_norm": 0.42305224575522893, "learning_rate": 1.3067907220017245e-06, "loss": 0.2627, "step": 5349 }, { "epoch": 2.365311613164781, "grad_norm": 0.2890393974378547, "learning_rate": 1.3050573467033995e-06, "loss": 0.2807, "step": 5350 }, { "epoch": 2.365753879040283, "grad_norm": 0.3472243858742114, "learning_rate": 1.3033249492085987e-06, "loss": 0.2974, "step": 5351 }, { "epoch": 2.366196144915785, "grad_norm": 0.29371457961991654, "learning_rate": 1.3015935299757698e-06, "loss": 0.2594, "step": 5352 }, { "epoch": 2.3666384107912872, "grad_norm": 0.3923489285535455, "learning_rate": 1.299863089463107e-06, "loss": 0.2739, "step": 5353 }, { "epoch": 2.3670806766667893, "grad_norm": 0.28463004423275223, "learning_rate": 1.2981336281285412e-06, "loss": 0.2454, "step": 5354 }, { "epoch": 2.3675229425422915, "grad_norm": 0.3859598738013691, "learning_rate": 1.2964051464297422e-06, "loss": 0.2919, "step": 5355 }, { "epoch": 2.367965208417794, "grad_norm": 0.31300912530229746, "learning_rate": 1.2946776448241244e-06, "loss": 0.2319, "step": 5356 }, { "epoch": 2.368407474293296, "grad_norm": 0.34646385346182096, "learning_rate": 1.2929511237688402e-06, "loss": 0.2601, "step": 5357 }, { "epoch": 2.3688497401687982, "grad_norm": 0.30063933327817355, "learning_rate": 1.2912255837207877e-06, "loss": 0.2489, "step": 5358 }, { "epoch": 2.3692920060443003, "grad_norm": 0.37230460053798176, "learning_rate": 1.2895010251365986e-06, "loss": 0.2871, "step": 5359 }, { "epoch": 2.3697342719198025, "grad_norm": 0.3709179167115677, "learning_rate": 1.2877774484726492e-06, "loss": 0.2564, "step": 5360 }, { "epoch": 2.3701765377953046, "grad_norm": 0.8425294976734312, "learning_rate": 1.2860548541850542e-06, "loss": 0.2947, "step": 5361 }, { "epoch": 2.3706188036708067, "grad_norm": 0.4592659374686397, "learning_rate": 1.2843332427296685e-06, "loss": 0.2841, "step": 5362 }, { "epoch": 2.371061069546309, "grad_norm": 0.39798285660162186, "learning_rate": 1.2826126145620877e-06, "loss": 0.2338, "step": 5363 }, { "epoch": 2.371503335421811, "grad_norm": 0.3212084858848717, "learning_rate": 1.2808929701376467e-06, "loss": 0.2429, "step": 5364 }, { "epoch": 2.3719456012973135, "grad_norm": 0.30398618939389066, "learning_rate": 1.279174309911419e-06, "loss": 0.2656, "step": 5365 }, { "epoch": 2.3723878671728156, "grad_norm": 0.39293172366221446, "learning_rate": 1.2774566343382205e-06, "loss": 0.2752, "step": 5366 }, { "epoch": 2.3728301330483177, "grad_norm": 0.3931616694395191, "learning_rate": 1.275739943872603e-06, "loss": 0.2485, "step": 5367 }, { "epoch": 2.37327239892382, "grad_norm": 0.3053278112000382, "learning_rate": 1.274024238968859e-06, "loss": 0.2536, "step": 5368 }, { "epoch": 2.373714664799322, "grad_norm": 0.3534784734336844, "learning_rate": 1.2723095200810232e-06, "loss": 0.2966, "step": 5369 }, { "epoch": 2.374156930674824, "grad_norm": 0.37861495115423077, "learning_rate": 1.270595787662866e-06, "loss": 0.2703, "step": 5370 }, { "epoch": 2.374599196550326, "grad_norm": 0.30870757043663916, "learning_rate": 1.2688830421678954e-06, "loss": 0.2659, "step": 5371 }, { "epoch": 2.3750414624258283, "grad_norm": 0.33091476567169337, "learning_rate": 1.2671712840493616e-06, "loss": 0.2916, "step": 5372 }, { "epoch": 2.3754837283013304, "grad_norm": 0.34172649439665365, "learning_rate": 1.2654605137602504e-06, "loss": 0.2631, "step": 5373 }, { "epoch": 2.3759259941768325, "grad_norm": 0.31742479641703936, "learning_rate": 1.2637507317532916e-06, "loss": 0.2632, "step": 5374 }, { "epoch": 2.3763682600523346, "grad_norm": 0.3360405461601076, "learning_rate": 1.2620419384809468e-06, "loss": 0.2782, "step": 5375 }, { "epoch": 2.376810525927837, "grad_norm": 0.3589010613683778, "learning_rate": 1.2603341343954206e-06, "loss": 0.2371, "step": 5376 }, { "epoch": 2.3772527918033393, "grad_norm": 0.3008207452419142, "learning_rate": 1.2586273199486538e-06, "loss": 0.2918, "step": 5377 }, { "epoch": 2.3776950576788414, "grad_norm": 0.3698192795361552, "learning_rate": 1.2569214955923248e-06, "loss": 0.2854, "step": 5378 }, { "epoch": 2.3781373235543435, "grad_norm": 0.3176580402117827, "learning_rate": 1.2552166617778522e-06, "loss": 0.26, "step": 5379 }, { "epoch": 2.3785795894298456, "grad_norm": 0.29532817479472095, "learning_rate": 1.2535128189563905e-06, "loss": 0.2773, "step": 5380 }, { "epoch": 2.3790218553053477, "grad_norm": 0.38388267320925246, "learning_rate": 1.2518099675788325e-06, "loss": 0.2927, "step": 5381 }, { "epoch": 2.37946412118085, "grad_norm": 0.3212940472674106, "learning_rate": 1.2501081080958088e-06, "loss": 0.2632, "step": 5382 }, { "epoch": 2.379906387056352, "grad_norm": 0.30225888252384414, "learning_rate": 1.248407240957688e-06, "loss": 0.2502, "step": 5383 }, { "epoch": 2.380348652931854, "grad_norm": 0.381217924824189, "learning_rate": 1.246707366614574e-06, "loss": 0.3047, "step": 5384 }, { "epoch": 2.3807909188073566, "grad_norm": 0.39836478775717915, "learning_rate": 1.245008485516313e-06, "loss": 0.2554, "step": 5385 }, { "epoch": 2.3812331846828583, "grad_norm": 0.2802750873206418, "learning_rate": 1.243310598112482e-06, "loss": 0.257, "step": 5386 }, { "epoch": 2.381675450558361, "grad_norm": 0.38627416499054074, "learning_rate": 1.241613704852398e-06, "loss": 0.2872, "step": 5387 }, { "epoch": 2.382117716433863, "grad_norm": 0.30092682008202687, "learning_rate": 1.2399178061851159e-06, "loss": 0.2637, "step": 5388 }, { "epoch": 2.382559982309365, "grad_norm": 0.31454815387774687, "learning_rate": 1.2382229025594239e-06, "loss": 0.2688, "step": 5389 }, { "epoch": 2.383002248184867, "grad_norm": 0.31523800126727186, "learning_rate": 1.2365289944238528e-06, "loss": 0.2414, "step": 5390 }, { "epoch": 2.3834445140603693, "grad_norm": 0.36932458695528786, "learning_rate": 1.2348360822266652e-06, "loss": 0.2658, "step": 5391 }, { "epoch": 2.3838867799358714, "grad_norm": 0.46283379543390507, "learning_rate": 1.233144166415861e-06, "loss": 0.2742, "step": 5392 }, { "epoch": 2.3843290458113735, "grad_norm": 0.3662374673273571, "learning_rate": 1.2314532474391778e-06, "loss": 0.2488, "step": 5393 }, { "epoch": 2.3847713116868756, "grad_norm": 0.30384582980454117, "learning_rate": 1.2297633257440866e-06, "loss": 0.2394, "step": 5394 }, { "epoch": 2.3852135775623777, "grad_norm": 0.309576403888022, "learning_rate": 1.2280744017777985e-06, "loss": 0.2385, "step": 5395 }, { "epoch": 2.3856558434378803, "grad_norm": 0.37073612200518985, "learning_rate": 1.2263864759872567e-06, "loss": 0.2669, "step": 5396 }, { "epoch": 2.3860981093133824, "grad_norm": 0.33085162428039916, "learning_rate": 1.2246995488191426e-06, "loss": 0.2783, "step": 5397 }, { "epoch": 2.3865403751888845, "grad_norm": 0.28748720568012653, "learning_rate": 1.2230136207198728e-06, "loss": 0.2779, "step": 5398 }, { "epoch": 2.3869826410643866, "grad_norm": 0.3361991351927183, "learning_rate": 1.2213286921355993e-06, "loss": 0.2624, "step": 5399 }, { "epoch": 2.3874249069398887, "grad_norm": 0.3713945544854088, "learning_rate": 1.2196447635122089e-06, "loss": 0.256, "step": 5400 }, { "epoch": 2.387867172815391, "grad_norm": 0.3903383054342945, "learning_rate": 1.2179618352953275e-06, "loss": 0.2588, "step": 5401 }, { "epoch": 2.388309438690893, "grad_norm": 0.28362158974269464, "learning_rate": 1.21627990793031e-06, "loss": 0.2634, "step": 5402 }, { "epoch": 2.388751704566395, "grad_norm": 0.38702249107430126, "learning_rate": 1.2145989818622512e-06, "loss": 0.2806, "step": 5403 }, { "epoch": 2.389193970441897, "grad_norm": 0.3696855796778974, "learning_rate": 1.212919057535979e-06, "loss": 0.2462, "step": 5404 }, { "epoch": 2.3896362363173993, "grad_norm": 0.38498954937052804, "learning_rate": 1.2112401353960556e-06, "loss": 0.2551, "step": 5405 }, { "epoch": 2.3900785021929014, "grad_norm": 0.31452492364398105, "learning_rate": 1.209562215886782e-06, "loss": 0.2653, "step": 5406 }, { "epoch": 2.390520768068404, "grad_norm": 0.3121557105795516, "learning_rate": 1.2078852994521894e-06, "loss": 0.2427, "step": 5407 }, { "epoch": 2.390963033943906, "grad_norm": 0.3617782164474428, "learning_rate": 1.2062093865360458e-06, "loss": 0.2516, "step": 5408 }, { "epoch": 2.391405299819408, "grad_norm": 0.36127497557508603, "learning_rate": 1.2045344775818518e-06, "loss": 0.2591, "step": 5409 }, { "epoch": 2.3918475656949103, "grad_norm": 0.292486546129355, "learning_rate": 1.202860573032844e-06, "loss": 0.2525, "step": 5410 }, { "epoch": 2.3922898315704124, "grad_norm": 0.27350442787998824, "learning_rate": 1.2011876733319928e-06, "loss": 0.2694, "step": 5411 }, { "epoch": 2.3927320974459145, "grad_norm": 0.37109177813410144, "learning_rate": 1.1995157789220025e-06, "loss": 0.2741, "step": 5412 }, { "epoch": 2.3931743633214166, "grad_norm": 0.3361113125320249, "learning_rate": 1.1978448902453111e-06, "loss": 0.2728, "step": 5413 }, { "epoch": 2.3936166291969188, "grad_norm": 0.30227587078495644, "learning_rate": 1.1961750077440915e-06, "loss": 0.2528, "step": 5414 }, { "epoch": 2.394058895072421, "grad_norm": 0.3469267129871987, "learning_rate": 1.1945061318602474e-06, "loss": 0.2382, "step": 5415 }, { "epoch": 2.3945011609479234, "grad_norm": 0.3317877891580278, "learning_rate": 1.192838263035423e-06, "loss": 0.2656, "step": 5416 }, { "epoch": 2.3949434268234255, "grad_norm": 0.3844938062199469, "learning_rate": 1.191171401710987e-06, "loss": 0.2705, "step": 5417 }, { "epoch": 2.3953856926989276, "grad_norm": 0.4017440642847551, "learning_rate": 1.1895055483280476e-06, "loss": 0.2467, "step": 5418 }, { "epoch": 2.3958279585744298, "grad_norm": 0.3213736584262775, "learning_rate": 1.1878407033274437e-06, "loss": 0.2475, "step": 5419 }, { "epoch": 2.396270224449932, "grad_norm": 0.4272255487406068, "learning_rate": 1.1861768671497487e-06, "loss": 0.2541, "step": 5420 }, { "epoch": 2.396712490325434, "grad_norm": 0.4489526045276576, "learning_rate": 1.1845140402352667e-06, "loss": 0.3096, "step": 5421 }, { "epoch": 2.397154756200936, "grad_norm": 0.28504031315595774, "learning_rate": 1.1828522230240396e-06, "loss": 0.2702, "step": 5422 }, { "epoch": 2.397597022076438, "grad_norm": 0.49764699499686055, "learning_rate": 1.1811914159558374e-06, "loss": 0.2787, "step": 5423 }, { "epoch": 2.3980392879519403, "grad_norm": 0.34466017748521366, "learning_rate": 1.1795316194701644e-06, "loss": 0.2709, "step": 5424 }, { "epoch": 2.3984815538274424, "grad_norm": 0.3168424335045117, "learning_rate": 1.177872834006259e-06, "loss": 0.2749, "step": 5425 }, { "epoch": 2.3989238197029445, "grad_norm": 0.30551898840115044, "learning_rate": 1.1762150600030858e-06, "loss": 0.2699, "step": 5426 }, { "epoch": 2.399366085578447, "grad_norm": 0.3158263242251181, "learning_rate": 1.1745582978993508e-06, "loss": 0.2464, "step": 5427 }, { "epoch": 2.399808351453949, "grad_norm": 0.3205585775720795, "learning_rate": 1.1729025481334866e-06, "loss": 0.2493, "step": 5428 }, { "epoch": 2.4002506173294513, "grad_norm": 0.33850498303732635, "learning_rate": 1.1712478111436587e-06, "loss": 0.2764, "step": 5429 }, { "epoch": 2.4006928832049534, "grad_norm": 0.29527464311234247, "learning_rate": 1.1695940873677652e-06, "loss": 0.2592, "step": 5430 }, { "epoch": 2.4011351490804556, "grad_norm": 0.392106101977807, "learning_rate": 1.167941377243435e-06, "loss": 0.2477, "step": 5431 }, { "epoch": 2.4015774149559577, "grad_norm": 0.43317262256069305, "learning_rate": 1.1662896812080337e-06, "loss": 0.2294, "step": 5432 }, { "epoch": 2.40201968083146, "grad_norm": 0.2871663589050267, "learning_rate": 1.1646389996986497e-06, "loss": 0.2586, "step": 5433 }, { "epoch": 2.402461946706962, "grad_norm": 0.29624848415689525, "learning_rate": 1.1629893331521092e-06, "loss": 0.2687, "step": 5434 }, { "epoch": 2.402904212582464, "grad_norm": 0.34555901973247577, "learning_rate": 1.1613406820049683e-06, "loss": 0.2649, "step": 5435 }, { "epoch": 2.4033464784579666, "grad_norm": 0.30029952979745894, "learning_rate": 1.1596930466935153e-06, "loss": 0.2494, "step": 5436 }, { "epoch": 2.4037887443334682, "grad_norm": 0.3647080321979302, "learning_rate": 1.1580464276537662e-06, "loss": 0.2587, "step": 5437 }, { "epoch": 2.404231010208971, "grad_norm": 0.402405973755451, "learning_rate": 1.1564008253214738e-06, "loss": 0.2474, "step": 5438 }, { "epoch": 2.404673276084473, "grad_norm": 0.27431501145978254, "learning_rate": 1.1547562401321173e-06, "loss": 0.2665, "step": 5439 }, { "epoch": 2.405115541959975, "grad_norm": 0.3765080640491979, "learning_rate": 1.1531126725209086e-06, "loss": 0.2936, "step": 5440 }, { "epoch": 2.405557807835477, "grad_norm": 0.3274338977961324, "learning_rate": 1.1514701229227905e-06, "loss": 0.2729, "step": 5441 }, { "epoch": 2.4060000737109792, "grad_norm": 0.3193667832228689, "learning_rate": 1.1498285917724316e-06, "loss": 0.2672, "step": 5442 }, { "epoch": 2.4064423395864813, "grad_norm": 0.46396263721465497, "learning_rate": 1.14818807950424e-06, "loss": 0.2471, "step": 5443 }, { "epoch": 2.4068846054619835, "grad_norm": 0.3681781257757596, "learning_rate": 1.1465485865523474e-06, "loss": 0.2708, "step": 5444 }, { "epoch": 2.4073268713374856, "grad_norm": 0.3227605682622056, "learning_rate": 1.1449101133506174e-06, "loss": 0.2605, "step": 5445 }, { "epoch": 2.4077691372129877, "grad_norm": 0.5078383926556829, "learning_rate": 1.1432726603326439e-06, "loss": 0.2792, "step": 5446 }, { "epoch": 2.4082114030884902, "grad_norm": 0.3271955575292313, "learning_rate": 1.1416362279317517e-06, "loss": 0.2701, "step": 5447 }, { "epoch": 2.4086536689639924, "grad_norm": 0.38861898646623794, "learning_rate": 1.140000816580994e-06, "loss": 0.2671, "step": 5448 }, { "epoch": 2.4090959348394945, "grad_norm": 0.39684252269116793, "learning_rate": 1.138366426713155e-06, "loss": 0.2772, "step": 5449 }, { "epoch": 2.4095382007149966, "grad_norm": 0.3846589789228408, "learning_rate": 1.1367330587607473e-06, "loss": 0.2508, "step": 5450 }, { "epoch": 2.4099804665904987, "grad_norm": 0.3510305376306423, "learning_rate": 1.1351007131560149e-06, "loss": 0.2597, "step": 5451 }, { "epoch": 2.410422732466001, "grad_norm": 0.3793080797329055, "learning_rate": 1.1334693903309279e-06, "loss": 0.2809, "step": 5452 }, { "epoch": 2.410864998341503, "grad_norm": 0.46943815980239506, "learning_rate": 1.1318390907171917e-06, "loss": 0.2341, "step": 5453 }, { "epoch": 2.411307264217005, "grad_norm": 0.32389591057913897, "learning_rate": 1.1302098147462348e-06, "loss": 0.2489, "step": 5454 }, { "epoch": 2.411749530092507, "grad_norm": 0.31348262563193546, "learning_rate": 1.1285815628492186e-06, "loss": 0.247, "step": 5455 }, { "epoch": 2.4121917959680093, "grad_norm": 0.3455213608310009, "learning_rate": 1.1269543354570323e-06, "loss": 0.2812, "step": 5456 }, { "epoch": 2.4126340618435114, "grad_norm": 0.3911472106718945, "learning_rate": 1.125328133000292e-06, "loss": 0.2634, "step": 5457 }, { "epoch": 2.413076327719014, "grad_norm": 0.34672139577996197, "learning_rate": 1.1237029559093438e-06, "loss": 0.2342, "step": 5458 }, { "epoch": 2.413518593594516, "grad_norm": 0.2868279027564801, "learning_rate": 1.1220788046142662e-06, "loss": 0.2257, "step": 5459 }, { "epoch": 2.413960859470018, "grad_norm": 0.4023465371565488, "learning_rate": 1.1204556795448617e-06, "loss": 0.2376, "step": 5460 }, { "epoch": 2.4144031253455203, "grad_norm": 0.4056384395568956, "learning_rate": 1.1188335811306628e-06, "loss": 0.2648, "step": 5461 }, { "epoch": 2.4148453912210224, "grad_norm": 0.3480384287981068, "learning_rate": 1.11721250980093e-06, "loss": 0.2328, "step": 5462 }, { "epoch": 2.4152876570965245, "grad_norm": 0.40608273674300027, "learning_rate": 1.1155924659846523e-06, "loss": 0.2666, "step": 5463 }, { "epoch": 2.4157299229720266, "grad_norm": 0.4239171247234556, "learning_rate": 1.113973450110546e-06, "loss": 0.2552, "step": 5464 }, { "epoch": 2.4161721888475287, "grad_norm": 0.3016826789935935, "learning_rate": 1.112355462607057e-06, "loss": 0.2666, "step": 5465 }, { "epoch": 2.416614454723031, "grad_norm": 0.3006509040442978, "learning_rate": 1.1107385039023578e-06, "loss": 0.2471, "step": 5466 }, { "epoch": 2.4170567205985334, "grad_norm": 0.3428165705605757, "learning_rate": 1.1091225744243483e-06, "loss": 0.2499, "step": 5467 }, { "epoch": 2.4174989864740355, "grad_norm": 0.3402861679599362, "learning_rate": 1.1075076746006557e-06, "loss": 0.2498, "step": 5468 }, { "epoch": 2.4179412523495376, "grad_norm": 0.3869705619395585, "learning_rate": 1.1058938048586388e-06, "loss": 0.2798, "step": 5469 }, { "epoch": 2.4183835182250397, "grad_norm": 0.31989936540648467, "learning_rate": 1.1042809656253789e-06, "loss": 0.2479, "step": 5470 }, { "epoch": 2.418825784100542, "grad_norm": 0.3893346850974202, "learning_rate": 1.1026691573276866e-06, "loss": 0.2972, "step": 5471 }, { "epoch": 2.419268049976044, "grad_norm": 0.3063433023768944, "learning_rate": 1.1010583803921004e-06, "loss": 0.2823, "step": 5472 }, { "epoch": 2.419710315851546, "grad_norm": 0.3115701047457834, "learning_rate": 1.0994486352448825e-06, "loss": 0.2621, "step": 5473 }, { "epoch": 2.420152581727048, "grad_norm": 0.34190609166630526, "learning_rate": 1.0978399223120246e-06, "loss": 0.2661, "step": 5474 }, { "epoch": 2.4205948476025503, "grad_norm": 0.42584872314092936, "learning_rate": 1.0962322420192478e-06, "loss": 0.2774, "step": 5475 }, { "epoch": 2.4210371134780524, "grad_norm": 0.4218523316365277, "learning_rate": 1.0946255947919959e-06, "loss": 0.2473, "step": 5476 }, { "epoch": 2.4214793793535545, "grad_norm": 0.38583657766644974, "learning_rate": 1.0930199810554403e-06, "loss": 0.232, "step": 5477 }, { "epoch": 2.421921645229057, "grad_norm": 0.3393450373956439, "learning_rate": 1.0914154012344802e-06, "loss": 0.2752, "step": 5478 }, { "epoch": 2.422363911104559, "grad_norm": 0.37897342324205757, "learning_rate": 1.089811855753739e-06, "loss": 0.2891, "step": 5479 }, { "epoch": 2.4228061769800613, "grad_norm": 0.30767560138625916, "learning_rate": 1.0882093450375692e-06, "loss": 0.259, "step": 5480 }, { "epoch": 2.4232484428555634, "grad_norm": 0.2977740864483657, "learning_rate": 1.0866078695100464e-06, "loss": 0.2734, "step": 5481 }, { "epoch": 2.4236907087310655, "grad_norm": 0.30890579644750155, "learning_rate": 1.085007429594974e-06, "loss": 0.2445, "step": 5482 }, { "epoch": 2.4241329746065676, "grad_norm": 0.2758901549167245, "learning_rate": 1.0834080257158825e-06, "loss": 0.2654, "step": 5483 }, { "epoch": 2.4245752404820697, "grad_norm": 0.38832246281537836, "learning_rate": 1.0818096582960237e-06, "loss": 0.2427, "step": 5484 }, { "epoch": 2.425017506357572, "grad_norm": 0.30105352416752074, "learning_rate": 1.080212327758382e-06, "loss": 0.2827, "step": 5485 }, { "epoch": 2.425459772233074, "grad_norm": 0.34343971813691465, "learning_rate": 1.078616034525662e-06, "loss": 0.2423, "step": 5486 }, { "epoch": 2.4259020381085765, "grad_norm": 0.6163651526673447, "learning_rate": 1.0770207790202974e-06, "loss": 0.2736, "step": 5487 }, { "epoch": 2.426344303984078, "grad_norm": 0.47609546962236754, "learning_rate": 1.0754265616644417e-06, "loss": 0.2889, "step": 5488 }, { "epoch": 2.4267865698595807, "grad_norm": 0.2957290517229188, "learning_rate": 1.073833382879978e-06, "loss": 0.2682, "step": 5489 }, { "epoch": 2.427228835735083, "grad_norm": 0.3782637650931899, "learning_rate": 1.0722412430885159e-06, "loss": 0.2624, "step": 5490 }, { "epoch": 2.427671101610585, "grad_norm": 0.3613416505817324, "learning_rate": 1.0706501427113874e-06, "loss": 0.2421, "step": 5491 }, { "epoch": 2.428113367486087, "grad_norm": 0.4826124313512447, "learning_rate": 1.06906008216965e-06, "loss": 0.3131, "step": 5492 }, { "epoch": 2.428555633361589, "grad_norm": 0.3716279548392287, "learning_rate": 1.0674710618840857e-06, "loss": 0.272, "step": 5493 }, { "epoch": 2.4289978992370913, "grad_norm": 0.3410182172624561, "learning_rate": 1.065883082275202e-06, "loss": 0.2651, "step": 5494 }, { "epoch": 2.4294401651125934, "grad_norm": 0.3574013928189775, "learning_rate": 1.0642961437632298e-06, "loss": 0.2546, "step": 5495 }, { "epoch": 2.4298824309880955, "grad_norm": 0.43014483191495523, "learning_rate": 1.0627102467681267e-06, "loss": 0.274, "step": 5496 }, { "epoch": 2.4303246968635976, "grad_norm": 0.3584874074207424, "learning_rate": 1.061125391709572e-06, "loss": 0.2646, "step": 5497 }, { "epoch": 2.4307669627391, "grad_norm": 0.30862298798008514, "learning_rate": 1.0595415790069712e-06, "loss": 0.2977, "step": 5498 }, { "epoch": 2.4312092286146023, "grad_norm": 0.3586320876008346, "learning_rate": 1.057958809079453e-06, "loss": 0.2389, "step": 5499 }, { "epoch": 2.4316514944901044, "grad_norm": 0.31483126578128423, "learning_rate": 1.0563770823458696e-06, "loss": 0.2655, "step": 5500 }, { "epoch": 2.4320937603656065, "grad_norm": 0.421136770870697, "learning_rate": 1.0547963992248e-06, "loss": 0.2613, "step": 5501 }, { "epoch": 2.4325360262411087, "grad_norm": 0.31427134552000147, "learning_rate": 1.053216760134544e-06, "loss": 0.2548, "step": 5502 }, { "epoch": 2.4329782921166108, "grad_norm": 0.3861329618440259, "learning_rate": 1.0516381654931274e-06, "loss": 0.2511, "step": 5503 }, { "epoch": 2.433420557992113, "grad_norm": 0.26384670765831947, "learning_rate": 1.0500606157182957e-06, "loss": 0.2478, "step": 5504 }, { "epoch": 2.433862823867615, "grad_norm": 0.2929118677828147, "learning_rate": 1.0484841112275207e-06, "loss": 0.2678, "step": 5505 }, { "epoch": 2.434305089743117, "grad_norm": 0.27291309789027934, "learning_rate": 1.046908652438e-06, "loss": 0.2884, "step": 5506 }, { "epoch": 2.434747355618619, "grad_norm": 0.3646777105419769, "learning_rate": 1.0453342397666506e-06, "loss": 0.2632, "step": 5507 }, { "epoch": 2.4351896214941213, "grad_norm": 0.3204573776483638, "learning_rate": 1.0437608736301136e-06, "loss": 0.2442, "step": 5508 }, { "epoch": 2.435631887369624, "grad_norm": 0.38702439372827907, "learning_rate": 1.0421885544447534e-06, "loss": 0.2934, "step": 5509 }, { "epoch": 2.436074153245126, "grad_norm": 0.40904366456586505, "learning_rate": 1.0406172826266581e-06, "loss": 0.2635, "step": 5510 }, { "epoch": 2.436516419120628, "grad_norm": 0.3621503880226485, "learning_rate": 1.0390470585916378e-06, "loss": 0.2918, "step": 5511 }, { "epoch": 2.43695868499613, "grad_norm": 0.3568792574099095, "learning_rate": 1.037477882755225e-06, "loss": 0.2683, "step": 5512 }, { "epoch": 2.4374009508716323, "grad_norm": 0.4082195746516879, "learning_rate": 1.0359097555326764e-06, "loss": 0.304, "step": 5513 }, { "epoch": 2.4378432167471344, "grad_norm": 0.27388828610854915, "learning_rate": 1.0343426773389687e-06, "loss": 0.2349, "step": 5514 }, { "epoch": 2.4382854826226366, "grad_norm": 0.4054840732699976, "learning_rate": 1.032776648588803e-06, "loss": 0.2707, "step": 5515 }, { "epoch": 2.4387277484981387, "grad_norm": 0.4050993333617965, "learning_rate": 1.0312116696966012e-06, "loss": 0.2983, "step": 5516 }, { "epoch": 2.439170014373641, "grad_norm": 0.25693536989462945, "learning_rate": 1.0296477410765105e-06, "loss": 0.2583, "step": 5517 }, { "epoch": 2.4396122802491433, "grad_norm": 0.29823868525989655, "learning_rate": 1.0280848631423979e-06, "loss": 0.2535, "step": 5518 }, { "epoch": 2.4400545461246455, "grad_norm": 0.28801165114601013, "learning_rate": 1.0265230363078498e-06, "loss": 0.2805, "step": 5519 }, { "epoch": 2.4404968120001476, "grad_norm": 0.30330960432408305, "learning_rate": 1.024962260986178e-06, "loss": 0.2624, "step": 5520 }, { "epoch": 2.4409390778756497, "grad_norm": 0.26818450977938635, "learning_rate": 1.0234025375904144e-06, "loss": 0.2764, "step": 5521 }, { "epoch": 2.441381343751152, "grad_norm": 0.5048474845689568, "learning_rate": 1.0218438665333159e-06, "loss": 0.2562, "step": 5522 }, { "epoch": 2.441823609626654, "grad_norm": 0.3972653260476032, "learning_rate": 1.0202862482273556e-06, "loss": 0.2932, "step": 5523 }, { "epoch": 2.442265875502156, "grad_norm": 0.3564478389881755, "learning_rate": 1.0187296830847322e-06, "loss": 0.2184, "step": 5524 }, { "epoch": 2.442708141377658, "grad_norm": 0.6082418849603981, "learning_rate": 1.0171741715173633e-06, "loss": 0.2564, "step": 5525 }, { "epoch": 2.4431504072531602, "grad_norm": 0.34607812205403726, "learning_rate": 1.0156197139368884e-06, "loss": 0.2666, "step": 5526 }, { "epoch": 2.4435926731286624, "grad_norm": 0.33486078079495807, "learning_rate": 1.0140663107546684e-06, "loss": 0.2693, "step": 5527 }, { "epoch": 2.4440349390041645, "grad_norm": 0.34019527328975707, "learning_rate": 1.0125139623817847e-06, "loss": 0.2668, "step": 5528 }, { "epoch": 2.444477204879667, "grad_norm": 0.3058829228135631, "learning_rate": 1.0109626692290404e-06, "loss": 0.2458, "step": 5529 }, { "epoch": 2.444919470755169, "grad_norm": 0.3375570875323558, "learning_rate": 1.0094124317069582e-06, "loss": 0.2872, "step": 5530 }, { "epoch": 2.4453617366306712, "grad_norm": 0.35398142043193037, "learning_rate": 1.0078632502257824e-06, "loss": 0.2509, "step": 5531 }, { "epoch": 2.4458040025061734, "grad_norm": 0.5107124932592204, "learning_rate": 1.0063151251954756e-06, "loss": 0.2774, "step": 5532 }, { "epoch": 2.4462462683816755, "grad_norm": 0.46849179107540145, "learning_rate": 1.0047680570257256e-06, "loss": 0.252, "step": 5533 }, { "epoch": 2.4466885342571776, "grad_norm": 0.35891057401880444, "learning_rate": 1.0032220461259374e-06, "loss": 0.2743, "step": 5534 }, { "epoch": 2.4471308001326797, "grad_norm": 0.3746613243547175, "learning_rate": 1.0016770929052345e-06, "loss": 0.2732, "step": 5535 }, { "epoch": 2.447573066008182, "grad_norm": 0.30617089487107835, "learning_rate": 1.0001331977724631e-06, "loss": 0.2639, "step": 5536 }, { "epoch": 2.448015331883684, "grad_norm": 0.3107656676852079, "learning_rate": 9.985903611361874e-07, "loss": 0.2833, "step": 5537 }, { "epoch": 2.4484575977591865, "grad_norm": 0.3533314498689717, "learning_rate": 9.97048583404695e-07, "loss": 0.2583, "step": 5538 }, { "epoch": 2.448899863634688, "grad_norm": 0.3400864958269832, "learning_rate": 9.955078649859912e-07, "loss": 0.2552, "step": 5539 }, { "epoch": 2.4493421295101907, "grad_norm": 0.36079584936665876, "learning_rate": 9.939682062877991e-07, "loss": 0.2507, "step": 5540 }, { "epoch": 2.449784395385693, "grad_norm": 0.2558357268364526, "learning_rate": 9.924296077175643e-07, "loss": 0.2587, "step": 5541 }, { "epoch": 2.450226661261195, "grad_norm": 0.35478810571368663, "learning_rate": 9.908920696824503e-07, "loss": 0.2676, "step": 5542 }, { "epoch": 2.450668927136697, "grad_norm": 0.34940900811535724, "learning_rate": 9.8935559258934e-07, "loss": 0.293, "step": 5543 }, { "epoch": 2.451111193012199, "grad_norm": 0.3118921905606024, "learning_rate": 9.878201768448364e-07, "loss": 0.2795, "step": 5544 }, { "epoch": 2.4515534588877013, "grad_norm": 0.4518467147077409, "learning_rate": 9.862858228552608e-07, "loss": 0.2647, "step": 5545 }, { "epoch": 2.4519957247632034, "grad_norm": 0.38097226467856, "learning_rate": 9.847525310266537e-07, "loss": 0.2564, "step": 5546 }, { "epoch": 2.4524379906387055, "grad_norm": 0.2834647604318961, "learning_rate": 9.832203017647746e-07, "loss": 0.2314, "step": 5547 }, { "epoch": 2.4528802565142076, "grad_norm": 0.31803825207862574, "learning_rate": 9.816891354751007e-07, "loss": 0.2922, "step": 5548 }, { "epoch": 2.45332252238971, "grad_norm": 0.28045324191664, "learning_rate": 9.80159032562833e-07, "loss": 0.2467, "step": 5549 }, { "epoch": 2.4537647882652123, "grad_norm": 0.36315872212042344, "learning_rate": 9.78629993432883e-07, "loss": 0.2589, "step": 5550 }, { "epoch": 2.4542070541407144, "grad_norm": 0.3196598864710218, "learning_rate": 9.77102018489886e-07, "loss": 0.2647, "step": 5551 }, { "epoch": 2.4546493200162165, "grad_norm": 0.34589177458050685, "learning_rate": 9.755751081381948e-07, "loss": 0.2831, "step": 5552 }, { "epoch": 2.4550915858917186, "grad_norm": 0.3103452779530683, "learning_rate": 9.740492627818793e-07, "loss": 0.2893, "step": 5553 }, { "epoch": 2.4555338517672207, "grad_norm": 0.3772792636338558, "learning_rate": 9.725244828247298e-07, "loss": 0.2799, "step": 5554 }, { "epoch": 2.455976117642723, "grad_norm": 0.2882517496910659, "learning_rate": 9.710007686702533e-07, "loss": 0.265, "step": 5555 }, { "epoch": 2.456418383518225, "grad_norm": 0.3727335186074405, "learning_rate": 9.694781207216736e-07, "loss": 0.296, "step": 5556 }, { "epoch": 2.456860649393727, "grad_norm": 0.3433832752504979, "learning_rate": 9.679565393819347e-07, "loss": 0.2904, "step": 5557 }, { "epoch": 2.457302915269229, "grad_norm": 0.29960125171359675, "learning_rate": 9.664360250536959e-07, "loss": 0.2622, "step": 5558 }, { "epoch": 2.4577451811447313, "grad_norm": 0.44077378979057, "learning_rate": 9.649165781393361e-07, "loss": 0.2778, "step": 5559 }, { "epoch": 2.458187447020234, "grad_norm": 0.37509353471933, "learning_rate": 9.633981990409508e-07, "loss": 0.2613, "step": 5560 }, { "epoch": 2.458629712895736, "grad_norm": 0.3497110236711511, "learning_rate": 9.618808881603537e-07, "loss": 0.2438, "step": 5561 }, { "epoch": 2.459071978771238, "grad_norm": 0.40783011301821154, "learning_rate": 9.60364645899074e-07, "loss": 0.2471, "step": 5562 }, { "epoch": 2.45951424464674, "grad_norm": 0.32973792140013036, "learning_rate": 9.588494726583592e-07, "loss": 0.2625, "step": 5563 }, { "epoch": 2.4599565105222423, "grad_norm": 0.3856207383106128, "learning_rate": 9.573353688391762e-07, "loss": 0.2469, "step": 5564 }, { "epoch": 2.4603987763977444, "grad_norm": 0.3094892607085166, "learning_rate": 9.55822334842207e-07, "loss": 0.289, "step": 5565 }, { "epoch": 2.4608410422732465, "grad_norm": 0.3765022183273263, "learning_rate": 9.543103710678469e-07, "loss": 0.2724, "step": 5566 }, { "epoch": 2.4612833081487486, "grad_norm": 0.36135077744148464, "learning_rate": 9.527994779162131e-07, "loss": 0.2865, "step": 5567 }, { "epoch": 2.4617255740242507, "grad_norm": 0.7520247308168936, "learning_rate": 9.512896557871381e-07, "loss": 0.2752, "step": 5568 }, { "epoch": 2.4621678398997533, "grad_norm": 0.39559265513807934, "learning_rate": 9.497809050801681e-07, "loss": 0.2547, "step": 5569 }, { "epoch": 2.4626101057752554, "grad_norm": 0.3142490023280314, "learning_rate": 9.482732261945726e-07, "loss": 0.2296, "step": 5570 }, { "epoch": 2.4630523716507575, "grad_norm": 0.2917975676165803, "learning_rate": 9.467666195293302e-07, "loss": 0.2684, "step": 5571 }, { "epoch": 2.4634946375262596, "grad_norm": 0.3757845907008518, "learning_rate": 9.45261085483139e-07, "loss": 0.2708, "step": 5572 }, { "epoch": 2.4639369034017617, "grad_norm": 0.32438781578628795, "learning_rate": 9.437566244544133e-07, "loss": 0.2475, "step": 5573 }, { "epoch": 2.464379169277264, "grad_norm": 0.31115873125665033, "learning_rate": 9.422532368412829e-07, "loss": 0.2682, "step": 5574 }, { "epoch": 2.464821435152766, "grad_norm": 0.32197490298416825, "learning_rate": 9.40750923041594e-07, "loss": 0.2857, "step": 5575 }, { "epoch": 2.465263701028268, "grad_norm": 0.2717452056389632, "learning_rate": 9.392496834529074e-07, "loss": 0.263, "step": 5576 }, { "epoch": 2.46570596690377, "grad_norm": 0.45623558463126773, "learning_rate": 9.37749518472501e-07, "loss": 0.3155, "step": 5577 }, { "epoch": 2.4661482327792723, "grad_norm": 0.36061751025386213, "learning_rate": 9.362504284973683e-07, "loss": 0.2764, "step": 5578 }, { "epoch": 2.4665904986547744, "grad_norm": 0.3442888194559739, "learning_rate": 9.347524139242159e-07, "loss": 0.2833, "step": 5579 }, { "epoch": 2.467032764530277, "grad_norm": 0.3004758890088855, "learning_rate": 9.332554751494727e-07, "loss": 0.2656, "step": 5580 }, { "epoch": 2.467475030405779, "grad_norm": 0.48831913935853627, "learning_rate": 9.317596125692724e-07, "loss": 0.2663, "step": 5581 }, { "epoch": 2.467917296281281, "grad_norm": 0.32579685983296747, "learning_rate": 9.302648265794728e-07, "loss": 0.2385, "step": 5582 }, { "epoch": 2.4683595621567833, "grad_norm": 0.3656760815706805, "learning_rate": 9.28771117575642e-07, "loss": 0.281, "step": 5583 }, { "epoch": 2.4688018280322854, "grad_norm": 0.3832675251406915, "learning_rate": 9.272784859530654e-07, "loss": 0.2791, "step": 5584 }, { "epoch": 2.4692440939077875, "grad_norm": 0.31034323683403675, "learning_rate": 9.257869321067404e-07, "loss": 0.2625, "step": 5585 }, { "epoch": 2.4696863597832897, "grad_norm": 0.35459171854252347, "learning_rate": 9.24296456431385e-07, "loss": 0.2731, "step": 5586 }, { "epoch": 2.4701286256587918, "grad_norm": 0.3931557170866955, "learning_rate": 9.228070593214261e-07, "loss": 0.2684, "step": 5587 }, { "epoch": 2.470570891534294, "grad_norm": 0.2869170394621469, "learning_rate": 9.213187411710072e-07, "loss": 0.2462, "step": 5588 }, { "epoch": 2.4710131574097964, "grad_norm": 0.3106167310987155, "learning_rate": 9.198315023739879e-07, "loss": 0.2544, "step": 5589 }, { "epoch": 2.471455423285298, "grad_norm": 0.36723719410237426, "learning_rate": 9.183453433239364e-07, "loss": 0.2594, "step": 5590 }, { "epoch": 2.4718976891608007, "grad_norm": 0.3669842783643791, "learning_rate": 9.168602644141439e-07, "loss": 0.2531, "step": 5591 }, { "epoch": 2.4723399550363028, "grad_norm": 0.42515826329608014, "learning_rate": 9.153762660376086e-07, "loss": 0.314, "step": 5592 }, { "epoch": 2.472782220911805, "grad_norm": 0.3504567856256131, "learning_rate": 9.138933485870466e-07, "loss": 0.2504, "step": 5593 }, { "epoch": 2.473224486787307, "grad_norm": 0.45815154745895026, "learning_rate": 9.12411512454886e-07, "loss": 0.2749, "step": 5594 }, { "epoch": 2.473666752662809, "grad_norm": 0.36076952930060996, "learning_rate": 9.109307580332688e-07, "loss": 0.2767, "step": 5595 }, { "epoch": 2.4741090185383112, "grad_norm": 0.31147953172627835, "learning_rate": 9.094510857140549e-07, "loss": 0.2947, "step": 5596 }, { "epoch": 2.4745512844138133, "grad_norm": 0.33846853572625274, "learning_rate": 9.079724958888098e-07, "loss": 0.254, "step": 5597 }, { "epoch": 2.4749935502893154, "grad_norm": 0.37694841063920886, "learning_rate": 9.064949889488195e-07, "loss": 0.2659, "step": 5598 }, { "epoch": 2.4754358161648176, "grad_norm": 0.35418861327108586, "learning_rate": 9.050185652850802e-07, "loss": 0.2475, "step": 5599 }, { "epoch": 2.47587808204032, "grad_norm": 0.3000324287765355, "learning_rate": 9.035432252883014e-07, "loss": 0.2439, "step": 5600 }, { "epoch": 2.4763203479158222, "grad_norm": 0.3304415289097863, "learning_rate": 9.020689693489087e-07, "loss": 0.2455, "step": 5601 }, { "epoch": 2.4767626137913243, "grad_norm": 0.3208651157474744, "learning_rate": 9.005957978570379e-07, "loss": 0.2521, "step": 5602 }, { "epoch": 2.4772048796668265, "grad_norm": 0.3062118669410128, "learning_rate": 8.991237112025392e-07, "loss": 0.2504, "step": 5603 }, { "epoch": 2.4776471455423286, "grad_norm": 0.377774063604695, "learning_rate": 8.976527097749737e-07, "loss": 0.2785, "step": 5604 }, { "epoch": 2.4780894114178307, "grad_norm": 0.325179842239, "learning_rate": 8.961827939636198e-07, "loss": 0.2614, "step": 5605 }, { "epoch": 2.478531677293333, "grad_norm": 0.27238045357473784, "learning_rate": 8.947139641574609e-07, "loss": 0.2712, "step": 5606 }, { "epoch": 2.478973943168835, "grad_norm": 0.3647329996682585, "learning_rate": 8.932462207452014e-07, "loss": 0.2864, "step": 5607 }, { "epoch": 2.479416209044337, "grad_norm": 0.32346580689901144, "learning_rate": 8.917795641152538e-07, "loss": 0.2651, "step": 5608 }, { "epoch": 2.479858474919839, "grad_norm": 0.3592601596985119, "learning_rate": 8.903139946557437e-07, "loss": 0.2456, "step": 5609 }, { "epoch": 2.4803007407953412, "grad_norm": 0.36621389929493636, "learning_rate": 8.888495127545088e-07, "loss": 0.2637, "step": 5610 }, { "epoch": 2.480743006670844, "grad_norm": 0.36928886522340054, "learning_rate": 8.873861187990989e-07, "loss": 0.2533, "step": 5611 }, { "epoch": 2.481185272546346, "grad_norm": 0.31343479527775386, "learning_rate": 8.859238131767772e-07, "loss": 0.2554, "step": 5612 }, { "epoch": 2.481627538421848, "grad_norm": 0.2813433541524397, "learning_rate": 8.844625962745174e-07, "loss": 0.2316, "step": 5613 }, { "epoch": 2.48206980429735, "grad_norm": 0.3675451654171158, "learning_rate": 8.830024684790056e-07, "loss": 0.2625, "step": 5614 }, { "epoch": 2.4825120701728522, "grad_norm": 0.4281912301509955, "learning_rate": 8.815434301766396e-07, "loss": 0.2713, "step": 5615 }, { "epoch": 2.4829543360483544, "grad_norm": 0.39119523822567864, "learning_rate": 8.800854817535281e-07, "loss": 0.237, "step": 5616 }, { "epoch": 2.4833966019238565, "grad_norm": 0.3682876074692755, "learning_rate": 8.786286235954943e-07, "loss": 0.2687, "step": 5617 }, { "epoch": 2.4838388677993586, "grad_norm": 0.35079174496855886, "learning_rate": 8.771728560880704e-07, "loss": 0.2614, "step": 5618 }, { "epoch": 2.4842811336748607, "grad_norm": 0.3093127624007145, "learning_rate": 8.757181796164993e-07, "loss": 0.2564, "step": 5619 }, { "epoch": 2.4847233995503633, "grad_norm": 0.41813466234118174, "learning_rate": 8.742645945657385e-07, "loss": 0.2768, "step": 5620 }, { "epoch": 2.4851656654258654, "grad_norm": 0.5477535194888661, "learning_rate": 8.72812101320451e-07, "loss": 0.2586, "step": 5621 }, { "epoch": 2.4856079313013675, "grad_norm": 0.35195886851697705, "learning_rate": 8.713607002650154e-07, "loss": 0.2912, "step": 5622 }, { "epoch": 2.4860501971768696, "grad_norm": 0.5022362216049552, "learning_rate": 8.699103917835228e-07, "loss": 0.2901, "step": 5623 }, { "epoch": 2.4864924630523717, "grad_norm": 0.3928680617627069, "learning_rate": 8.6846117625977e-07, "loss": 0.2653, "step": 5624 }, { "epoch": 2.486934728927874, "grad_norm": 0.5205236859946165, "learning_rate": 8.670130540772686e-07, "loss": 0.2564, "step": 5625 }, { "epoch": 2.487376994803376, "grad_norm": 0.4234247779047252, "learning_rate": 8.655660256192383e-07, "loss": 0.2702, "step": 5626 }, { "epoch": 2.487819260678878, "grad_norm": 0.2956915000151826, "learning_rate": 8.641200912686116e-07, "loss": 0.2528, "step": 5627 }, { "epoch": 2.48826152655438, "grad_norm": 0.35335455799480897, "learning_rate": 8.626752514080295e-07, "loss": 0.2833, "step": 5628 }, { "epoch": 2.4887037924298823, "grad_norm": 0.27227450713070256, "learning_rate": 8.61231506419844e-07, "loss": 0.2519, "step": 5629 }, { "epoch": 2.4891460583053844, "grad_norm": 0.3510185351898216, "learning_rate": 8.597888566861184e-07, "loss": 0.2898, "step": 5630 }, { "epoch": 2.489588324180887, "grad_norm": 0.4237268171141241, "learning_rate": 8.583473025886252e-07, "loss": 0.2876, "step": 5631 }, { "epoch": 2.490030590056389, "grad_norm": 0.35769653132658863, "learning_rate": 8.569068445088458e-07, "loss": 0.276, "step": 5632 }, { "epoch": 2.490472855931891, "grad_norm": 0.38720955076740027, "learning_rate": 8.554674828279747e-07, "loss": 0.2676, "step": 5633 }, { "epoch": 2.4909151218073933, "grad_norm": 0.3269559797306377, "learning_rate": 8.540292179269139e-07, "loss": 0.2852, "step": 5634 }, { "epoch": 2.4913573876828954, "grad_norm": 0.2809843484358712, "learning_rate": 8.525920501862755e-07, "loss": 0.2324, "step": 5635 }, { "epoch": 2.4917996535583975, "grad_norm": 0.3829409791615502, "learning_rate": 8.511559799863822e-07, "loss": 0.3051, "step": 5636 }, { "epoch": 2.4922419194338996, "grad_norm": 0.35235790851358034, "learning_rate": 8.497210077072626e-07, "loss": 0.2718, "step": 5637 }, { "epoch": 2.4926841853094017, "grad_norm": 0.269190333378876, "learning_rate": 8.482871337286607e-07, "loss": 0.2574, "step": 5638 }, { "epoch": 2.493126451184904, "grad_norm": 0.3496412858492224, "learning_rate": 8.468543584300259e-07, "loss": 0.2786, "step": 5639 }, { "epoch": 2.4935687170604064, "grad_norm": 0.3385703926930264, "learning_rate": 8.454226821905171e-07, "loss": 0.2753, "step": 5640 }, { "epoch": 2.494010982935908, "grad_norm": 0.36100541500378014, "learning_rate": 8.439921053890037e-07, "loss": 0.2659, "step": 5641 }, { "epoch": 2.4944532488114106, "grad_norm": 0.3088073184780034, "learning_rate": 8.425626284040634e-07, "loss": 0.2432, "step": 5642 }, { "epoch": 2.4948955146869127, "grad_norm": 0.27238352292148804, "learning_rate": 8.41134251613982e-07, "loss": 0.2637, "step": 5643 }, { "epoch": 2.495337780562415, "grad_norm": 0.43153349216857584, "learning_rate": 8.397069753967562e-07, "loss": 0.2835, "step": 5644 }, { "epoch": 2.495780046437917, "grad_norm": 0.3740844713709341, "learning_rate": 8.382808001300897e-07, "loss": 0.2824, "step": 5645 }, { "epoch": 2.496222312313419, "grad_norm": 0.45864245601304526, "learning_rate": 8.368557261913957e-07, "loss": 0.2959, "step": 5646 }, { "epoch": 2.496664578188921, "grad_norm": 0.3403562896664517, "learning_rate": 8.354317539577955e-07, "loss": 0.2836, "step": 5647 }, { "epoch": 2.4971068440644233, "grad_norm": 0.35424592987857345, "learning_rate": 8.340088838061178e-07, "loss": 0.229, "step": 5648 }, { "epoch": 2.4975491099399254, "grad_norm": 0.34422452549692173, "learning_rate": 8.325871161129034e-07, "loss": 0.2805, "step": 5649 }, { "epoch": 2.4979913758154275, "grad_norm": 0.2989328285673686, "learning_rate": 8.311664512543977e-07, "loss": 0.33, "step": 5650 }, { "epoch": 2.49843364169093, "grad_norm": 0.4400186642274676, "learning_rate": 8.297468896065564e-07, "loss": 0.2515, "step": 5651 }, { "epoch": 2.498875907566432, "grad_norm": 0.32123254547054797, "learning_rate": 8.283284315450407e-07, "loss": 0.2689, "step": 5652 }, { "epoch": 2.4993181734419343, "grad_norm": 0.33481183854989166, "learning_rate": 8.269110774452199e-07, "loss": 0.284, "step": 5653 }, { "epoch": 2.4997604393174364, "grad_norm": 0.3044067938957695, "learning_rate": 8.254948276821761e-07, "loss": 0.288, "step": 5654 }, { "epoch": 2.5002027051929385, "grad_norm": 0.31628748123062456, "learning_rate": 8.24079682630694e-07, "loss": 0.2672, "step": 5655 }, { "epoch": 2.5006449710684406, "grad_norm": 0.28610437197690974, "learning_rate": 8.226656426652679e-07, "loss": 0.2478, "step": 5656 }, { "epoch": 2.5010872369439427, "grad_norm": 0.4147545976509776, "learning_rate": 8.21252708160099e-07, "loss": 0.2862, "step": 5657 }, { "epoch": 2.501529502819445, "grad_norm": 0.41413885930641037, "learning_rate": 8.198408794890966e-07, "loss": 0.2829, "step": 5658 }, { "epoch": 2.501971768694947, "grad_norm": 0.3134478643380484, "learning_rate": 8.184301570258768e-07, "loss": 0.2555, "step": 5659 }, { "epoch": 2.5024140345704495, "grad_norm": 0.28624127628509555, "learning_rate": 8.170205411437631e-07, "loss": 0.2862, "step": 5660 }, { "epoch": 2.502856300445951, "grad_norm": 0.3265598709617427, "learning_rate": 8.156120322157868e-07, "loss": 0.2676, "step": 5661 }, { "epoch": 2.5032985663214538, "grad_norm": 0.30044392646718987, "learning_rate": 8.142046306146856e-07, "loss": 0.2723, "step": 5662 }, { "epoch": 2.503740832196956, "grad_norm": 0.32749397563879656, "learning_rate": 8.127983367129039e-07, "loss": 0.2744, "step": 5663 }, { "epoch": 2.504183098072458, "grad_norm": 0.3585165412724209, "learning_rate": 8.113931508825929e-07, "loss": 0.2436, "step": 5664 }, { "epoch": 2.50462536394796, "grad_norm": 0.2914637419438621, "learning_rate": 8.099890734956123e-07, "loss": 0.2448, "step": 5665 }, { "epoch": 2.505067629823462, "grad_norm": 0.37259891615378277, "learning_rate": 8.085861049235272e-07, "loss": 0.2649, "step": 5666 }, { "epoch": 2.5055098956989643, "grad_norm": 0.3529952813799786, "learning_rate": 8.0718424553761e-07, "loss": 0.2664, "step": 5667 }, { "epoch": 2.5059521615744664, "grad_norm": 0.42503447371025865, "learning_rate": 8.057834957088367e-07, "loss": 0.2428, "step": 5668 }, { "epoch": 2.5063944274499685, "grad_norm": 0.3192465065037529, "learning_rate": 8.043838558078915e-07, "loss": 0.28, "step": 5669 }, { "epoch": 2.5068366933254707, "grad_norm": 0.3149534986723207, "learning_rate": 8.029853262051679e-07, "loss": 0.2566, "step": 5670 }, { "epoch": 2.507278959200973, "grad_norm": 0.4521141502891238, "learning_rate": 8.015879072707611e-07, "loss": 0.2684, "step": 5671 }, { "epoch": 2.507721225076475, "grad_norm": 0.33242450565172704, "learning_rate": 8.001915993744752e-07, "loss": 0.2811, "step": 5672 }, { "epoch": 2.5081634909519774, "grad_norm": 0.30212288907900137, "learning_rate": 7.98796402885818e-07, "loss": 0.2606, "step": 5673 }, { "epoch": 2.5086057568274795, "grad_norm": 0.31075343560849994, "learning_rate": 7.974023181740059e-07, "loss": 0.2336, "step": 5674 }, { "epoch": 2.5090480227029817, "grad_norm": 0.38352954711163084, "learning_rate": 7.960093456079587e-07, "loss": 0.3097, "step": 5675 }, { "epoch": 2.5094902885784838, "grad_norm": 0.3222645177808848, "learning_rate": 7.946174855563027e-07, "loss": 0.2514, "step": 5676 }, { "epoch": 2.509932554453986, "grad_norm": 0.339410658019023, "learning_rate": 7.932267383873704e-07, "loss": 0.2689, "step": 5677 }, { "epoch": 2.510374820329488, "grad_norm": 0.32941628766943465, "learning_rate": 7.918371044691992e-07, "loss": 0.2529, "step": 5678 }, { "epoch": 2.51081708620499, "grad_norm": 0.32462289157681784, "learning_rate": 7.904485841695315e-07, "loss": 0.2528, "step": 5679 }, { "epoch": 2.5112593520804922, "grad_norm": 0.3166034727232006, "learning_rate": 7.89061177855815e-07, "loss": 0.2883, "step": 5680 }, { "epoch": 2.5117016179559943, "grad_norm": 0.2999105720673099, "learning_rate": 7.876748858952055e-07, "loss": 0.249, "step": 5681 }, { "epoch": 2.512143883831497, "grad_norm": 0.27611207018957934, "learning_rate": 7.862897086545606e-07, "loss": 0.2713, "step": 5682 }, { "epoch": 2.512586149706999, "grad_norm": 0.3084757284140895, "learning_rate": 7.849056465004424e-07, "loss": 0.26, "step": 5683 }, { "epoch": 2.513028415582501, "grad_norm": 0.424567697390641, "learning_rate": 7.835226997991202e-07, "loss": 0.2766, "step": 5684 }, { "epoch": 2.5134706814580032, "grad_norm": 0.34544007447809694, "learning_rate": 7.82140868916566e-07, "loss": 0.2422, "step": 5685 }, { "epoch": 2.5139129473335053, "grad_norm": 0.4836984671804525, "learning_rate": 7.807601542184606e-07, "loss": 0.2909, "step": 5686 }, { "epoch": 2.5143552132090075, "grad_norm": 0.3816438874415674, "learning_rate": 7.793805560701845e-07, "loss": 0.2454, "step": 5687 }, { "epoch": 2.5147974790845096, "grad_norm": 0.3733596746567231, "learning_rate": 7.780020748368256e-07, "loss": 0.2762, "step": 5688 }, { "epoch": 2.5152397449600117, "grad_norm": 0.34996324603898255, "learning_rate": 7.766247108831749e-07, "loss": 0.2784, "step": 5689 }, { "epoch": 2.515682010835514, "grad_norm": 0.331299948176832, "learning_rate": 7.752484645737285e-07, "loss": 0.245, "step": 5690 }, { "epoch": 2.5161242767110163, "grad_norm": 0.3786114532071442, "learning_rate": 7.738733362726869e-07, "loss": 0.2607, "step": 5691 }, { "epoch": 2.516566542586518, "grad_norm": 0.3214841543973751, "learning_rate": 7.724993263439534e-07, "loss": 0.276, "step": 5692 }, { "epoch": 2.5170088084620206, "grad_norm": 0.3345659551999126, "learning_rate": 7.711264351511371e-07, "loss": 0.2527, "step": 5693 }, { "epoch": 2.5174510743375227, "grad_norm": 0.37646534424485817, "learning_rate": 7.697546630575497e-07, "loss": 0.2619, "step": 5694 }, { "epoch": 2.517893340213025, "grad_norm": 0.3058725313020821, "learning_rate": 7.683840104262081e-07, "loss": 0.2886, "step": 5695 }, { "epoch": 2.518335606088527, "grad_norm": 0.4078835395534565, "learning_rate": 7.670144776198301e-07, "loss": 0.2414, "step": 5696 }, { "epoch": 2.518777871964029, "grad_norm": 0.32857788132441906, "learning_rate": 7.656460650008418e-07, "loss": 0.2697, "step": 5697 }, { "epoch": 2.519220137839531, "grad_norm": 0.28632978750260507, "learning_rate": 7.6427877293137e-07, "loss": 0.2603, "step": 5698 }, { "epoch": 2.5196624037150332, "grad_norm": 0.33736128967400253, "learning_rate": 7.629126017732436e-07, "loss": 0.3255, "step": 5699 }, { "epoch": 2.5201046695905354, "grad_norm": 0.3012979363186249, "learning_rate": 7.615475518879967e-07, "loss": 0.2545, "step": 5700 }, { "epoch": 2.5205469354660375, "grad_norm": 0.2895484200894655, "learning_rate": 7.601836236368659e-07, "loss": 0.2582, "step": 5701 }, { "epoch": 2.52098920134154, "grad_norm": 0.37656182165767155, "learning_rate": 7.588208173807943e-07, "loss": 0.2535, "step": 5702 }, { "epoch": 2.521431467217042, "grad_norm": 0.32795654915519035, "learning_rate": 7.574591334804233e-07, "loss": 0.2669, "step": 5703 }, { "epoch": 2.5218737330925443, "grad_norm": 0.2911400143730167, "learning_rate": 7.560985722960995e-07, "loss": 0.2671, "step": 5704 }, { "epoch": 2.5223159989680464, "grad_norm": 0.3636584893359287, "learning_rate": 7.547391341878724e-07, "loss": 0.2809, "step": 5705 }, { "epoch": 2.5227582648435485, "grad_norm": 0.3333502785749958, "learning_rate": 7.53380819515494e-07, "loss": 0.2585, "step": 5706 }, { "epoch": 2.5232005307190506, "grad_norm": 0.3706515805089631, "learning_rate": 7.520236286384191e-07, "loss": 0.2654, "step": 5707 }, { "epoch": 2.5236427965945527, "grad_norm": 0.322465702892497, "learning_rate": 7.506675619158055e-07, "loss": 0.2854, "step": 5708 }, { "epoch": 2.524085062470055, "grad_norm": 0.28815703288126787, "learning_rate": 7.49312619706512e-07, "loss": 0.2616, "step": 5709 }, { "epoch": 2.524527328345557, "grad_norm": 0.35751346852624233, "learning_rate": 7.479588023691026e-07, "loss": 0.2364, "step": 5710 }, { "epoch": 2.5249695942210595, "grad_norm": 0.34633475677532743, "learning_rate": 7.466061102618404e-07, "loss": 0.2687, "step": 5711 }, { "epoch": 2.525411860096561, "grad_norm": 0.4125402097409979, "learning_rate": 7.452545437426922e-07, "loss": 0.2851, "step": 5712 }, { "epoch": 2.5258541259720637, "grad_norm": 0.31838268159666605, "learning_rate": 7.439041031693294e-07, "loss": 0.2618, "step": 5713 }, { "epoch": 2.526296391847566, "grad_norm": 0.37835914835220946, "learning_rate": 7.425547888991208e-07, "loss": 0.2914, "step": 5714 }, { "epoch": 2.526738657723068, "grad_norm": 0.44564887581280405, "learning_rate": 7.4120660128914e-07, "loss": 0.246, "step": 5715 }, { "epoch": 2.52718092359857, "grad_norm": 0.37635148774872856, "learning_rate": 7.398595406961617e-07, "loss": 0.2346, "step": 5716 }, { "epoch": 2.527623189474072, "grad_norm": 0.3516831493019079, "learning_rate": 7.385136074766614e-07, "loss": 0.2634, "step": 5717 }, { "epoch": 2.5280654553495743, "grad_norm": 0.3381953804238181, "learning_rate": 7.371688019868201e-07, "loss": 0.2662, "step": 5718 }, { "epoch": 2.5285077212250764, "grad_norm": 0.2822347887541551, "learning_rate": 7.35825124582516e-07, "loss": 0.2253, "step": 5719 }, { "epoch": 2.5289499871005785, "grad_norm": 0.37112617730715786, "learning_rate": 7.344825756193302e-07, "loss": 0.2751, "step": 5720 }, { "epoch": 2.5293922529760806, "grad_norm": 0.2846027176940892, "learning_rate": 7.331411554525458e-07, "loss": 0.2606, "step": 5721 }, { "epoch": 2.529834518851583, "grad_norm": 0.3325046200597836, "learning_rate": 7.318008644371471e-07, "loss": 0.2592, "step": 5722 }, { "epoch": 2.530276784727085, "grad_norm": 0.41618750365889623, "learning_rate": 7.304617029278182e-07, "loss": 0.2922, "step": 5723 }, { "epoch": 2.5307190506025874, "grad_norm": 0.2676330103732248, "learning_rate": 7.291236712789462e-07, "loss": 0.2397, "step": 5724 }, { "epoch": 2.5311613164780895, "grad_norm": 0.3220877617108733, "learning_rate": 7.277867698446184e-07, "loss": 0.2644, "step": 5725 }, { "epoch": 2.5316035823535916, "grad_norm": 0.30877878943777765, "learning_rate": 7.264509989786222e-07, "loss": 0.2883, "step": 5726 }, { "epoch": 2.5320458482290937, "grad_norm": 0.31632961468571263, "learning_rate": 7.251163590344462e-07, "loss": 0.2467, "step": 5727 }, { "epoch": 2.532488114104596, "grad_norm": 0.30018011440677556, "learning_rate": 7.237828503652822e-07, "loss": 0.2807, "step": 5728 }, { "epoch": 2.532930379980098, "grad_norm": 0.34945320552183917, "learning_rate": 7.2245047332402e-07, "loss": 0.2596, "step": 5729 }, { "epoch": 2.5333726458556, "grad_norm": 0.31450861968119115, "learning_rate": 7.211192282632485e-07, "loss": 0.2641, "step": 5730 }, { "epoch": 2.533814911731102, "grad_norm": 0.4582091192230337, "learning_rate": 7.197891155352604e-07, "loss": 0.24, "step": 5731 }, { "epoch": 2.5342571776066043, "grad_norm": 0.3589197310754844, "learning_rate": 7.184601354920467e-07, "loss": 0.2569, "step": 5732 }, { "epoch": 2.534699443482107, "grad_norm": 0.3622720520559415, "learning_rate": 7.171322884852988e-07, "loss": 0.2476, "step": 5733 }, { "epoch": 2.535141709357609, "grad_norm": 0.3300178579217263, "learning_rate": 7.158055748664105e-07, "loss": 0.253, "step": 5734 }, { "epoch": 2.535583975233111, "grad_norm": 0.35982810860512865, "learning_rate": 7.144799949864734e-07, "loss": 0.2808, "step": 5735 }, { "epoch": 2.536026241108613, "grad_norm": 0.38268879517666177, "learning_rate": 7.131555491962794e-07, "loss": 0.2567, "step": 5736 }, { "epoch": 2.5364685069841153, "grad_norm": 0.43774396151851097, "learning_rate": 7.118322378463199e-07, "loss": 0.2613, "step": 5737 }, { "epoch": 2.5369107728596174, "grad_norm": 0.29233044695716603, "learning_rate": 7.105100612867877e-07, "loss": 0.2537, "step": 5738 }, { "epoch": 2.5373530387351195, "grad_norm": 0.3032725269325722, "learning_rate": 7.091890198675733e-07, "loss": 0.2483, "step": 5739 }, { "epoch": 2.5377953046106216, "grad_norm": 0.3921609380885764, "learning_rate": 7.078691139382676e-07, "loss": 0.2722, "step": 5740 }, { "epoch": 2.5382375704861238, "grad_norm": 0.34357603736734216, "learning_rate": 7.065503438481625e-07, "loss": 0.2838, "step": 5741 }, { "epoch": 2.5386798363616263, "grad_norm": 0.3132688887473724, "learning_rate": 7.05232709946247e-07, "loss": 0.3029, "step": 5742 }, { "epoch": 2.539122102237128, "grad_norm": 0.37531325452403724, "learning_rate": 7.039162125812093e-07, "loss": 0.2848, "step": 5743 }, { "epoch": 2.5395643681126305, "grad_norm": 0.32811160814877854, "learning_rate": 7.026008521014404e-07, "loss": 0.2771, "step": 5744 }, { "epoch": 2.5400066339881326, "grad_norm": 0.3517642410857932, "learning_rate": 7.012866288550286e-07, "loss": 0.2824, "step": 5745 }, { "epoch": 2.5404488998636348, "grad_norm": 0.3274072430678713, "learning_rate": 6.999735431897575e-07, "loss": 0.2584, "step": 5746 }, { "epoch": 2.540891165739137, "grad_norm": 0.3080326462246038, "learning_rate": 6.986615954531139e-07, "loss": 0.2548, "step": 5747 }, { "epoch": 2.541333431614639, "grad_norm": 0.29757281274217207, "learning_rate": 6.973507859922835e-07, "loss": 0.2266, "step": 5748 }, { "epoch": 2.541775697490141, "grad_norm": 0.3137035163007523, "learning_rate": 6.960411151541474e-07, "loss": 0.2556, "step": 5749 }, { "epoch": 2.542217963365643, "grad_norm": 0.3273951366167265, "learning_rate": 6.947325832852903e-07, "loss": 0.2656, "step": 5750 }, { "epoch": 2.5426602292411453, "grad_norm": 0.3633924838443234, "learning_rate": 6.934251907319917e-07, "loss": 0.2732, "step": 5751 }, { "epoch": 2.5431024951166474, "grad_norm": 0.31336311652092347, "learning_rate": 6.921189378402304e-07, "loss": 0.2601, "step": 5752 }, { "epoch": 2.54354476099215, "grad_norm": 0.3502217124187296, "learning_rate": 6.908138249556862e-07, "loss": 0.2654, "step": 5753 }, { "epoch": 2.543987026867652, "grad_norm": 0.45445186503536145, "learning_rate": 6.895098524237298e-07, "loss": 0.3089, "step": 5754 }, { "epoch": 2.544429292743154, "grad_norm": 0.3034285599511004, "learning_rate": 6.882070205894404e-07, "loss": 0.2343, "step": 5755 }, { "epoch": 2.5448715586186563, "grad_norm": 0.28530060145857994, "learning_rate": 6.86905329797588e-07, "loss": 0.2623, "step": 5756 }, { "epoch": 2.5453138244941584, "grad_norm": 0.39241436120441237, "learning_rate": 6.85604780392643e-07, "loss": 0.2942, "step": 5757 }, { "epoch": 2.5457560903696606, "grad_norm": 0.3537899316052233, "learning_rate": 6.843053727187743e-07, "loss": 0.2658, "step": 5758 }, { "epoch": 2.5461983562451627, "grad_norm": 0.48549886505620476, "learning_rate": 6.830071071198458e-07, "loss": 0.2596, "step": 5759 }, { "epoch": 2.5466406221206648, "grad_norm": 0.3558284658554342, "learning_rate": 6.817099839394259e-07, "loss": 0.257, "step": 5760 }, { "epoch": 2.547082887996167, "grad_norm": 0.35381791912109695, "learning_rate": 6.804140035207713e-07, "loss": 0.2799, "step": 5761 }, { "epoch": 2.5475251538716694, "grad_norm": 0.46339535476445326, "learning_rate": 6.791191662068436e-07, "loss": 0.2366, "step": 5762 }, { "epoch": 2.547967419747171, "grad_norm": 0.3647087462050346, "learning_rate": 6.778254723402983e-07, "loss": 0.25, "step": 5763 }, { "epoch": 2.5484096856226737, "grad_norm": 0.7745914266805539, "learning_rate": 6.765329222634892e-07, "loss": 0.2733, "step": 5764 }, { "epoch": 2.548851951498176, "grad_norm": 0.4314924353000768, "learning_rate": 6.752415163184689e-07, "loss": 0.2463, "step": 5765 }, { "epoch": 2.549294217373678, "grad_norm": 0.36456535323589373, "learning_rate": 6.739512548469862e-07, "loss": 0.2523, "step": 5766 }, { "epoch": 2.54973648324918, "grad_norm": 0.43414016979864517, "learning_rate": 6.726621381904852e-07, "loss": 0.2575, "step": 5767 }, { "epoch": 2.550178749124682, "grad_norm": 0.31733359622688373, "learning_rate": 6.713741666901096e-07, "loss": 0.2767, "step": 5768 }, { "epoch": 2.5506210150001842, "grad_norm": 0.35471428553215834, "learning_rate": 6.700873406867004e-07, "loss": 0.2441, "step": 5769 }, { "epoch": 2.5510632808756863, "grad_norm": 0.3513460567180963, "learning_rate": 6.688016605207903e-07, "loss": 0.2836, "step": 5770 }, { "epoch": 2.5515055467511885, "grad_norm": 0.29890777776775657, "learning_rate": 6.675171265326158e-07, "loss": 0.2536, "step": 5771 }, { "epoch": 2.5519478126266906, "grad_norm": 0.3119080587847136, "learning_rate": 6.66233739062106e-07, "loss": 0.2732, "step": 5772 }, { "epoch": 2.552390078502193, "grad_norm": 0.32640761284821346, "learning_rate": 6.649514984488876e-07, "loss": 0.2582, "step": 5773 }, { "epoch": 2.552832344377695, "grad_norm": 0.3245116848556874, "learning_rate": 6.636704050322834e-07, "loss": 0.2762, "step": 5774 }, { "epoch": 2.5532746102531974, "grad_norm": 0.34741106052272586, "learning_rate": 6.623904591513125e-07, "loss": 0.2696, "step": 5775 }, { "epoch": 2.5537168761286995, "grad_norm": 0.31723466441824316, "learning_rate": 6.611116611446933e-07, "loss": 0.2882, "step": 5776 }, { "epoch": 2.5541591420042016, "grad_norm": 0.4095507287607178, "learning_rate": 6.598340113508351e-07, "loss": 0.2808, "step": 5777 }, { "epoch": 2.5546014078797037, "grad_norm": 0.41957550277994604, "learning_rate": 6.585575101078467e-07, "loss": 0.2357, "step": 5778 }, { "epoch": 2.555043673755206, "grad_norm": 0.3507897094700749, "learning_rate": 6.572821577535327e-07, "loss": 0.2672, "step": 5779 }, { "epoch": 2.555485939630708, "grad_norm": 0.324742362146611, "learning_rate": 6.560079546253916e-07, "loss": 0.264, "step": 5780 }, { "epoch": 2.55592820550621, "grad_norm": 0.405139760088357, "learning_rate": 6.54734901060623e-07, "loss": 0.2704, "step": 5781 }, { "epoch": 2.5563704713817126, "grad_norm": 0.38560820886712494, "learning_rate": 6.534629973961171e-07, "loss": 0.2659, "step": 5782 }, { "epoch": 2.5568127372572143, "grad_norm": 0.3339909010489933, "learning_rate": 6.521922439684608e-07, "loss": 0.2771, "step": 5783 }, { "epoch": 2.557255003132717, "grad_norm": 0.3828603010002521, "learning_rate": 6.509226411139396e-07, "loss": 0.262, "step": 5784 }, { "epoch": 2.557697269008219, "grad_norm": 0.3388936723519341, "learning_rate": 6.496541891685293e-07, "loss": 0.2907, "step": 5785 }, { "epoch": 2.558139534883721, "grad_norm": 0.33980994550430843, "learning_rate": 6.48386888467904e-07, "loss": 0.2914, "step": 5786 }, { "epoch": 2.558581800759223, "grad_norm": 0.337479934664289, "learning_rate": 6.471207393474355e-07, "loss": 0.277, "step": 5787 }, { "epoch": 2.5590240666347253, "grad_norm": 0.28405397744875105, "learning_rate": 6.458557421421879e-07, "loss": 0.2447, "step": 5788 }, { "epoch": 2.5594663325102274, "grad_norm": 0.3444329561241975, "learning_rate": 6.445918971869209e-07, "loss": 0.2698, "step": 5789 }, { "epoch": 2.5599085983857295, "grad_norm": 0.3346882636697464, "learning_rate": 6.433292048160888e-07, "loss": 0.2944, "step": 5790 }, { "epoch": 2.5603508642612316, "grad_norm": 0.3329271958802929, "learning_rate": 6.420676653638425e-07, "loss": 0.2383, "step": 5791 }, { "epoch": 2.5607931301367337, "grad_norm": 0.33632881103937606, "learning_rate": 6.408072791640269e-07, "loss": 0.2654, "step": 5792 }, { "epoch": 2.5612353960122363, "grad_norm": 0.29431832998626317, "learning_rate": 6.395480465501808e-07, "loss": 0.2586, "step": 5793 }, { "epoch": 2.561677661887738, "grad_norm": 0.3612828331910208, "learning_rate": 6.382899678555398e-07, "loss": 0.2531, "step": 5794 }, { "epoch": 2.5621199277632405, "grad_norm": 0.31408447054965544, "learning_rate": 6.370330434130317e-07, "loss": 0.2656, "step": 5795 }, { "epoch": 2.5625621936387426, "grad_norm": 0.31861295329232214, "learning_rate": 6.357772735552803e-07, "loss": 0.3054, "step": 5796 }, { "epoch": 2.5630044595142447, "grad_norm": 0.3150239172304592, "learning_rate": 6.345226586146047e-07, "loss": 0.2813, "step": 5797 }, { "epoch": 2.563446725389747, "grad_norm": 0.35808688605995803, "learning_rate": 6.332691989230172e-07, "loss": 0.2613, "step": 5798 }, { "epoch": 2.563888991265249, "grad_norm": 0.4852937228639185, "learning_rate": 6.320168948122235e-07, "loss": 0.2462, "step": 5799 }, { "epoch": 2.564331257140751, "grad_norm": 0.6875061865826065, "learning_rate": 6.307657466136269e-07, "loss": 0.2734, "step": 5800 }, { "epoch": 2.564773523016253, "grad_norm": 0.4593161962575283, "learning_rate": 6.295157546583181e-07, "loss": 0.2711, "step": 5801 }, { "epoch": 2.5652157888917553, "grad_norm": 0.34771295839801125, "learning_rate": 6.282669192770896e-07, "loss": 0.291, "step": 5802 }, { "epoch": 2.5656580547672574, "grad_norm": 0.32364347600571497, "learning_rate": 6.270192408004233e-07, "loss": 0.2748, "step": 5803 }, { "epoch": 2.56610032064276, "grad_norm": 0.32068490038923964, "learning_rate": 6.257727195584967e-07, "loss": 0.2552, "step": 5804 }, { "epoch": 2.566542586518262, "grad_norm": 0.5302295294867605, "learning_rate": 6.245273558811793e-07, "loss": 0.3082, "step": 5805 }, { "epoch": 2.566984852393764, "grad_norm": 0.4088656501759794, "learning_rate": 6.232831500980363e-07, "loss": 0.2676, "step": 5806 }, { "epoch": 2.5674271182692663, "grad_norm": 0.3052127928107752, "learning_rate": 6.220401025383249e-07, "loss": 0.2796, "step": 5807 }, { "epoch": 2.5678693841447684, "grad_norm": 0.35551123639521753, "learning_rate": 6.207982135309964e-07, "loss": 0.2433, "step": 5808 }, { "epoch": 2.5683116500202705, "grad_norm": 0.3625644313076304, "learning_rate": 6.195574834046964e-07, "loss": 0.2642, "step": 5809 }, { "epoch": 2.5687539158957726, "grad_norm": 0.35247821695735876, "learning_rate": 6.18317912487762e-07, "loss": 0.2823, "step": 5810 }, { "epoch": 2.5691961817712747, "grad_norm": 0.30209515177392426, "learning_rate": 6.170795011082248e-07, "loss": 0.2696, "step": 5811 }, { "epoch": 2.569638447646777, "grad_norm": 0.3683438323390054, "learning_rate": 6.158422495938076e-07, "loss": 0.3026, "step": 5812 }, { "epoch": 2.5700807135222794, "grad_norm": 0.3330403581771962, "learning_rate": 6.146061582719315e-07, "loss": 0.2559, "step": 5813 }, { "epoch": 2.570522979397781, "grad_norm": 0.2805323620305868, "learning_rate": 6.133712274697046e-07, "loss": 0.285, "step": 5814 }, { "epoch": 2.5709652452732836, "grad_norm": 0.358076948125364, "learning_rate": 6.121374575139311e-07, "loss": 0.2487, "step": 5815 }, { "epoch": 2.5714075111487857, "grad_norm": 0.3653191096815574, "learning_rate": 6.109048487311075e-07, "loss": 0.2638, "step": 5816 }, { "epoch": 2.571849777024288, "grad_norm": 0.3724472131792029, "learning_rate": 6.096734014474198e-07, "loss": 0.253, "step": 5817 }, { "epoch": 2.57229204289979, "grad_norm": 0.4014301660714565, "learning_rate": 6.084431159887528e-07, "loss": 0.2609, "step": 5818 }, { "epoch": 2.572734308775292, "grad_norm": 0.5976481008291328, "learning_rate": 6.072139926806792e-07, "loss": 0.3194, "step": 5819 }, { "epoch": 2.573176574650794, "grad_norm": 0.34489818472885814, "learning_rate": 6.059860318484662e-07, "loss": 0.2753, "step": 5820 }, { "epoch": 2.5736188405262963, "grad_norm": 0.38297819520786197, "learning_rate": 6.047592338170715e-07, "loss": 0.2558, "step": 5821 }, { "epoch": 2.5740611064017984, "grad_norm": 0.3308502759822303, "learning_rate": 6.035335989111473e-07, "loss": 0.2672, "step": 5822 }, { "epoch": 2.5745033722773005, "grad_norm": 0.2857652774399557, "learning_rate": 6.023091274550369e-07, "loss": 0.2369, "step": 5823 }, { "epoch": 2.574945638152803, "grad_norm": 0.33178614095921494, "learning_rate": 6.010858197727753e-07, "loss": 0.2482, "step": 5824 }, { "epoch": 2.5753879040283048, "grad_norm": 0.2960429949693921, "learning_rate": 5.998636761880899e-07, "loss": 0.2938, "step": 5825 }, { "epoch": 2.5758301699038073, "grad_norm": 0.598207428225949, "learning_rate": 5.986426970244009e-07, "loss": 0.2752, "step": 5826 }, { "epoch": 2.5762724357793094, "grad_norm": 0.37175121004401107, "learning_rate": 5.97422882604819e-07, "loss": 0.2627, "step": 5827 }, { "epoch": 2.5767147016548115, "grad_norm": 0.3114275372611911, "learning_rate": 5.962042332521467e-07, "loss": 0.2566, "step": 5828 }, { "epoch": 2.5771569675303136, "grad_norm": 0.3493385674802645, "learning_rate": 5.9498674928888e-07, "loss": 0.2604, "step": 5829 }, { "epoch": 2.5775992334058158, "grad_norm": 0.27968334061815764, "learning_rate": 5.93770431037205e-07, "loss": 0.2942, "step": 5830 }, { "epoch": 2.578041499281318, "grad_norm": 0.34859337800711826, "learning_rate": 5.925552788190003e-07, "loss": 0.2444, "step": 5831 }, { "epoch": 2.57848376515682, "grad_norm": 0.3174100557681211, "learning_rate": 5.913412929558332e-07, "loss": 0.2803, "step": 5832 }, { "epoch": 2.5789260310323225, "grad_norm": 0.40242709668079324, "learning_rate": 5.901284737689644e-07, "loss": 0.2994, "step": 5833 }, { "epoch": 2.579368296907824, "grad_norm": 0.3453231401499252, "learning_rate": 5.889168215793473e-07, "loss": 0.2634, "step": 5834 }, { "epoch": 2.5798105627833268, "grad_norm": 0.3008220539187619, "learning_rate": 5.877063367076252e-07, "loss": 0.2793, "step": 5835 }, { "epoch": 2.580252828658829, "grad_norm": 0.39958439122267025, "learning_rate": 5.864970194741309e-07, "loss": 0.3, "step": 5836 }, { "epoch": 2.580695094534331, "grad_norm": 0.44228679555064343, "learning_rate": 5.852888701988906e-07, "loss": 0.2538, "step": 5837 }, { "epoch": 2.581137360409833, "grad_norm": 0.43318876310159193, "learning_rate": 5.840818892016198e-07, "loss": 0.3206, "step": 5838 }, { "epoch": 2.581579626285335, "grad_norm": 0.3970682433244259, "learning_rate": 5.828760768017256e-07, "loss": 0.2963, "step": 5839 }, { "epoch": 2.5820218921608373, "grad_norm": 0.3597766463193483, "learning_rate": 5.816714333183055e-07, "loss": 0.2554, "step": 5840 }, { "epoch": 2.5824641580363394, "grad_norm": 0.35794938774545726, "learning_rate": 5.804679590701479e-07, "loss": 0.2898, "step": 5841 }, { "epoch": 2.5829064239118416, "grad_norm": 0.4456364176066104, "learning_rate": 5.792656543757324e-07, "loss": 0.2958, "step": 5842 }, { "epoch": 2.5833486897873437, "grad_norm": 0.3685446493103738, "learning_rate": 5.780645195532276e-07, "loss": 0.244, "step": 5843 }, { "epoch": 2.5837909556628462, "grad_norm": 0.3202921742894392, "learning_rate": 5.768645549204921e-07, "loss": 0.269, "step": 5844 }, { "epoch": 2.584233221538348, "grad_norm": 0.36623250080868747, "learning_rate": 5.756657607950788e-07, "loss": 0.2529, "step": 5845 }, { "epoch": 2.5846754874138504, "grad_norm": 0.3217912432366482, "learning_rate": 5.744681374942273e-07, "loss": 0.2789, "step": 5846 }, { "epoch": 2.5851177532893526, "grad_norm": 0.33882277473762246, "learning_rate": 5.732716853348685e-07, "loss": 0.2528, "step": 5847 }, { "epoch": 2.5855600191648547, "grad_norm": 0.35289850733165806, "learning_rate": 5.720764046336214e-07, "loss": 0.2792, "step": 5848 }, { "epoch": 2.586002285040357, "grad_norm": 0.34333211575397937, "learning_rate": 5.708822957067961e-07, "loss": 0.2545, "step": 5849 }, { "epoch": 2.586444550915859, "grad_norm": 0.30814652258466163, "learning_rate": 5.696893588703956e-07, "loss": 0.2644, "step": 5850 }, { "epoch": 2.586886816791361, "grad_norm": 0.37625006722716514, "learning_rate": 5.684975944401094e-07, "loss": 0.2616, "step": 5851 }, { "epoch": 2.587329082666863, "grad_norm": 0.3022760570164476, "learning_rate": 5.673070027313177e-07, "loss": 0.2607, "step": 5852 }, { "epoch": 2.5877713485423652, "grad_norm": 0.31731645519603785, "learning_rate": 5.661175840590899e-07, "loss": 0.2269, "step": 5853 }, { "epoch": 2.5882136144178673, "grad_norm": 0.3088685934594086, "learning_rate": 5.649293387381854e-07, "loss": 0.2964, "step": 5854 }, { "epoch": 2.58865588029337, "grad_norm": 0.3262176498704167, "learning_rate": 5.637422670830534e-07, "loss": 0.2459, "step": 5855 }, { "epoch": 2.589098146168872, "grad_norm": 0.45036272119342446, "learning_rate": 5.625563694078318e-07, "loss": 0.2807, "step": 5856 }, { "epoch": 2.589540412044374, "grad_norm": 0.3026033070179577, "learning_rate": 5.613716460263485e-07, "loss": 0.2688, "step": 5857 }, { "epoch": 2.5899826779198762, "grad_norm": 0.34377313128948966, "learning_rate": 5.601880972521195e-07, "loss": 0.2766, "step": 5858 }, { "epoch": 2.5904249437953784, "grad_norm": 0.37778392824748747, "learning_rate": 5.59005723398352e-07, "loss": 0.2522, "step": 5859 }, { "epoch": 2.5908672096708805, "grad_norm": 0.3194073150219822, "learning_rate": 5.57824524777939e-07, "loss": 0.2539, "step": 5860 }, { "epoch": 2.5913094755463826, "grad_norm": 0.32117305046114075, "learning_rate": 5.566445017034677e-07, "loss": 0.2468, "step": 5861 }, { "epoch": 2.5917517414218847, "grad_norm": 0.33274286223182603, "learning_rate": 5.5546565448721e-07, "loss": 0.2548, "step": 5862 }, { "epoch": 2.592194007297387, "grad_norm": 0.2889197952114806, "learning_rate": 5.542879834411258e-07, "loss": 0.2527, "step": 5863 }, { "epoch": 2.5926362731728894, "grad_norm": 0.38383597357682725, "learning_rate": 5.531114888768674e-07, "loss": 0.2879, "step": 5864 }, { "epoch": 2.593078539048391, "grad_norm": 0.3424762858149328, "learning_rate": 5.519361711057725e-07, "loss": 0.281, "step": 5865 }, { "epoch": 2.5935208049238936, "grad_norm": 0.41593544812707134, "learning_rate": 5.507620304388711e-07, "loss": 0.2683, "step": 5866 }, { "epoch": 2.5939630707993957, "grad_norm": 0.3424759423778889, "learning_rate": 5.495890671868781e-07, "loss": 0.2566, "step": 5867 }, { "epoch": 2.594405336674898, "grad_norm": 0.3200799460586557, "learning_rate": 5.484172816601985e-07, "loss": 0.2834, "step": 5868 }, { "epoch": 2.5948476025504, "grad_norm": 0.3483588519126988, "learning_rate": 5.472466741689253e-07, "loss": 0.2688, "step": 5869 }, { "epoch": 2.595289868425902, "grad_norm": 0.393524243550273, "learning_rate": 5.460772450228397e-07, "loss": 0.2511, "step": 5870 }, { "epoch": 2.595732134301404, "grad_norm": 0.36142154068201965, "learning_rate": 5.449089945314117e-07, "loss": 0.2679, "step": 5871 }, { "epoch": 2.5961744001769063, "grad_norm": 0.36079456924217956, "learning_rate": 5.437419230037982e-07, "loss": 0.3061, "step": 5872 }, { "epoch": 2.5966166660524084, "grad_norm": 0.3479304953367216, "learning_rate": 5.425760307488448e-07, "loss": 0.2575, "step": 5873 }, { "epoch": 2.5970589319279105, "grad_norm": 0.3421084238069376, "learning_rate": 5.414113180750852e-07, "loss": 0.2303, "step": 5874 }, { "epoch": 2.597501197803413, "grad_norm": 0.28861609735634397, "learning_rate": 5.402477852907401e-07, "loss": 0.2774, "step": 5875 }, { "epoch": 2.5979434636789147, "grad_norm": 0.35400107796190683, "learning_rate": 5.390854327037198e-07, "loss": 0.2817, "step": 5876 }, { "epoch": 2.5983857295544173, "grad_norm": 0.3260362496719406, "learning_rate": 5.379242606216206e-07, "loss": 0.2928, "step": 5877 }, { "epoch": 2.5988279954299194, "grad_norm": 0.31882580349873807, "learning_rate": 5.367642693517278e-07, "loss": 0.2713, "step": 5878 }, { "epoch": 2.5992702613054215, "grad_norm": 0.2903880897272109, "learning_rate": 5.356054592010113e-07, "loss": 0.282, "step": 5879 }, { "epoch": 2.5997125271809236, "grad_norm": 0.406041792641403, "learning_rate": 5.344478304761319e-07, "loss": 0.2677, "step": 5880 }, { "epoch": 2.6001547930564257, "grad_norm": 0.3647326972772114, "learning_rate": 5.332913834834341e-07, "loss": 0.2398, "step": 5881 }, { "epoch": 2.600597058931928, "grad_norm": 0.33700036584599863, "learning_rate": 5.321361185289547e-07, "loss": 0.2656, "step": 5882 }, { "epoch": 2.60103932480743, "grad_norm": 0.32145845477810386, "learning_rate": 5.309820359184137e-07, "loss": 0.2788, "step": 5883 }, { "epoch": 2.6014815906829325, "grad_norm": 0.3656958375852359, "learning_rate": 5.298291359572195e-07, "loss": 0.26, "step": 5884 }, { "epoch": 2.601923856558434, "grad_norm": 0.33300232819254055, "learning_rate": 5.286774189504673e-07, "loss": 0.2788, "step": 5885 }, { "epoch": 2.6023661224339367, "grad_norm": 0.3566404043808171, "learning_rate": 5.275268852029392e-07, "loss": 0.2335, "step": 5886 }, { "epoch": 2.602808388309439, "grad_norm": 0.38884214472736045, "learning_rate": 5.263775350191047e-07, "loss": 0.3, "step": 5887 }, { "epoch": 2.603250654184941, "grad_norm": 0.3810470957658784, "learning_rate": 5.252293687031196e-07, "loss": 0.2884, "step": 5888 }, { "epoch": 2.603692920060443, "grad_norm": 0.39810590894516096, "learning_rate": 5.240823865588257e-07, "loss": 0.2848, "step": 5889 }, { "epoch": 2.604135185935945, "grad_norm": 0.3632304526482087, "learning_rate": 5.229365888897536e-07, "loss": 0.2327, "step": 5890 }, { "epoch": 2.6045774518114473, "grad_norm": 0.3473450424955047, "learning_rate": 5.217919759991174e-07, "loss": 0.2646, "step": 5891 }, { "epoch": 2.6050197176869494, "grad_norm": 0.3764792716587389, "learning_rate": 5.20648548189821e-07, "loss": 0.2473, "step": 5892 }, { "epoch": 2.6054619835624515, "grad_norm": 0.33364680664743024, "learning_rate": 5.19506305764454e-07, "loss": 0.2729, "step": 5893 }, { "epoch": 2.6059042494379536, "grad_norm": 0.28026406315228697, "learning_rate": 5.183652490252888e-07, "loss": 0.2891, "step": 5894 }, { "epoch": 2.606346515313456, "grad_norm": 0.31418249042806146, "learning_rate": 5.172253782742875e-07, "loss": 0.2694, "step": 5895 }, { "epoch": 2.606788781188958, "grad_norm": 0.33654239108275386, "learning_rate": 5.160866938130977e-07, "loss": 0.2556, "step": 5896 }, { "epoch": 2.6072310470644604, "grad_norm": 0.29479050011303853, "learning_rate": 5.149491959430514e-07, "loss": 0.2591, "step": 5897 }, { "epoch": 2.6076733129399625, "grad_norm": 0.3157685007721685, "learning_rate": 5.138128849651702e-07, "loss": 0.2837, "step": 5898 }, { "epoch": 2.6081155788154646, "grad_norm": 0.35580058155415367, "learning_rate": 5.126777611801587e-07, "loss": 0.2775, "step": 5899 }, { "epoch": 2.6085578446909667, "grad_norm": 0.30345173675396986, "learning_rate": 5.11543824888408e-07, "loss": 0.2669, "step": 5900 }, { "epoch": 2.609000110566469, "grad_norm": 0.3127044261489459, "learning_rate": 5.10411076389995e-07, "loss": 0.2449, "step": 5901 }, { "epoch": 2.609442376441971, "grad_norm": 0.347807396117175, "learning_rate": 5.09279515984682e-07, "loss": 0.2911, "step": 5902 }, { "epoch": 2.609884642317473, "grad_norm": 0.454897969938363, "learning_rate": 5.081491439719172e-07, "loss": 0.2733, "step": 5903 }, { "epoch": 2.610326908192975, "grad_norm": 0.3371425677671565, "learning_rate": 5.070199606508341e-07, "loss": 0.2601, "step": 5904 }, { "epoch": 2.6107691740684773, "grad_norm": 0.37630512072627, "learning_rate": 5.058919663202522e-07, "loss": 0.2465, "step": 5905 }, { "epoch": 2.61121143994398, "grad_norm": 0.3257772447882722, "learning_rate": 5.047651612786758e-07, "loss": 0.2845, "step": 5906 }, { "epoch": 2.611653705819482, "grad_norm": 0.3162045234309733, "learning_rate": 5.036395458242937e-07, "loss": 0.2604, "step": 5907 }, { "epoch": 2.612095971694984, "grad_norm": 0.3774737511598758, "learning_rate": 5.025151202549827e-07, "loss": 0.2824, "step": 5908 }, { "epoch": 2.612538237570486, "grad_norm": 0.355639225564383, "learning_rate": 5.013918848683024e-07, "loss": 0.2738, "step": 5909 }, { "epoch": 2.6129805034459883, "grad_norm": 0.3196098144277579, "learning_rate": 5.002698399614969e-07, "loss": 0.2547, "step": 5910 }, { "epoch": 2.6134227693214904, "grad_norm": 0.4173203912156896, "learning_rate": 4.991489858314968e-07, "loss": 0.2836, "step": 5911 }, { "epoch": 2.6138650351969925, "grad_norm": 0.4673463142580349, "learning_rate": 4.980293227749156e-07, "loss": 0.2782, "step": 5912 }, { "epoch": 2.6143073010724946, "grad_norm": 0.42552245711571807, "learning_rate": 4.969108510880555e-07, "loss": 0.2557, "step": 5913 }, { "epoch": 2.6147495669479968, "grad_norm": 0.327951665823192, "learning_rate": 4.957935710668992e-07, "loss": 0.2789, "step": 5914 }, { "epoch": 2.6151918328234993, "grad_norm": 0.29198619077617843, "learning_rate": 4.946774830071171e-07, "loss": 0.2459, "step": 5915 }, { "epoch": 2.615634098699001, "grad_norm": 0.3436692480735151, "learning_rate": 4.935625872040617e-07, "loss": 0.2684, "step": 5916 }, { "epoch": 2.6160763645745035, "grad_norm": 0.36529315514668886, "learning_rate": 4.924488839527719e-07, "loss": 0.2613, "step": 5917 }, { "epoch": 2.6165186304500057, "grad_norm": 0.4390842658291592, "learning_rate": 4.913363735479698e-07, "loss": 0.2653, "step": 5918 }, { "epoch": 2.6169608963255078, "grad_norm": 0.27739894585920705, "learning_rate": 4.902250562840622e-07, "loss": 0.2566, "step": 5919 }, { "epoch": 2.61740316220101, "grad_norm": 0.31518635876112133, "learning_rate": 4.891149324551403e-07, "loss": 0.2648, "step": 5920 }, { "epoch": 2.617845428076512, "grad_norm": 0.4647964608269594, "learning_rate": 4.880060023549804e-07, "loss": 0.2735, "step": 5921 }, { "epoch": 2.618287693952014, "grad_norm": 0.3674768903893479, "learning_rate": 4.868982662770406e-07, "loss": 0.2431, "step": 5922 }, { "epoch": 2.618729959827516, "grad_norm": 0.3414260559582169, "learning_rate": 4.857917245144639e-07, "loss": 0.2306, "step": 5923 }, { "epoch": 2.6191722257030183, "grad_norm": 0.328225665008469, "learning_rate": 4.846863773600807e-07, "loss": 0.2682, "step": 5924 }, { "epoch": 2.6196144915785204, "grad_norm": 0.3572312460771169, "learning_rate": 4.835822251063993e-07, "loss": 0.2819, "step": 5925 }, { "epoch": 2.620056757454023, "grad_norm": 0.32764839515233324, "learning_rate": 4.824792680456158e-07, "loss": 0.25, "step": 5926 }, { "epoch": 2.6204990233295247, "grad_norm": 0.40039324850309, "learning_rate": 4.813775064696091e-07, "loss": 0.2914, "step": 5927 }, { "epoch": 2.6209412892050272, "grad_norm": 0.37045320612433835, "learning_rate": 4.802769406699398e-07, "loss": 0.2491, "step": 5928 }, { "epoch": 2.6213835550805293, "grad_norm": 0.2888791982220122, "learning_rate": 4.791775709378571e-07, "loss": 0.2669, "step": 5929 }, { "epoch": 2.6218258209560314, "grad_norm": 0.3497999571376568, "learning_rate": 4.780793975642883e-07, "loss": 0.2498, "step": 5930 }, { "epoch": 2.6222680868315336, "grad_norm": 0.32029396419705614, "learning_rate": 4.769824208398466e-07, "loss": 0.2547, "step": 5931 }, { "epoch": 2.6227103527070357, "grad_norm": 0.29390023429080475, "learning_rate": 4.758866410548285e-07, "loss": 0.2593, "step": 5932 }, { "epoch": 2.623152618582538, "grad_norm": 0.4107359473570294, "learning_rate": 4.747920584992144e-07, "loss": 0.2799, "step": 5933 }, { "epoch": 2.62359488445804, "grad_norm": 0.30420950732064644, "learning_rate": 4.7369867346266365e-07, "loss": 0.2437, "step": 5934 }, { "epoch": 2.6240371503335425, "grad_norm": 0.29360648936538836, "learning_rate": 4.7260648623452474e-07, "loss": 0.2544, "step": 5935 }, { "epoch": 2.624479416209044, "grad_norm": 0.31025759775042616, "learning_rate": 4.715154971038255e-07, "loss": 0.2402, "step": 5936 }, { "epoch": 2.6249216820845467, "grad_norm": 0.43841768348973337, "learning_rate": 4.704257063592782e-07, "loss": 0.2704, "step": 5937 }, { "epoch": 2.625363947960049, "grad_norm": 0.341547971713102, "learning_rate": 4.6933711428927663e-07, "loss": 0.2684, "step": 5938 }, { "epoch": 2.625806213835551, "grad_norm": 0.43611353640849576, "learning_rate": 4.682497211818976e-07, "loss": 0.2746, "step": 5939 }, { "epoch": 2.626248479711053, "grad_norm": 0.30972483268130324, "learning_rate": 4.6716352732490367e-07, "loss": 0.2792, "step": 5940 }, { "epoch": 2.626690745586555, "grad_norm": 0.413326174992317, "learning_rate": 4.6607853300573433e-07, "loss": 0.2613, "step": 5941 }, { "epoch": 2.6271330114620572, "grad_norm": 0.41127400055990737, "learning_rate": 4.6499473851151723e-07, "loss": 0.2768, "step": 5942 }, { "epoch": 2.6275752773375594, "grad_norm": 0.3611231278385071, "learning_rate": 4.6391214412905836e-07, "loss": 0.2609, "step": 5943 }, { "epoch": 2.6280175432130615, "grad_norm": 0.34154758451246253, "learning_rate": 4.6283075014484755e-07, "loss": 0.282, "step": 5944 }, { "epoch": 2.6284598090885636, "grad_norm": 0.32443137210759637, "learning_rate": 4.6175055684505975e-07, "loss": 0.2523, "step": 5945 }, { "epoch": 2.628902074964066, "grad_norm": 0.30852911039911046, "learning_rate": 4.6067156451554794e-07, "loss": 0.2588, "step": 5946 }, { "epoch": 2.629344340839568, "grad_norm": 0.3905902612170922, "learning_rate": 4.5959377344184984e-07, "loss": 0.2859, "step": 5947 }, { "epoch": 2.6297866067150704, "grad_norm": 0.3595549788188878, "learning_rate": 4.585171839091834e-07, "loss": 0.2465, "step": 5948 }, { "epoch": 2.6302288725905725, "grad_norm": 0.4005999383004607, "learning_rate": 4.574417962024502e-07, "loss": 0.2187, "step": 5949 }, { "epoch": 2.6306711384660746, "grad_norm": 0.403637467660229, "learning_rate": 4.5636761060623314e-07, "loss": 0.2677, "step": 5950 }, { "epoch": 2.6311134043415767, "grad_norm": 0.34196110411936226, "learning_rate": 4.5529462740479757e-07, "loss": 0.2277, "step": 5951 }, { "epoch": 2.631555670217079, "grad_norm": 0.3545609471627134, "learning_rate": 4.5422284688208974e-07, "loss": 0.2614, "step": 5952 }, { "epoch": 2.631997936092581, "grad_norm": 0.31593617129774626, "learning_rate": 4.531522693217383e-07, "loss": 0.263, "step": 5953 }, { "epoch": 2.632440201968083, "grad_norm": 0.3221118363094815, "learning_rate": 4.5208289500705327e-07, "loss": 0.2543, "step": 5954 }, { "epoch": 2.632882467843585, "grad_norm": 0.41408435470923544, "learning_rate": 4.510147242210261e-07, "loss": 0.2713, "step": 5955 }, { "epoch": 2.6333247337190873, "grad_norm": 0.3663411517953394, "learning_rate": 4.499477572463301e-07, "loss": 0.2809, "step": 5956 }, { "epoch": 2.63376699959459, "grad_norm": 0.5041225637793244, "learning_rate": 4.4888199436532e-07, "loss": 0.2957, "step": 5957 }, { "epoch": 2.634209265470092, "grad_norm": 0.3079721545179338, "learning_rate": 4.478174358600318e-07, "loss": 0.2777, "step": 5958 }, { "epoch": 2.634651531345594, "grad_norm": 0.28130800911444936, "learning_rate": 4.467540820121824e-07, "loss": 0.2485, "step": 5959 }, { "epoch": 2.635093797221096, "grad_norm": 0.37250163745220566, "learning_rate": 4.456919331031695e-07, "loss": 0.285, "step": 5960 }, { "epoch": 2.6355360630965983, "grad_norm": 0.28722964539830115, "learning_rate": 4.4463098941407435e-07, "loss": 0.2588, "step": 5961 }, { "epoch": 2.6359783289721004, "grad_norm": 0.33347112053581146, "learning_rate": 4.4357125122565734e-07, "loss": 0.2745, "step": 5962 }, { "epoch": 2.6364205948476025, "grad_norm": 0.2878187481765559, "learning_rate": 4.425127188183592e-07, "loss": 0.2649, "step": 5963 }, { "epoch": 2.6368628607231046, "grad_norm": 0.38678690315764735, "learning_rate": 4.414553924723042e-07, "loss": 0.2746, "step": 5964 }, { "epoch": 2.6373051265986067, "grad_norm": 0.28938775395163957, "learning_rate": 4.4039927246729284e-07, "loss": 0.2618, "step": 5965 }, { "epoch": 2.6377473924741093, "grad_norm": 0.5958846381397478, "learning_rate": 4.3934435908281113e-07, "loss": 0.2837, "step": 5966 }, { "epoch": 2.638189658349611, "grad_norm": 0.43453058264211775, "learning_rate": 4.382906525980241e-07, "loss": 0.2394, "step": 5967 }, { "epoch": 2.6386319242251135, "grad_norm": 0.2800351721228033, "learning_rate": 4.372381532917763e-07, "loss": 0.2742, "step": 5968 }, { "epoch": 2.6390741901006156, "grad_norm": 0.2768678048614319, "learning_rate": 4.36186861442594e-07, "loss": 0.2945, "step": 5969 }, { "epoch": 2.6395164559761177, "grad_norm": 0.36023282811211277, "learning_rate": 4.3513677732868407e-07, "loss": 0.2513, "step": 5970 }, { "epoch": 2.63995872185162, "grad_norm": 0.33467250892567607, "learning_rate": 4.3408790122793307e-07, "loss": 0.2651, "step": 5971 }, { "epoch": 2.640400987727122, "grad_norm": 0.30754327213908733, "learning_rate": 4.3304023341790844e-07, "loss": 0.2513, "step": 5972 }, { "epoch": 2.640843253602624, "grad_norm": 0.3353947461405004, "learning_rate": 4.3199377417585677e-07, "loss": 0.2714, "step": 5973 }, { "epoch": 2.641285519478126, "grad_norm": 0.3116278611170824, "learning_rate": 4.309485237787064e-07, "loss": 0.2915, "step": 5974 }, { "epoch": 2.6417277853536283, "grad_norm": 0.4280029538829324, "learning_rate": 4.299044825030646e-07, "loss": 0.3045, "step": 5975 }, { "epoch": 2.6421700512291304, "grad_norm": 0.30237168891089855, "learning_rate": 4.2886165062521857e-07, "loss": 0.2728, "step": 5976 }, { "epoch": 2.642612317104633, "grad_norm": 0.3255138656083672, "learning_rate": 4.2782002842113755e-07, "loss": 0.2501, "step": 5977 }, { "epoch": 2.6430545829801346, "grad_norm": 0.33485075235570344, "learning_rate": 4.267796161664678e-07, "loss": 0.2704, "step": 5978 }, { "epoch": 2.643496848855637, "grad_norm": 0.3933286956990757, "learning_rate": 4.2574041413653734e-07, "loss": 0.2796, "step": 5979 }, { "epoch": 2.6439391147311393, "grad_norm": 0.3348198754572881, "learning_rate": 4.2470242260635397e-07, "loss": 0.2728, "step": 5980 }, { "epoch": 2.6443813806066414, "grad_norm": 0.30821898429740274, "learning_rate": 4.2366564185060134e-07, "loss": 0.2746, "step": 5981 }, { "epoch": 2.6448236464821435, "grad_norm": 0.3363779534613407, "learning_rate": 4.226300721436488e-07, "loss": 0.2864, "step": 5982 }, { "epoch": 2.6452659123576456, "grad_norm": 0.3562764501159435, "learning_rate": 4.2159571375954167e-07, "loss": 0.2684, "step": 5983 }, { "epoch": 2.6457081782331477, "grad_norm": 0.36537100929653643, "learning_rate": 4.205625669720048e-07, "loss": 0.2783, "step": 5984 }, { "epoch": 2.64615044410865, "grad_norm": 0.372347521539284, "learning_rate": 4.195306320544423e-07, "loss": 0.2719, "step": 5985 }, { "epoch": 2.6465927099841524, "grad_norm": 0.36626157971378615, "learning_rate": 4.1849990927993966e-07, "loss": 0.2356, "step": 5986 }, { "epoch": 2.647034975859654, "grad_norm": 0.33050007638739315, "learning_rate": 4.174703989212586e-07, "loss": 0.2814, "step": 5987 }, { "epoch": 2.6474772417351566, "grad_norm": 0.33217084199191255, "learning_rate": 4.1644210125084186e-07, "loss": 0.2501, "step": 5988 }, { "epoch": 2.6479195076106588, "grad_norm": 0.35530442248881056, "learning_rate": 4.154150165408116e-07, "loss": 0.26, "step": 5989 }, { "epoch": 2.648361773486161, "grad_norm": 0.3422543521868326, "learning_rate": 4.1438914506296723e-07, "loss": 0.2739, "step": 5990 }, { "epoch": 2.648804039361663, "grad_norm": 0.36642357825945376, "learning_rate": 4.1336448708878876e-07, "loss": 0.256, "step": 5991 }, { "epoch": 2.649246305237165, "grad_norm": 0.36401500549611476, "learning_rate": 4.1234104288943364e-07, "loss": 0.2723, "step": 5992 }, { "epoch": 2.649688571112667, "grad_norm": 0.3681446243216617, "learning_rate": 4.1131881273574037e-07, "loss": 0.2914, "step": 5993 }, { "epoch": 2.6501308369881693, "grad_norm": 0.4255050012450791, "learning_rate": 4.1029779689822414e-07, "loss": 0.2672, "step": 5994 }, { "epoch": 2.6505731028636714, "grad_norm": 0.3419632072607582, "learning_rate": 4.092779956470794e-07, "loss": 0.2998, "step": 5995 }, { "epoch": 2.6510153687391735, "grad_norm": 0.3794504449973385, "learning_rate": 4.0825940925217855e-07, "loss": 0.254, "step": 5996 }, { "epoch": 2.651457634614676, "grad_norm": 0.33978464659711605, "learning_rate": 4.0724203798307217e-07, "loss": 0.2487, "step": 5997 }, { "epoch": 2.6518999004901778, "grad_norm": 0.2614602259251213, "learning_rate": 4.062258821089926e-07, "loss": 0.2376, "step": 5998 }, { "epoch": 2.6523421663656803, "grad_norm": 0.36483850591138706, "learning_rate": 4.05210941898847e-07, "loss": 0.2697, "step": 5999 }, { "epoch": 2.6527844322411824, "grad_norm": 0.31714797586528204, "learning_rate": 4.041972176212222e-07, "loss": 0.2587, "step": 6000 }, { "epoch": 2.6532266981166845, "grad_norm": 0.33400922894667284, "learning_rate": 4.031847095443825e-07, "loss": 0.274, "step": 6001 }, { "epoch": 2.6536689639921867, "grad_norm": 0.29149749884218906, "learning_rate": 4.0217341793627127e-07, "loss": 0.2575, "step": 6002 }, { "epoch": 2.6541112298676888, "grad_norm": 0.3147971352686926, "learning_rate": 4.0116334306450953e-07, "loss": 0.2757, "step": 6003 }, { "epoch": 2.654553495743191, "grad_norm": 0.42796163423518047, "learning_rate": 4.001544851963968e-07, "loss": 0.3009, "step": 6004 }, { "epoch": 2.654995761618693, "grad_norm": 0.3479397445304035, "learning_rate": 3.9914684459890884e-07, "loss": 0.2754, "step": 6005 }, { "epoch": 2.655438027494195, "grad_norm": 0.3903978047936694, "learning_rate": 3.981404215387019e-07, "loss": 0.3033, "step": 6006 }, { "epoch": 2.655880293369697, "grad_norm": 0.29298005345972167, "learning_rate": 3.971352162821085e-07, "loss": 0.2523, "step": 6007 }, { "epoch": 2.6563225592451998, "grad_norm": 0.3031564862406429, "learning_rate": 3.961312290951369e-07, "loss": 0.2912, "step": 6008 }, { "epoch": 2.656764825120702, "grad_norm": 0.34289904630950235, "learning_rate": 3.951284602434785e-07, "loss": 0.2727, "step": 6009 }, { "epoch": 2.657207090996204, "grad_norm": 0.33583097554959973, "learning_rate": 3.9412690999249715e-07, "loss": 0.2686, "step": 6010 }, { "epoch": 2.657649356871706, "grad_norm": 0.2995119259857668, "learning_rate": 3.9312657860723744e-07, "loss": 0.2468, "step": 6011 }, { "epoch": 2.6580916227472082, "grad_norm": 0.3211077140732192, "learning_rate": 3.921274663524183e-07, "loss": 0.2892, "step": 6012 }, { "epoch": 2.6585338886227103, "grad_norm": 0.31816803397533205, "learning_rate": 3.9112957349243697e-07, "loss": 0.2753, "step": 6013 }, { "epoch": 2.6589761544982125, "grad_norm": 0.32153401522380703, "learning_rate": 3.9013290029137065e-07, "loss": 0.2649, "step": 6014 }, { "epoch": 2.6594184203737146, "grad_norm": 0.3690886215628098, "learning_rate": 3.8913744701297216e-07, "loss": 0.2698, "step": 6015 }, { "epoch": 2.6598606862492167, "grad_norm": 0.3547500514704673, "learning_rate": 3.881432139206698e-07, "loss": 0.2525, "step": 6016 }, { "epoch": 2.6603029521247192, "grad_norm": 0.40642527435500386, "learning_rate": 3.871502012775713e-07, "loss": 0.2399, "step": 6017 }, { "epoch": 2.660745218000221, "grad_norm": 0.3809978681396917, "learning_rate": 3.861584093464604e-07, "loss": 0.2738, "step": 6018 }, { "epoch": 2.6611874838757235, "grad_norm": 0.32228161986738607, "learning_rate": 3.851678383897972e-07, "loss": 0.2742, "step": 6019 }, { "epoch": 2.6616297497512256, "grad_norm": 0.39048279807013647, "learning_rate": 3.8417848866972027e-07, "loss": 0.2748, "step": 6020 }, { "epoch": 2.6620720156267277, "grad_norm": 0.42548944609955036, "learning_rate": 3.8319036044804413e-07, "loss": 0.295, "step": 6021 }, { "epoch": 2.66251428150223, "grad_norm": 0.34540736853449255, "learning_rate": 3.822034539862596e-07, "loss": 0.2656, "step": 6022 }, { "epoch": 2.662956547377732, "grad_norm": 0.5116921955825198, "learning_rate": 3.812177695455338e-07, "loss": 0.2859, "step": 6023 }, { "epoch": 2.663398813253234, "grad_norm": 0.42792945657531656, "learning_rate": 3.802333073867126e-07, "loss": 0.2813, "step": 6024 }, { "epoch": 2.663841079128736, "grad_norm": 0.3062652008511477, "learning_rate": 3.792500677703176e-07, "loss": 0.2485, "step": 6025 }, { "epoch": 2.6642833450042382, "grad_norm": 0.34852259759394694, "learning_rate": 3.782680509565462e-07, "loss": 0.2564, "step": 6026 }, { "epoch": 2.6647256108797404, "grad_norm": 0.3366670175507866, "learning_rate": 3.772872572052705e-07, "loss": 0.2432, "step": 6027 }, { "epoch": 2.665167876755243, "grad_norm": 0.35032213644258126, "learning_rate": 3.7630768677604236e-07, "loss": 0.2558, "step": 6028 }, { "epoch": 2.6656101426307446, "grad_norm": 0.3482868174745312, "learning_rate": 3.7532933992808717e-07, "loss": 0.2326, "step": 6029 }, { "epoch": 2.666052408506247, "grad_norm": 0.32771535354025094, "learning_rate": 3.7435221692030956e-07, "loss": 0.2563, "step": 6030 }, { "epoch": 2.6664946743817493, "grad_norm": 0.32348250412856516, "learning_rate": 3.733763180112876e-07, "loss": 0.2687, "step": 6031 }, { "epoch": 2.6669369402572514, "grad_norm": 0.2835012132995, "learning_rate": 3.724016434592764e-07, "loss": 0.2583, "step": 6032 }, { "epoch": 2.6673792061327535, "grad_norm": 0.3484142987654154, "learning_rate": 3.714281935222064e-07, "loss": 0.2615, "step": 6033 }, { "epoch": 2.6678214720082556, "grad_norm": 0.3511776040969961, "learning_rate": 3.704559684576853e-07, "loss": 0.2845, "step": 6034 }, { "epoch": 2.6682637378837577, "grad_norm": 0.4527065573311213, "learning_rate": 3.694849685229951e-07, "loss": 0.2788, "step": 6035 }, { "epoch": 2.66870600375926, "grad_norm": 0.35678705082730017, "learning_rate": 3.685151939750953e-07, "loss": 0.2971, "step": 6036 }, { "epoch": 2.6691482696347624, "grad_norm": 0.35959562664660616, "learning_rate": 3.6754664507061945e-07, "loss": 0.2642, "step": 6037 }, { "epoch": 2.669590535510264, "grad_norm": 0.3389707842066638, "learning_rate": 3.6657932206587753e-07, "loss": 0.2701, "step": 6038 }, { "epoch": 2.6700328013857666, "grad_norm": 0.3761840816132344, "learning_rate": 3.6561322521685425e-07, "loss": 0.3135, "step": 6039 }, { "epoch": 2.6704750672612687, "grad_norm": 0.37911157174836985, "learning_rate": 3.6464835477921233e-07, "loss": 0.2667, "step": 6040 }, { "epoch": 2.670917333136771, "grad_norm": 0.3372596312984688, "learning_rate": 3.636847110082881e-07, "loss": 0.2313, "step": 6041 }, { "epoch": 2.671359599012273, "grad_norm": 0.33873300843949283, "learning_rate": 3.6272229415909356e-07, "loss": 0.2372, "step": 6042 }, { "epoch": 2.671801864887775, "grad_norm": 0.3365932748057223, "learning_rate": 3.61761104486314e-07, "loss": 0.2817, "step": 6043 }, { "epoch": 2.672244130763277, "grad_norm": 0.6328786394979511, "learning_rate": 3.608011422443136e-07, "loss": 0.3253, "step": 6044 }, { "epoch": 2.6726863966387793, "grad_norm": 0.5964299955118236, "learning_rate": 3.5984240768712876e-07, "loss": 0.2764, "step": 6045 }, { "epoch": 2.6731286625142814, "grad_norm": 0.2980014745110206, "learning_rate": 3.5888490106847307e-07, "loss": 0.258, "step": 6046 }, { "epoch": 2.6735709283897835, "grad_norm": 0.3153666038941953, "learning_rate": 3.57928622641735e-07, "loss": 0.2583, "step": 6047 }, { "epoch": 2.674013194265286, "grad_norm": 0.356384198896269, "learning_rate": 3.5697357265997603e-07, "loss": 0.2818, "step": 6048 }, { "epoch": 2.6744554601407877, "grad_norm": 0.3033065699760108, "learning_rate": 3.560197513759345e-07, "loss": 0.2922, "step": 6049 }, { "epoch": 2.6748977260162903, "grad_norm": 0.3155635512235249, "learning_rate": 3.550671590420229e-07, "loss": 0.2698, "step": 6050 }, { "epoch": 2.6753399918917924, "grad_norm": 0.33207752872309926, "learning_rate": 3.5411579591032806e-07, "loss": 0.2809, "step": 6051 }, { "epoch": 2.6757822577672945, "grad_norm": 0.35011228815734313, "learning_rate": 3.531656622326124e-07, "loss": 0.2853, "step": 6052 }, { "epoch": 2.6762245236427966, "grad_norm": 0.5287042794588545, "learning_rate": 3.522167582603131e-07, "loss": 0.2685, "step": 6053 }, { "epoch": 2.6766667895182987, "grad_norm": 0.33364854455647797, "learning_rate": 3.5126908424454043e-07, "loss": 0.2473, "step": 6054 }, { "epoch": 2.677109055393801, "grad_norm": 0.347794941453931, "learning_rate": 3.503226404360793e-07, "loss": 0.2753, "step": 6055 }, { "epoch": 2.677551321269303, "grad_norm": 0.3718538788501584, "learning_rate": 3.493774270853928e-07, "loss": 0.2583, "step": 6056 }, { "epoch": 2.677993587144805, "grad_norm": 0.3535317073216845, "learning_rate": 3.4843344444261405e-07, "loss": 0.2444, "step": 6057 }, { "epoch": 2.678435853020307, "grad_norm": 0.3221163723989363, "learning_rate": 3.4749069275755166e-07, "loss": 0.2795, "step": 6058 }, { "epoch": 2.6788781188958097, "grad_norm": 0.34742732081839756, "learning_rate": 3.465491722796882e-07, "loss": 0.2643, "step": 6059 }, { "epoch": 2.679320384771312, "grad_norm": 0.3496946720509187, "learning_rate": 3.456088832581811e-07, "loss": 0.254, "step": 6060 }, { "epoch": 2.679762650646814, "grad_norm": 0.3499971363699362, "learning_rate": 3.446698259418635e-07, "loss": 0.2513, "step": 6061 }, { "epoch": 2.680204916522316, "grad_norm": 0.40379783635241506, "learning_rate": 3.437320005792394e-07, "loss": 0.2429, "step": 6062 }, { "epoch": 2.680647182397818, "grad_norm": 0.2991874728554719, "learning_rate": 3.4279540741848916e-07, "loss": 0.2649, "step": 6063 }, { "epoch": 2.6810894482733203, "grad_norm": 0.44580449276932005, "learning_rate": 3.418600467074651e-07, "loss": 0.283, "step": 6064 }, { "epoch": 2.6815317141488224, "grad_norm": 0.3637603116925985, "learning_rate": 3.409259186936953e-07, "loss": 0.2484, "step": 6065 }, { "epoch": 2.6819739800243245, "grad_norm": 0.3900271664551135, "learning_rate": 3.3999302362438094e-07, "loss": 0.2784, "step": 6066 }, { "epoch": 2.6824162458998266, "grad_norm": 0.4375213026510722, "learning_rate": 3.3906136174639615e-07, "loss": 0.2591, "step": 6067 }, { "epoch": 2.682858511775329, "grad_norm": 0.3163090475163727, "learning_rate": 3.381309333062893e-07, "loss": 0.241, "step": 6068 }, { "epoch": 2.683300777650831, "grad_norm": 0.3778004501030964, "learning_rate": 3.3720173855028293e-07, "loss": 0.257, "step": 6069 }, { "epoch": 2.6837430435263334, "grad_norm": 0.3866214097386847, "learning_rate": 3.362737777242725e-07, "loss": 0.268, "step": 6070 }, { "epoch": 2.6841853094018355, "grad_norm": 0.34156042900038314, "learning_rate": 3.3534705107382606e-07, "loss": 0.3063, "step": 6071 }, { "epoch": 2.6846275752773376, "grad_norm": 0.37158896502869004, "learning_rate": 3.344215588441874e-07, "loss": 0.2472, "step": 6072 }, { "epoch": 2.6850698411528398, "grad_norm": 0.40715454997058614, "learning_rate": 3.334973012802728e-07, "loss": 0.3175, "step": 6073 }, { "epoch": 2.685512107028342, "grad_norm": 0.3149154379965672, "learning_rate": 3.3257427862666894e-07, "loss": 0.2444, "step": 6074 }, { "epoch": 2.685954372903844, "grad_norm": 0.4278532289932696, "learning_rate": 3.316524911276392e-07, "loss": 0.3213, "step": 6075 }, { "epoch": 2.686396638779346, "grad_norm": 0.38474240831169465, "learning_rate": 3.3073193902711854e-07, "loss": 0.2781, "step": 6076 }, { "epoch": 2.686838904654848, "grad_norm": 0.31538295589625165, "learning_rate": 3.298126225687159e-07, "loss": 0.2632, "step": 6077 }, { "epoch": 2.6872811705303503, "grad_norm": 0.3162738811667121, "learning_rate": 3.288945419957129e-07, "loss": 0.2747, "step": 6078 }, { "epoch": 2.687723436405853, "grad_norm": 0.3074987739789955, "learning_rate": 3.2797769755106345e-07, "loss": 0.2815, "step": 6079 }, { "epoch": 2.6881657022813545, "grad_norm": 0.5647123204016855, "learning_rate": 3.2706208947739457e-07, "loss": 0.2414, "step": 6080 }, { "epoch": 2.688607968156857, "grad_norm": 0.3476112672114892, "learning_rate": 3.2614771801700697e-07, "loss": 0.2576, "step": 6081 }, { "epoch": 2.689050234032359, "grad_norm": 0.3343141547026732, "learning_rate": 3.2523458341187306e-07, "loss": 0.2748, "step": 6082 }, { "epoch": 2.6894924999078613, "grad_norm": 0.38991666092995725, "learning_rate": 3.24322685903638e-07, "loss": 0.2835, "step": 6083 }, { "epoch": 2.6899347657833634, "grad_norm": 0.3452310915555471, "learning_rate": 3.234120257336204e-07, "loss": 0.2639, "step": 6084 }, { "epoch": 2.6903770316588655, "grad_norm": 0.3172079128150255, "learning_rate": 3.2250260314281134e-07, "loss": 0.263, "step": 6085 }, { "epoch": 2.6908192975343677, "grad_norm": 0.3454743246335968, "learning_rate": 3.2159441837187276e-07, "loss": 0.2704, "step": 6086 }, { "epoch": 2.6912615634098698, "grad_norm": 0.32140942300094205, "learning_rate": 3.206874716611402e-07, "loss": 0.2714, "step": 6087 }, { "epoch": 2.6917038292853723, "grad_norm": 0.29134993994166464, "learning_rate": 3.197817632506245e-07, "loss": 0.2763, "step": 6088 }, { "epoch": 2.692146095160874, "grad_norm": 0.28373146491596274, "learning_rate": 3.1887729338000215e-07, "loss": 0.2381, "step": 6089 }, { "epoch": 2.6925883610363766, "grad_norm": 0.3438542674028451, "learning_rate": 3.179740622886279e-07, "loss": 0.2705, "step": 6090 }, { "epoch": 2.6930306269118787, "grad_norm": 0.26529033558659615, "learning_rate": 3.170720702155261e-07, "loss": 0.2562, "step": 6091 }, { "epoch": 2.693472892787381, "grad_norm": 0.2773977611139816, "learning_rate": 3.161713173993919e-07, "loss": 0.2833, "step": 6092 }, { "epoch": 2.693915158662883, "grad_norm": 0.3656277839231638, "learning_rate": 3.152718040785968e-07, "loss": 0.2662, "step": 6093 }, { "epoch": 2.694357424538385, "grad_norm": 0.366296026425634, "learning_rate": 3.1437353049118045e-07, "loss": 0.2551, "step": 6094 }, { "epoch": 2.694799690413887, "grad_norm": 0.3202959011320936, "learning_rate": 3.134764968748555e-07, "loss": 0.2953, "step": 6095 }, { "epoch": 2.6952419562893892, "grad_norm": 0.2860073564538517, "learning_rate": 3.12580703467007e-07, "loss": 0.279, "step": 6096 }, { "epoch": 2.6956842221648913, "grad_norm": 0.34176236071767224, "learning_rate": 3.116861505046909e-07, "loss": 0.2593, "step": 6097 }, { "epoch": 2.6961264880403935, "grad_norm": 0.3043249350695795, "learning_rate": 3.107928382246356e-07, "loss": 0.2443, "step": 6098 }, { "epoch": 2.696568753915896, "grad_norm": 0.31010872724803457, "learning_rate": 3.099007668632409e-07, "loss": 0.2812, "step": 6099 }, { "epoch": 2.6970110197913977, "grad_norm": 0.2596867982883875, "learning_rate": 3.09009936656578e-07, "loss": 0.2578, "step": 6100 }, { "epoch": 2.6974532856669002, "grad_norm": 0.2925436794589028, "learning_rate": 3.081203478403905e-07, "loss": 0.2584, "step": 6101 }, { "epoch": 2.6978955515424023, "grad_norm": 0.3010942456451035, "learning_rate": 3.0723200065009284e-07, "loss": 0.2828, "step": 6102 }, { "epoch": 2.6983378174179045, "grad_norm": 0.3651177361334821, "learning_rate": 3.0634489532076984e-07, "loss": 0.2668, "step": 6103 }, { "epoch": 2.6987800832934066, "grad_norm": 0.2811230850515277, "learning_rate": 3.054590320871809e-07, "loss": 0.2542, "step": 6104 }, { "epoch": 2.6992223491689087, "grad_norm": 0.35537798641186497, "learning_rate": 3.045744111837529e-07, "loss": 0.2792, "step": 6105 }, { "epoch": 2.699664615044411, "grad_norm": 0.3445321502419141, "learning_rate": 3.036910328445858e-07, "loss": 0.2788, "step": 6106 }, { "epoch": 2.700106880919913, "grad_norm": 0.4677049542928385, "learning_rate": 3.0280889730345153e-07, "loss": 0.3148, "step": 6107 }, { "epoch": 2.700549146795415, "grad_norm": 0.3359712191944673, "learning_rate": 3.019280047937906e-07, "loss": 0.2512, "step": 6108 }, { "epoch": 2.700991412670917, "grad_norm": 0.28746480099348865, "learning_rate": 3.010483555487176e-07, "loss": 0.2703, "step": 6109 }, { "epoch": 2.7014336785464197, "grad_norm": 0.4376492743384325, "learning_rate": 3.0016994980101634e-07, "loss": 0.2924, "step": 6110 }, { "epoch": 2.701875944421922, "grad_norm": 0.4175417821121843, "learning_rate": 2.9929278778314255e-07, "loss": 0.2543, "step": 6111 }, { "epoch": 2.702318210297424, "grad_norm": 0.447418695035736, "learning_rate": 2.9841686972722106e-07, "loss": 0.2724, "step": 6112 }, { "epoch": 2.702760476172926, "grad_norm": 0.34587906542554037, "learning_rate": 2.9754219586504873e-07, "loss": 0.2792, "step": 6113 }, { "epoch": 2.703202742048428, "grad_norm": 0.4449407939346553, "learning_rate": 2.9666876642809426e-07, "loss": 0.2792, "step": 6114 }, { "epoch": 2.7036450079239303, "grad_norm": 0.3995432729240275, "learning_rate": 2.9579658164749446e-07, "loss": 0.2648, "step": 6115 }, { "epoch": 2.7040872737994324, "grad_norm": 0.3263819228119533, "learning_rate": 2.9492564175405856e-07, "loss": 0.2559, "step": 6116 }, { "epoch": 2.7045295396749345, "grad_norm": 0.353708158197794, "learning_rate": 2.9405594697826666e-07, "loss": 0.2449, "step": 6117 }, { "epoch": 2.7049718055504366, "grad_norm": 0.3567506651146154, "learning_rate": 2.9318749755026746e-07, "loss": 0.2697, "step": 6118 }, { "epoch": 2.705414071425939, "grad_norm": 0.31068086330596845, "learning_rate": 2.923202936998826e-07, "loss": 0.276, "step": 6119 }, { "epoch": 2.705856337301441, "grad_norm": 0.29957666190856325, "learning_rate": 2.914543356566013e-07, "loss": 0.2263, "step": 6120 }, { "epoch": 2.7062986031769434, "grad_norm": 0.434461054782563, "learning_rate": 2.905896236495859e-07, "loss": 0.2749, "step": 6121 }, { "epoch": 2.7067408690524455, "grad_norm": 0.38888279190023517, "learning_rate": 2.8972615790766714e-07, "loss": 0.2781, "step": 6122 }, { "epoch": 2.7071831349279476, "grad_norm": 0.384013468225612, "learning_rate": 2.8886393865934617e-07, "loss": 0.2827, "step": 6123 }, { "epoch": 2.7076254008034497, "grad_norm": 0.5389468656874551, "learning_rate": 2.8800296613279433e-07, "loss": 0.2692, "step": 6124 }, { "epoch": 2.708067666678952, "grad_norm": 0.44982669435249795, "learning_rate": 2.8714324055585443e-07, "loss": 0.2652, "step": 6125 }, { "epoch": 2.708509932554454, "grad_norm": 0.33420467043974167, "learning_rate": 2.862847621560377e-07, "loss": 0.2789, "step": 6126 }, { "epoch": 2.708952198429956, "grad_norm": 0.3578849643941665, "learning_rate": 2.854275311605259e-07, "loss": 0.2818, "step": 6127 }, { "epoch": 2.709394464305458, "grad_norm": 0.3355390549643105, "learning_rate": 2.8457154779617133e-07, "loss": 0.2638, "step": 6128 }, { "epoch": 2.7098367301809603, "grad_norm": 0.36270350033958854, "learning_rate": 2.8371681228949286e-07, "loss": 0.2845, "step": 6129 }, { "epoch": 2.710278996056463, "grad_norm": 0.3725440742255514, "learning_rate": 2.8286332486668397e-07, "loss": 0.2525, "step": 6130 }, { "epoch": 2.7107212619319645, "grad_norm": 0.323568672247368, "learning_rate": 2.8201108575360457e-07, "loss": 0.2804, "step": 6131 }, { "epoch": 2.711163527807467, "grad_norm": 0.29570640611779847, "learning_rate": 2.811600951757859e-07, "loss": 0.2835, "step": 6132 }, { "epoch": 2.711605793682969, "grad_norm": 0.3937673926878258, "learning_rate": 2.803103533584278e-07, "loss": 0.2726, "step": 6133 }, { "epoch": 2.7120480595584713, "grad_norm": 0.39813005638411136, "learning_rate": 2.7946186052639933e-07, "loss": 0.2457, "step": 6134 }, { "epoch": 2.7124903254339734, "grad_norm": 0.33292563691620747, "learning_rate": 2.786146169042403e-07, "loss": 0.2323, "step": 6135 }, { "epoch": 2.7129325913094755, "grad_norm": 0.2876356168180536, "learning_rate": 2.777686227161591e-07, "loss": 0.2659, "step": 6136 }, { "epoch": 2.7133748571849776, "grad_norm": 0.43491223631832193, "learning_rate": 2.7692387818603396e-07, "loss": 0.3046, "step": 6137 }, { "epoch": 2.7138171230604797, "grad_norm": 0.304747134580932, "learning_rate": 2.760803835374115e-07, "loss": 0.2614, "step": 6138 }, { "epoch": 2.7142593889359823, "grad_norm": 1.4441276306311215, "learning_rate": 2.752381389935083e-07, "loss": 0.2537, "step": 6139 }, { "epoch": 2.714701654811484, "grad_norm": 0.38159924162117653, "learning_rate": 2.743971447772098e-07, "loss": 0.2747, "step": 6140 }, { "epoch": 2.7151439206869865, "grad_norm": 0.2704570202820522, "learning_rate": 2.7355740111107145e-07, "loss": 0.2287, "step": 6141 }, { "epoch": 2.7155861865624886, "grad_norm": 0.3319286095405196, "learning_rate": 2.727189082173171e-07, "loss": 0.2233, "step": 6142 }, { "epoch": 2.7160284524379907, "grad_norm": 0.7178838084999236, "learning_rate": 2.718816663178397e-07, "loss": 0.2938, "step": 6143 }, { "epoch": 2.716470718313493, "grad_norm": 0.4108991317848428, "learning_rate": 2.710456756342006e-07, "loss": 0.2831, "step": 6144 }, { "epoch": 2.716912984188995, "grad_norm": 0.4266621517105274, "learning_rate": 2.702109363876293e-07, "loss": 0.2943, "step": 6145 }, { "epoch": 2.717355250064497, "grad_norm": 0.366823742768942, "learning_rate": 2.693774487990275e-07, "loss": 0.2765, "step": 6146 }, { "epoch": 2.717797515939999, "grad_norm": 0.3457039225916395, "learning_rate": 2.685452130889621e-07, "loss": 0.2829, "step": 6147 }, { "epoch": 2.7182397818155013, "grad_norm": 0.32634019682468945, "learning_rate": 2.677142294776708e-07, "loss": 0.268, "step": 6148 }, { "epoch": 2.7186820476910034, "grad_norm": 0.31137014388552187, "learning_rate": 2.668844981850588e-07, "loss": 0.2646, "step": 6149 }, { "epoch": 2.719124313566506, "grad_norm": 0.3332429744054116, "learning_rate": 2.660560194306999e-07, "loss": 0.2862, "step": 6150 }, { "epoch": 2.7195665794420076, "grad_norm": 0.3862919244113495, "learning_rate": 2.652287934338377e-07, "loss": 0.297, "step": 6151 }, { "epoch": 2.72000884531751, "grad_norm": 0.3097239069853736, "learning_rate": 2.6440282041338307e-07, "loss": 0.2881, "step": 6152 }, { "epoch": 2.7204511111930123, "grad_norm": 0.29140619760802294, "learning_rate": 2.6357810058791564e-07, "loss": 0.2641, "step": 6153 }, { "epoch": 2.7208933770685144, "grad_norm": 0.3113172327618287, "learning_rate": 2.627546341756837e-07, "loss": 0.2803, "step": 6154 }, { "epoch": 2.7213356429440165, "grad_norm": 0.3295282036456247, "learning_rate": 2.619324213946034e-07, "loss": 0.2519, "step": 6155 }, { "epoch": 2.7217779088195186, "grad_norm": 0.542853540725736, "learning_rate": 2.611114624622585e-07, "loss": 0.274, "step": 6156 }, { "epoch": 2.7222201746950208, "grad_norm": 0.3639195149264314, "learning_rate": 2.6029175759590354e-07, "loss": 0.2897, "step": 6157 }, { "epoch": 2.722662440570523, "grad_norm": 0.46060959442079735, "learning_rate": 2.5947330701245775e-07, "loss": 0.2949, "step": 6158 }, { "epoch": 2.723104706446025, "grad_norm": 0.3131486660188938, "learning_rate": 2.586561109285124e-07, "loss": 0.28, "step": 6159 }, { "epoch": 2.723546972321527, "grad_norm": 0.3434400857783589, "learning_rate": 2.5784016956032163e-07, "loss": 0.2878, "step": 6160 }, { "epoch": 2.7239892381970296, "grad_norm": 0.28920336315125467, "learning_rate": 2.5702548312381117e-07, "loss": 0.264, "step": 6161 }, { "epoch": 2.7244315040725318, "grad_norm": 0.38093085124590637, "learning_rate": 2.5621205183457507e-07, "loss": 0.2794, "step": 6162 }, { "epoch": 2.724873769948034, "grad_norm": 0.33073008552114314, "learning_rate": 2.553998759078735e-07, "loss": 0.251, "step": 6163 }, { "epoch": 2.725316035823536, "grad_norm": 0.3215390975996645, "learning_rate": 2.5458895555863504e-07, "loss": 0.3004, "step": 6164 }, { "epoch": 2.725758301699038, "grad_norm": 0.40552680300251087, "learning_rate": 2.5377929100145526e-07, "loss": 0.2617, "step": 6165 }, { "epoch": 2.72620056757454, "grad_norm": 0.3374895974989695, "learning_rate": 2.5297088245059886e-07, "loss": 0.2747, "step": 6166 }, { "epoch": 2.7266428334500423, "grad_norm": 0.5542990498083997, "learning_rate": 2.5216373011999697e-07, "loss": 0.2594, "step": 6167 }, { "epoch": 2.7270850993255444, "grad_norm": 0.32160327793992893, "learning_rate": 2.513578342232487e-07, "loss": 0.2545, "step": 6168 }, { "epoch": 2.7275273652010465, "grad_norm": 0.356807524252808, "learning_rate": 2.505531949736206e-07, "loss": 0.2639, "step": 6169 }, { "epoch": 2.727969631076549, "grad_norm": 0.31037266148893383, "learning_rate": 2.49749812584047e-07, "loss": 0.2836, "step": 6170 }, { "epoch": 2.7284118969520508, "grad_norm": 0.38395258128577703, "learning_rate": 2.489476872671287e-07, "loss": 0.2676, "step": 6171 }, { "epoch": 2.7288541628275533, "grad_norm": 0.3376995178818875, "learning_rate": 2.4814681923513493e-07, "loss": 0.2501, "step": 6172 }, { "epoch": 2.7292964287030554, "grad_norm": 0.39163145539877176, "learning_rate": 2.473472087000023e-07, "loss": 0.2623, "step": 6173 }, { "epoch": 2.7297386945785576, "grad_norm": 0.3375516636538508, "learning_rate": 2.465488558733331e-07, "loss": 0.2591, "step": 6174 }, { "epoch": 2.7301809604540597, "grad_norm": 0.3705348784537606, "learning_rate": 2.4575176096639897e-07, "loss": 0.2693, "step": 6175 }, { "epoch": 2.730623226329562, "grad_norm": 0.3100412545022364, "learning_rate": 2.449559241901367e-07, "loss": 0.2439, "step": 6176 }, { "epoch": 2.731065492205064, "grad_norm": 0.301705705726356, "learning_rate": 2.441613457551495e-07, "loss": 0.2614, "step": 6177 }, { "epoch": 2.731507758080566, "grad_norm": 0.3228799448825708, "learning_rate": 2.433680258717114e-07, "loss": 0.2941, "step": 6178 }, { "epoch": 2.731950023956068, "grad_norm": 0.31862534507325574, "learning_rate": 2.4257596474976055e-07, "loss": 0.2332, "step": 6179 }, { "epoch": 2.7323922898315702, "grad_norm": 0.3482243153478689, "learning_rate": 2.4178516259890097e-07, "loss": 0.28, "step": 6180 }, { "epoch": 2.732834555707073, "grad_norm": 0.29875306384179173, "learning_rate": 2.409956196284063e-07, "loss": 0.2435, "step": 6181 }, { "epoch": 2.7332768215825745, "grad_norm": 0.3662698033894095, "learning_rate": 2.4020733604721513e-07, "loss": 0.3163, "step": 6182 }, { "epoch": 2.733719087458077, "grad_norm": 0.36350699176763845, "learning_rate": 2.394203120639338e-07, "loss": 0.2683, "step": 6183 }, { "epoch": 2.734161353333579, "grad_norm": 0.3887160681985706, "learning_rate": 2.386345478868335e-07, "loss": 0.2485, "step": 6184 }, { "epoch": 2.7346036192090812, "grad_norm": 0.3412960359956296, "learning_rate": 2.3785004372385456e-07, "loss": 0.2581, "step": 6185 }, { "epoch": 2.7350458850845834, "grad_norm": 0.34750335593755405, "learning_rate": 2.3706679978260206e-07, "loss": 0.2925, "step": 6186 }, { "epoch": 2.7354881509600855, "grad_norm": 0.3795814522370479, "learning_rate": 2.362848162703474e-07, "loss": 0.3175, "step": 6187 }, { "epoch": 2.7359304168355876, "grad_norm": 0.289348166425223, "learning_rate": 2.3550409339403058e-07, "loss": 0.2444, "step": 6188 }, { "epoch": 2.7363726827110897, "grad_norm": 0.34647487654765496, "learning_rate": 2.3472463136025636e-07, "loss": 0.2494, "step": 6189 }, { "epoch": 2.7368149485865922, "grad_norm": 0.3592553142556816, "learning_rate": 2.3394643037529642e-07, "loss": 0.3053, "step": 6190 }, { "epoch": 2.737257214462094, "grad_norm": 0.312735409298561, "learning_rate": 2.331694906450871e-07, "loss": 0.2686, "step": 6191 }, { "epoch": 2.7376994803375965, "grad_norm": 0.3015192811546745, "learning_rate": 2.3239381237523228e-07, "loss": 0.2536, "step": 6192 }, { "epoch": 2.7381417462130986, "grad_norm": 0.31952078894168146, "learning_rate": 2.3161939577100222e-07, "loss": 0.293, "step": 6193 }, { "epoch": 2.7385840120886007, "grad_norm": 0.3433770464043433, "learning_rate": 2.3084624103733412e-07, "loss": 0.2787, "step": 6194 }, { "epoch": 2.739026277964103, "grad_norm": 0.3218546589775642, "learning_rate": 2.3007434837882926e-07, "loss": 0.3008, "step": 6195 }, { "epoch": 2.739468543839605, "grad_norm": 0.28394427747601975, "learning_rate": 2.2930371799975593e-07, "loss": 0.2745, "step": 6196 }, { "epoch": 2.739910809715107, "grad_norm": 0.29302683530394785, "learning_rate": 2.2853435010404823e-07, "loss": 0.2593, "step": 6197 }, { "epoch": 2.740353075590609, "grad_norm": 0.3492895591795883, "learning_rate": 2.2776624489530664e-07, "loss": 0.2798, "step": 6198 }, { "epoch": 2.7407953414661113, "grad_norm": 0.3128565867278914, "learning_rate": 2.2699940257679742e-07, "loss": 0.2582, "step": 6199 }, { "epoch": 2.7412376073416134, "grad_norm": 0.303579519304344, "learning_rate": 2.2623382335145105e-07, "loss": 0.2853, "step": 6200 }, { "epoch": 2.741679873217116, "grad_norm": 0.3438160717536974, "learning_rate": 2.25469507421866e-07, "loss": 0.264, "step": 6201 }, { "epoch": 2.7421221390926176, "grad_norm": 0.385018425633065, "learning_rate": 2.2470645499030498e-07, "loss": 0.2526, "step": 6202 }, { "epoch": 2.74256440496812, "grad_norm": 0.3040016704985247, "learning_rate": 2.2394466625869637e-07, "loss": 0.2803, "step": 6203 }, { "epoch": 2.7430066708436223, "grad_norm": 0.3066319526216355, "learning_rate": 2.231841414286362e-07, "loss": 0.2425, "step": 6204 }, { "epoch": 2.7434489367191244, "grad_norm": 0.35183158185809216, "learning_rate": 2.224248807013829e-07, "loss": 0.2269, "step": 6205 }, { "epoch": 2.7438912025946265, "grad_norm": 0.38718822716752066, "learning_rate": 2.21666884277863e-07, "loss": 0.2489, "step": 6206 }, { "epoch": 2.7443334684701286, "grad_norm": 0.28954660943474386, "learning_rate": 2.2091015235866597e-07, "loss": 0.2422, "step": 6207 }, { "epoch": 2.7447757343456307, "grad_norm": 0.33315536385769245, "learning_rate": 2.201546851440489e-07, "loss": 0.2529, "step": 6208 }, { "epoch": 2.745218000221133, "grad_norm": 0.2751586801267218, "learning_rate": 2.1940048283393244e-07, "loss": 0.25, "step": 6209 }, { "epoch": 2.745660266096635, "grad_norm": 0.31754856239712537, "learning_rate": 2.1864754562790468e-07, "loss": 0.2525, "step": 6210 }, { "epoch": 2.746102531972137, "grad_norm": 0.29504576420149153, "learning_rate": 2.1789587372521627e-07, "loss": 0.2802, "step": 6211 }, { "epoch": 2.7465447978476396, "grad_norm": 0.2853128732167867, "learning_rate": 2.1714546732478581e-07, "loss": 0.269, "step": 6212 }, { "epoch": 2.7469870637231417, "grad_norm": 0.29244928755553457, "learning_rate": 2.163963266251945e-07, "loss": 0.2773, "step": 6213 }, { "epoch": 2.747429329598644, "grad_norm": 0.47044760081601195, "learning_rate": 2.156484518246893e-07, "loss": 0.2581, "step": 6214 }, { "epoch": 2.747871595474146, "grad_norm": 0.36828796963742855, "learning_rate": 2.1490184312118356e-07, "loss": 0.263, "step": 6215 }, { "epoch": 2.748313861349648, "grad_norm": 0.3819191325918007, "learning_rate": 2.1415650071225425e-07, "loss": 0.2842, "step": 6216 }, { "epoch": 2.74875612722515, "grad_norm": 0.3054502147957532, "learning_rate": 2.13412424795143e-07, "loss": 0.2635, "step": 6217 }, { "epoch": 2.7491983931006523, "grad_norm": 0.3052933802562839, "learning_rate": 2.1266961556675736e-07, "loss": 0.2903, "step": 6218 }, { "epoch": 2.7496406589761544, "grad_norm": 0.3099304265379618, "learning_rate": 2.1192807322366838e-07, "loss": 0.2645, "step": 6219 }, { "epoch": 2.7500829248516565, "grad_norm": 0.3816059463114143, "learning_rate": 2.1118779796211352e-07, "loss": 0.2741, "step": 6220 }, { "epoch": 2.750525190727159, "grad_norm": 0.2983618070766435, "learning_rate": 2.1044878997799389e-07, "loss": 0.2414, "step": 6221 }, { "epoch": 2.7509674566026607, "grad_norm": 0.4271651675825456, "learning_rate": 2.0971104946687525e-07, "loss": 0.2759, "step": 6222 }, { "epoch": 2.7514097224781633, "grad_norm": 0.4257780580492268, "learning_rate": 2.0897457662398702e-07, "loss": 0.2572, "step": 6223 }, { "epoch": 2.7518519883536654, "grad_norm": 0.33502507352656197, "learning_rate": 2.0823937164422492e-07, "loss": 0.2828, "step": 6224 }, { "epoch": 2.7522942542291675, "grad_norm": 0.32367334349085175, "learning_rate": 2.0750543472214836e-07, "loss": 0.2583, "step": 6225 }, { "epoch": 2.7527365201046696, "grad_norm": 0.2822104825925981, "learning_rate": 2.0677276605198193e-07, "loss": 0.2926, "step": 6226 }, { "epoch": 2.7531787859801717, "grad_norm": 0.3584223256440699, "learning_rate": 2.0604136582761335e-07, "loss": 0.2336, "step": 6227 }, { "epoch": 2.753621051855674, "grad_norm": 0.3546939958403491, "learning_rate": 2.0531123424259448e-07, "loss": 0.246, "step": 6228 }, { "epoch": 2.754063317731176, "grad_norm": 0.3061941116010482, "learning_rate": 2.0458237149014347e-07, "loss": 0.2733, "step": 6229 }, { "epoch": 2.754505583606678, "grad_norm": 0.3503336165906419, "learning_rate": 2.0385477776314056e-07, "loss": 0.25, "step": 6230 }, { "epoch": 2.75494784948218, "grad_norm": 0.36252621204520025, "learning_rate": 2.031284532541311e-07, "loss": 0.2748, "step": 6231 }, { "epoch": 2.7553901153576827, "grad_norm": 0.37791382699082526, "learning_rate": 2.0240339815532472e-07, "loss": 0.2635, "step": 6232 }, { "epoch": 2.7558323812331844, "grad_norm": 0.32160802813623096, "learning_rate": 2.0167961265859515e-07, "loss": 0.2604, "step": 6233 }, { "epoch": 2.756274647108687, "grad_norm": 0.3605608622647959, "learning_rate": 2.0095709695547915e-07, "loss": 0.2857, "step": 6234 }, { "epoch": 2.756716912984189, "grad_norm": 0.36839502015846304, "learning_rate": 2.002358512371777e-07, "loss": 0.2538, "step": 6235 }, { "epoch": 2.757159178859691, "grad_norm": 0.32698853091043933, "learning_rate": 1.995158756945581e-07, "loss": 0.2677, "step": 6236 }, { "epoch": 2.7576014447351933, "grad_norm": 0.3448134688007764, "learning_rate": 1.9879717051814851e-07, "loss": 0.3068, "step": 6237 }, { "epoch": 2.7580437106106954, "grad_norm": 0.3021411543197005, "learning_rate": 1.9807973589814178e-07, "loss": 0.2663, "step": 6238 }, { "epoch": 2.7584859764861975, "grad_norm": 0.28898500863111387, "learning_rate": 1.9736357202439492e-07, "loss": 0.2668, "step": 6239 }, { "epoch": 2.7589282423616996, "grad_norm": 0.3374944714906738, "learning_rate": 1.9664867908642749e-07, "loss": 0.2703, "step": 6240 }, { "epoch": 2.759370508237202, "grad_norm": 0.2954425922583511, "learning_rate": 1.9593505727342587e-07, "loss": 0.2209, "step": 6241 }, { "epoch": 2.759812774112704, "grad_norm": 0.3577841845446629, "learning_rate": 1.9522270677423628e-07, "loss": 0.2713, "step": 6242 }, { "epoch": 2.7602550399882064, "grad_norm": 0.29069446479872074, "learning_rate": 1.9451162777737064e-07, "loss": 0.2729, "step": 6243 }, { "epoch": 2.7606973058637085, "grad_norm": 0.3598513781373461, "learning_rate": 1.9380182047100404e-07, "loss": 0.2767, "step": 6244 }, { "epoch": 2.7611395717392107, "grad_norm": 0.3966058715190196, "learning_rate": 1.9309328504297453e-07, "loss": 0.2719, "step": 6245 }, { "epoch": 2.7615818376147128, "grad_norm": 0.34946618437693633, "learning_rate": 1.9238602168078434e-07, "loss": 0.2794, "step": 6246 }, { "epoch": 2.762024103490215, "grad_norm": 0.355111109855697, "learning_rate": 1.9168003057159767e-07, "loss": 0.2528, "step": 6247 }, { "epoch": 2.762466369365717, "grad_norm": 0.4174019139532519, "learning_rate": 1.9097531190224395e-07, "loss": 0.2624, "step": 6248 }, { "epoch": 2.762908635241219, "grad_norm": 0.2726524707379807, "learning_rate": 1.902718658592151e-07, "loss": 0.2498, "step": 6249 }, { "epoch": 2.763350901116721, "grad_norm": 0.4087191010778335, "learning_rate": 1.8956969262866499e-07, "loss": 0.2535, "step": 6250 }, { "epoch": 2.7637931669922233, "grad_norm": 0.30261606986201167, "learning_rate": 1.888687923964122e-07, "loss": 0.2699, "step": 6251 }, { "epoch": 2.764235432867726, "grad_norm": 0.4376486137406308, "learning_rate": 1.8816916534794005e-07, "loss": 0.2585, "step": 6252 }, { "epoch": 2.7646776987432276, "grad_norm": 0.7069483197792326, "learning_rate": 1.8747081166838986e-07, "loss": 0.2989, "step": 6253 }, { "epoch": 2.76511996461873, "grad_norm": 0.3462786201768452, "learning_rate": 1.867737315425705e-07, "loss": 0.2343, "step": 6254 }, { "epoch": 2.765562230494232, "grad_norm": 0.28898675638617966, "learning_rate": 1.8607792515495216e-07, "loss": 0.2828, "step": 6255 }, { "epoch": 2.7660044963697343, "grad_norm": 0.34124263544530203, "learning_rate": 1.8538339268966755e-07, "loss": 0.2664, "step": 6256 }, { "epoch": 2.7664467622452364, "grad_norm": 0.3276778497961183, "learning_rate": 1.846901343305141e-07, "loss": 0.2694, "step": 6257 }, { "epoch": 2.7668890281207386, "grad_norm": 0.3647096409903832, "learning_rate": 1.8399815026095004e-07, "loss": 0.2876, "step": 6258 }, { "epoch": 2.7673312939962407, "grad_norm": 0.2811894392193494, "learning_rate": 1.8330744066409723e-07, "loss": 0.2566, "step": 6259 }, { "epoch": 2.767773559871743, "grad_norm": 0.29552697932189126, "learning_rate": 1.8261800572274001e-07, "loss": 0.2636, "step": 6260 }, { "epoch": 2.768215825747245, "grad_norm": 0.324906555908087, "learning_rate": 1.819298456193258e-07, "loss": 0.2815, "step": 6261 }, { "epoch": 2.768658091622747, "grad_norm": 0.47259299638747965, "learning_rate": 1.8124296053596446e-07, "loss": 0.2601, "step": 6262 }, { "epoch": 2.7691003574982496, "grad_norm": 0.35008333297977445, "learning_rate": 1.805573506544278e-07, "loss": 0.2691, "step": 6263 }, { "epoch": 2.7695426233737517, "grad_norm": 0.39131836426464217, "learning_rate": 1.7987301615615184e-07, "loss": 0.2986, "step": 6264 }, { "epoch": 2.769984889249254, "grad_norm": 0.27975638607154296, "learning_rate": 1.7918995722223332e-07, "loss": 0.2406, "step": 6265 }, { "epoch": 2.770427155124756, "grad_norm": 0.37446849737292, "learning_rate": 1.7850817403343212e-07, "loss": 0.2853, "step": 6266 }, { "epoch": 2.770869421000258, "grad_norm": 0.29382134247922936, "learning_rate": 1.7782766677017006e-07, "loss": 0.2676, "step": 6267 }, { "epoch": 2.77131168687576, "grad_norm": 0.3495564078730832, "learning_rate": 1.7714843561253413e-07, "loss": 0.2601, "step": 6268 }, { "epoch": 2.7717539527512622, "grad_norm": 0.332553273040729, "learning_rate": 1.764704807402684e-07, "loss": 0.284, "step": 6269 }, { "epoch": 2.7721962186267644, "grad_norm": 0.34461638989524174, "learning_rate": 1.757938023327832e-07, "loss": 0.2626, "step": 6270 }, { "epoch": 2.7726384845022665, "grad_norm": 0.29295720710987483, "learning_rate": 1.751184005691503e-07, "loss": 0.2774, "step": 6271 }, { "epoch": 2.773080750377769, "grad_norm": 0.357310246002941, "learning_rate": 1.7444427562810173e-07, "loss": 0.2615, "step": 6272 }, { "epoch": 2.7735230162532707, "grad_norm": 0.2959029620216933, "learning_rate": 1.7377142768803535e-07, "loss": 0.2504, "step": 6273 }, { "epoch": 2.7739652821287732, "grad_norm": 0.3320876241489794, "learning_rate": 1.730998569270076e-07, "loss": 0.2823, "step": 6274 }, { "epoch": 2.7744075480042754, "grad_norm": 0.3591039571625444, "learning_rate": 1.7242956352273854e-07, "loss": 0.2513, "step": 6275 }, { "epoch": 2.7748498138797775, "grad_norm": 0.3444365766165889, "learning_rate": 1.7176054765260963e-07, "loss": 0.2431, "step": 6276 }, { "epoch": 2.7752920797552796, "grad_norm": 0.3822684939769085, "learning_rate": 1.710928094936648e-07, "loss": 0.2546, "step": 6277 }, { "epoch": 2.7757343456307817, "grad_norm": 0.4089287399800861, "learning_rate": 1.7042634922260935e-07, "loss": 0.2563, "step": 6278 }, { "epoch": 2.776176611506284, "grad_norm": 0.35944610895957874, "learning_rate": 1.6976116701581114e-07, "loss": 0.2637, "step": 6279 }, { "epoch": 2.776618877381786, "grad_norm": 0.32283701784472607, "learning_rate": 1.690972630492982e-07, "loss": 0.2666, "step": 6280 }, { "epoch": 2.777061143257288, "grad_norm": 0.3665418726565398, "learning_rate": 1.6843463749876221e-07, "loss": 0.2726, "step": 6281 }, { "epoch": 2.77750340913279, "grad_norm": 0.3181258087014156, "learning_rate": 1.6777329053955572e-07, "loss": 0.2479, "step": 6282 }, { "epoch": 2.7779456750082927, "grad_norm": 0.3534260660592554, "learning_rate": 1.6711322234669258e-07, "loss": 0.2469, "step": 6283 }, { "epoch": 2.7783879408837944, "grad_norm": 0.3227608021429406, "learning_rate": 1.6645443309484866e-07, "loss": 0.2704, "step": 6284 }, { "epoch": 2.778830206759297, "grad_norm": 0.3655525792233656, "learning_rate": 1.6579692295836168e-07, "loss": 0.252, "step": 6285 }, { "epoch": 2.779272472634799, "grad_norm": 0.33197672522204036, "learning_rate": 1.6514069211122975e-07, "loss": 0.2703, "step": 6286 }, { "epoch": 2.779714738510301, "grad_norm": 0.41207750060373194, "learning_rate": 1.6448574072711388e-07, "loss": 0.2746, "step": 6287 }, { "epoch": 2.7801570043858033, "grad_norm": 0.41491549018468754, "learning_rate": 1.638320689793349e-07, "loss": 0.2669, "step": 6288 }, { "epoch": 2.7805992702613054, "grad_norm": 0.282688317213578, "learning_rate": 1.6317967704087723e-07, "loss": 0.2479, "step": 6289 }, { "epoch": 2.7810415361368075, "grad_norm": 0.3150346960782387, "learning_rate": 1.6252856508438385e-07, "loss": 0.2746, "step": 6290 }, { "epoch": 2.7814838020123096, "grad_norm": 0.40988042241347894, "learning_rate": 1.6187873328216142e-07, "loss": 0.2778, "step": 6291 }, { "epoch": 2.781926067887812, "grad_norm": 0.32235704510360935, "learning_rate": 1.6123018180617734e-07, "loss": 0.2495, "step": 6292 }, { "epoch": 2.782368333763314, "grad_norm": 0.32435737344602095, "learning_rate": 1.605829108280571e-07, "loss": 0.2711, "step": 6293 }, { "epoch": 2.7828105996388164, "grad_norm": 0.35166339759990317, "learning_rate": 1.5993692051909259e-07, "loss": 0.2889, "step": 6294 }, { "epoch": 2.7832528655143185, "grad_norm": 0.39007009280205074, "learning_rate": 1.5929221105023318e-07, "loss": 0.2985, "step": 6295 }, { "epoch": 2.7836951313898206, "grad_norm": 0.42633566336615586, "learning_rate": 1.5864878259209015e-07, "loss": 0.2614, "step": 6296 }, { "epoch": 2.7841373972653227, "grad_norm": 0.3281768904257308, "learning_rate": 1.580066353149362e-07, "loss": 0.2892, "step": 6297 }, { "epoch": 2.784579663140825, "grad_norm": 0.36526868407658425, "learning_rate": 1.5736576938870373e-07, "loss": 0.2641, "step": 6298 }, { "epoch": 2.785021929016327, "grad_norm": 0.3929846091552261, "learning_rate": 1.567261849829882e-07, "loss": 0.2922, "step": 6299 }, { "epoch": 2.785464194891829, "grad_norm": 0.3191719503677726, "learning_rate": 1.560878822670442e-07, "loss": 0.2793, "step": 6300 }, { "epoch": 2.785906460767331, "grad_norm": 0.32552122588861415, "learning_rate": 1.5545086140978772e-07, "loss": 0.242, "step": 6301 }, { "epoch": 2.7863487266428333, "grad_norm": 0.3321519340426613, "learning_rate": 1.5481512257979502e-07, "loss": 0.2619, "step": 6302 }, { "epoch": 2.786790992518336, "grad_norm": 0.2898475159529574, "learning_rate": 1.5418066594530378e-07, "loss": 0.2869, "step": 6303 }, { "epoch": 2.7872332583938375, "grad_norm": 0.33812824884275305, "learning_rate": 1.5354749167421191e-07, "loss": 0.283, "step": 6304 }, { "epoch": 2.78767552426934, "grad_norm": 0.24928492607609645, "learning_rate": 1.5291559993407868e-07, "loss": 0.2397, "step": 6305 }, { "epoch": 2.788117790144842, "grad_norm": 0.4815883973405139, "learning_rate": 1.5228499089212367e-07, "loss": 0.2766, "step": 6306 }, { "epoch": 2.7885600560203443, "grad_norm": 0.3764652700121932, "learning_rate": 1.516556647152262e-07, "loss": 0.2608, "step": 6307 }, { "epoch": 2.7890023218958464, "grad_norm": 0.4078093774347163, "learning_rate": 1.5102762156992746e-07, "loss": 0.2695, "step": 6308 }, { "epoch": 2.7894445877713485, "grad_norm": 0.3501567787965059, "learning_rate": 1.5040086162242673e-07, "loss": 0.2773, "step": 6309 }, { "epoch": 2.7898868536468506, "grad_norm": 0.34739921802581075, "learning_rate": 1.497753850385869e-07, "loss": 0.2586, "step": 6310 }, { "epoch": 2.7903291195223527, "grad_norm": 0.37105904341513457, "learning_rate": 1.4915119198392892e-07, "loss": 0.2418, "step": 6311 }, { "epoch": 2.790771385397855, "grad_norm": 0.33628329393176326, "learning_rate": 1.485282826236356e-07, "loss": 0.277, "step": 6312 }, { "epoch": 2.791213651273357, "grad_norm": 0.32387985008014664, "learning_rate": 1.4790665712254793e-07, "loss": 0.293, "step": 6313 }, { "epoch": 2.7916559171488595, "grad_norm": 0.32334049898709094, "learning_rate": 1.4728631564516982e-07, "loss": 0.2476, "step": 6314 }, { "epoch": 2.7920981830243616, "grad_norm": 0.3136290466061874, "learning_rate": 1.4666725835566388e-07, "loss": 0.2474, "step": 6315 }, { "epoch": 2.7925404488998637, "grad_norm": 0.46786559262453026, "learning_rate": 1.460494854178518e-07, "loss": 0.2798, "step": 6316 }, { "epoch": 2.792982714775366, "grad_norm": 0.3362426438034283, "learning_rate": 1.454329969952184e-07, "loss": 0.243, "step": 6317 }, { "epoch": 2.793424980650868, "grad_norm": 0.31828434220170093, "learning_rate": 1.4481779325090538e-07, "loss": 0.2549, "step": 6318 }, { "epoch": 2.79386724652637, "grad_norm": 0.3303065735581948, "learning_rate": 1.4420387434771698e-07, "loss": 0.2646, "step": 6319 }, { "epoch": 2.794309512401872, "grad_norm": 0.3454085130964522, "learning_rate": 1.435912404481149e-07, "loss": 0.2652, "step": 6320 }, { "epoch": 2.7947517782773743, "grad_norm": 0.40293768020403065, "learning_rate": 1.4297989171422443e-07, "loss": 0.2708, "step": 6321 }, { "epoch": 2.7951940441528764, "grad_norm": 0.3616318172545895, "learning_rate": 1.4236982830782676e-07, "loss": 0.239, "step": 6322 }, { "epoch": 2.795636310028379, "grad_norm": 0.3424657047476419, "learning_rate": 1.4176105039036659e-07, "loss": 0.2558, "step": 6323 }, { "epoch": 2.7960785759038806, "grad_norm": 0.3239695445859339, "learning_rate": 1.4115355812294395e-07, "loss": 0.2524, "step": 6324 }, { "epoch": 2.796520841779383, "grad_norm": 0.4977817788275513, "learning_rate": 1.4054735166632304e-07, "loss": 0.2458, "step": 6325 }, { "epoch": 2.7969631076548853, "grad_norm": 0.3598903096510685, "learning_rate": 1.399424311809261e-07, "loss": 0.2533, "step": 6326 }, { "epoch": 2.7974053735303874, "grad_norm": 0.3015482646832331, "learning_rate": 1.393387968268345e-07, "loss": 0.2721, "step": 6327 }, { "epoch": 2.7978476394058895, "grad_norm": 0.27479164304648035, "learning_rate": 1.3873644876379046e-07, "loss": 0.2663, "step": 6328 }, { "epoch": 2.7982899052813917, "grad_norm": 0.3538885928296776, "learning_rate": 1.3813538715119423e-07, "loss": 0.2503, "step": 6329 }, { "epoch": 2.7987321711568938, "grad_norm": 0.4622952502244271, "learning_rate": 1.3753561214810696e-07, "loss": 0.2707, "step": 6330 }, { "epoch": 2.799174437032396, "grad_norm": 0.3707649909687692, "learning_rate": 1.3693712391324832e-07, "loss": 0.2571, "step": 6331 }, { "epoch": 2.799616702907898, "grad_norm": 0.29646935755424275, "learning_rate": 1.363399226049994e-07, "loss": 0.2797, "step": 6332 }, { "epoch": 2.8000589687834, "grad_norm": 0.35446436613912025, "learning_rate": 1.3574400838139768e-07, "loss": 0.2651, "step": 6333 }, { "epoch": 2.8005012346589027, "grad_norm": 0.3197861142604679, "learning_rate": 1.351493814001431e-07, "loss": 0.2583, "step": 6334 }, { "epoch": 2.8009435005344043, "grad_norm": 0.35367651627749896, "learning_rate": 1.3455604181859206e-07, "loss": 0.2655, "step": 6335 }, { "epoch": 2.801385766409907, "grad_norm": 0.3748566707302233, "learning_rate": 1.339639897937628e-07, "loss": 0.272, "step": 6336 }, { "epoch": 2.801828032285409, "grad_norm": 0.2535734616295034, "learning_rate": 1.3337322548233223e-07, "loss": 0.2688, "step": 6337 }, { "epoch": 2.802270298160911, "grad_norm": 0.31035476894090763, "learning_rate": 1.3278374904063473e-07, "loss": 0.2342, "step": 6338 }, { "epoch": 2.8027125640364132, "grad_norm": 0.3275668524561321, "learning_rate": 1.3219556062466665e-07, "loss": 0.2663, "step": 6339 }, { "epoch": 2.8031548299119153, "grad_norm": 0.2943388704722098, "learning_rate": 1.3160866039008123e-07, "loss": 0.2662, "step": 6340 }, { "epoch": 2.8035970957874174, "grad_norm": 0.4128546710557439, "learning_rate": 1.310230484921904e-07, "loss": 0.2415, "step": 6341 }, { "epoch": 2.8040393616629196, "grad_norm": 0.36550572193650654, "learning_rate": 1.304387250859679e-07, "loss": 0.263, "step": 6342 }, { "epoch": 2.804481627538422, "grad_norm": 0.3838609128833974, "learning_rate": 1.2985569032604507e-07, "loss": 0.2453, "step": 6343 }, { "epoch": 2.804923893413924, "grad_norm": 0.4804446511547815, "learning_rate": 1.292739443667118e-07, "loss": 0.2614, "step": 6344 }, { "epoch": 2.8053661592894263, "grad_norm": 0.32693911287166716, "learning_rate": 1.2869348736191666e-07, "loss": 0.2829, "step": 6345 }, { "epoch": 2.8058084251649285, "grad_norm": 0.3684809651736643, "learning_rate": 1.2811431946526786e-07, "loss": 0.2726, "step": 6346 }, { "epoch": 2.8062506910404306, "grad_norm": 0.3095743334779584, "learning_rate": 1.275364408300328e-07, "loss": 0.2596, "step": 6347 }, { "epoch": 2.8066929569159327, "grad_norm": 0.2970853175313266, "learning_rate": 1.2695985160913692e-07, "loss": 0.2552, "step": 6348 }, { "epoch": 2.807135222791435, "grad_norm": 0.3655547936210256, "learning_rate": 1.2638455195516487e-07, "loss": 0.2757, "step": 6349 }, { "epoch": 2.807577488666937, "grad_norm": 0.37626357361499685, "learning_rate": 1.258105420203598e-07, "loss": 0.2551, "step": 6350 }, { "epoch": 2.808019754542439, "grad_norm": 0.3056334868182495, "learning_rate": 1.2523782195662303e-07, "loss": 0.2376, "step": 6351 }, { "epoch": 2.808462020417941, "grad_norm": 0.3152750448642099, "learning_rate": 1.246663919155161e-07, "loss": 0.2675, "step": 6352 }, { "epoch": 2.8089042862934432, "grad_norm": 0.3497835942971208, "learning_rate": 1.2409625204825802e-07, "loss": 0.2595, "step": 6353 }, { "epoch": 2.809346552168946, "grad_norm": 0.34955124051281555, "learning_rate": 1.2352740250572647e-07, "loss": 0.2451, "step": 6354 }, { "epoch": 2.8097888180444475, "grad_norm": 0.6384783291091912, "learning_rate": 1.2295984343845768e-07, "loss": 0.273, "step": 6355 }, { "epoch": 2.81023108391995, "grad_norm": 0.42360630266516414, "learning_rate": 1.2239357499664596e-07, "loss": 0.2633, "step": 6356 }, { "epoch": 2.810673349795452, "grad_norm": 0.3307666476968353, "learning_rate": 1.218285973301453e-07, "loss": 0.2679, "step": 6357 }, { "epoch": 2.8111156156709542, "grad_norm": 0.3635129545717591, "learning_rate": 1.2126491058846723e-07, "loss": 0.235, "step": 6358 }, { "epoch": 2.8115578815464564, "grad_norm": 0.39810923737992204, "learning_rate": 1.2070251492078189e-07, "loss": 0.265, "step": 6359 }, { "epoch": 2.8120001474219585, "grad_norm": 0.3628817183093948, "learning_rate": 1.2014141047591742e-07, "loss": 0.2645, "step": 6360 }, { "epoch": 2.8124424132974606, "grad_norm": 0.34795514222142093, "learning_rate": 1.195815974023612e-07, "loss": 0.2519, "step": 6361 }, { "epoch": 2.8128846791729627, "grad_norm": 0.2926384764418154, "learning_rate": 1.1902307584825745e-07, "loss": 0.2521, "step": 6362 }, { "epoch": 2.813326945048465, "grad_norm": 0.28502220954371277, "learning_rate": 1.1846584596140964e-07, "loss": 0.2636, "step": 6363 }, { "epoch": 2.813769210923967, "grad_norm": 0.3733805092068417, "learning_rate": 1.1790990788927981e-07, "loss": 0.2617, "step": 6364 }, { "epoch": 2.8142114767994695, "grad_norm": 0.37142465114602985, "learning_rate": 1.1735526177898582e-07, "loss": 0.2527, "step": 6365 }, { "epoch": 2.8146537426749716, "grad_norm": 0.31400648090437755, "learning_rate": 1.1680190777730694e-07, "loss": 0.2721, "step": 6366 }, { "epoch": 2.8150960085504737, "grad_norm": 0.35362354088586717, "learning_rate": 1.1624984603067768e-07, "loss": 0.2631, "step": 6367 }, { "epoch": 2.815538274425976, "grad_norm": 0.2978018891391144, "learning_rate": 1.1569907668519287e-07, "loss": 0.2605, "step": 6368 }, { "epoch": 2.815980540301478, "grad_norm": 0.3823258933044803, "learning_rate": 1.1514959988660368e-07, "loss": 0.2368, "step": 6369 }, { "epoch": 2.81642280617698, "grad_norm": 0.3485431210129929, "learning_rate": 1.1460141578032047e-07, "loss": 0.2802, "step": 6370 }, { "epoch": 2.816865072052482, "grad_norm": 0.362612076363403, "learning_rate": 1.1405452451140941e-07, "loss": 0.2806, "step": 6371 }, { "epoch": 2.8173073379279843, "grad_norm": 0.3285240349184043, "learning_rate": 1.1350892622459587e-07, "loss": 0.2926, "step": 6372 }, { "epoch": 2.8177496038034864, "grad_norm": 0.35810544885337675, "learning_rate": 1.1296462106426487e-07, "loss": 0.2687, "step": 6373 }, { "epoch": 2.818191869678989, "grad_norm": 0.36633386137671825, "learning_rate": 1.1242160917445621e-07, "loss": 0.2938, "step": 6374 }, { "epoch": 2.8186341355544906, "grad_norm": 0.30243465012052156, "learning_rate": 1.1187989069886884e-07, "loss": 0.2672, "step": 6375 }, { "epoch": 2.819076401429993, "grad_norm": 0.4191520462943535, "learning_rate": 1.1133946578085975e-07, "loss": 0.2689, "step": 6376 }, { "epoch": 2.8195186673054953, "grad_norm": 0.37591043566191434, "learning_rate": 1.1080033456344286e-07, "loss": 0.3028, "step": 6377 }, { "epoch": 2.8199609331809974, "grad_norm": 0.35895864706619635, "learning_rate": 1.1026249718929072e-07, "loss": 0.2602, "step": 6378 }, { "epoch": 2.8204031990564995, "grad_norm": 0.3035815092540295, "learning_rate": 1.0972595380073115e-07, "loss": 0.2437, "step": 6379 }, { "epoch": 2.8208454649320016, "grad_norm": 0.3464369633498638, "learning_rate": 1.0919070453975278e-07, "loss": 0.2662, "step": 6380 }, { "epoch": 2.8212877308075037, "grad_norm": 0.26190522577800945, "learning_rate": 1.086567495480001e-07, "loss": 0.247, "step": 6381 }, { "epoch": 2.821729996683006, "grad_norm": 0.37051505141531466, "learning_rate": 1.0812408896677396e-07, "loss": 0.2717, "step": 6382 }, { "epoch": 2.822172262558508, "grad_norm": 0.3462234575557001, "learning_rate": 1.0759272293703494e-07, "loss": 0.254, "step": 6383 }, { "epoch": 2.82261452843401, "grad_norm": 0.35299939259623, "learning_rate": 1.0706265159939944e-07, "loss": 0.2492, "step": 6384 }, { "epoch": 2.8230567943095126, "grad_norm": 0.38897138377297286, "learning_rate": 1.0653387509414303e-07, "loss": 0.2603, "step": 6385 }, { "epoch": 2.8234990601850143, "grad_norm": 0.3309716114539192, "learning_rate": 1.0600639356119601e-07, "loss": 0.2551, "step": 6386 }, { "epoch": 2.823941326060517, "grad_norm": 0.3068428612335317, "learning_rate": 1.0548020714014783e-07, "loss": 0.2607, "step": 6387 }, { "epoch": 2.824383591936019, "grad_norm": 0.37647078701782555, "learning_rate": 1.0495531597024433e-07, "loss": 0.2851, "step": 6388 }, { "epoch": 2.824825857811521, "grad_norm": 0.35853393739695893, "learning_rate": 1.0443172019038939e-07, "loss": 0.2624, "step": 6389 }, { "epoch": 2.825268123687023, "grad_norm": 0.3305893467969975, "learning_rate": 1.0390941993914383e-07, "loss": 0.2648, "step": 6390 }, { "epoch": 2.8257103895625253, "grad_norm": 0.40685460916780125, "learning_rate": 1.0338841535472599e-07, "loss": 0.2616, "step": 6391 }, { "epoch": 2.8261526554380274, "grad_norm": 0.3555673491561749, "learning_rate": 1.0286870657500947e-07, "loss": 0.2892, "step": 6392 }, { "epoch": 2.8265949213135295, "grad_norm": 0.25103864421833433, "learning_rate": 1.0235029373752758e-07, "loss": 0.2775, "step": 6393 }, { "epoch": 2.827037187189032, "grad_norm": 0.33272941545374907, "learning_rate": 1.0183317697946893e-07, "loss": 0.2516, "step": 6394 }, { "epoch": 2.8274794530645337, "grad_norm": 0.3614754112451186, "learning_rate": 1.0131735643767959e-07, "loss": 0.2439, "step": 6395 }, { "epoch": 2.8279217189400363, "grad_norm": 0.2920840572607736, "learning_rate": 1.0080283224866261e-07, "loss": 0.2562, "step": 6396 }, { "epoch": 2.8283639848155384, "grad_norm": 0.38121809852247396, "learning_rate": 1.0028960454857795e-07, "loss": 0.2815, "step": 6397 }, { "epoch": 2.8288062506910405, "grad_norm": 0.3611065034545852, "learning_rate": 9.977767347324308e-08, "loss": 0.2551, "step": 6398 }, { "epoch": 2.8292485165665426, "grad_norm": 0.2777127071044596, "learning_rate": 9.926703915813129e-08, "loss": 0.2312, "step": 6399 }, { "epoch": 2.8296907824420447, "grad_norm": 0.3178009089035904, "learning_rate": 9.875770173837341e-08, "loss": 0.2689, "step": 6400 }, { "epoch": 2.830133048317547, "grad_norm": 0.41509687508273124, "learning_rate": 9.824966134875769e-08, "loss": 0.2903, "step": 6401 }, { "epoch": 2.830575314193049, "grad_norm": 0.36121508306521877, "learning_rate": 9.774291812372716e-08, "loss": 0.2886, "step": 6402 }, { "epoch": 2.831017580068551, "grad_norm": 0.4068232129031206, "learning_rate": 9.723747219738344e-08, "loss": 0.2764, "step": 6403 }, { "epoch": 2.831459845944053, "grad_norm": 0.3825630122673971, "learning_rate": 9.673332370348343e-08, "loss": 0.2482, "step": 6404 }, { "epoch": 2.8319021118195558, "grad_norm": 0.3678987122984374, "learning_rate": 9.623047277544207e-08, "loss": 0.2265, "step": 6405 }, { "epoch": 2.8323443776950574, "grad_norm": 0.3013213292221254, "learning_rate": 9.572891954633012e-08, "loss": 0.2651, "step": 6406 }, { "epoch": 2.83278664357056, "grad_norm": 0.3682538986506459, "learning_rate": 9.522866414887533e-08, "loss": 0.2778, "step": 6407 }, { "epoch": 2.833228909446062, "grad_norm": 0.3956821766986362, "learning_rate": 9.472970671546178e-08, "loss": 0.2627, "step": 6408 }, { "epoch": 2.833671175321564, "grad_norm": 0.4584434089781507, "learning_rate": 9.423204737812941e-08, "loss": 0.2357, "step": 6409 }, { "epoch": 2.8341134411970663, "grad_norm": 0.32312359039043803, "learning_rate": 9.373568626857566e-08, "loss": 0.2627, "step": 6410 }, { "epoch": 2.8345557070725684, "grad_norm": 0.3668480270092999, "learning_rate": 9.324062351815433e-08, "loss": 0.2713, "step": 6411 }, { "epoch": 2.8349979729480705, "grad_norm": 0.33381615956325683, "learning_rate": 9.274685925787507e-08, "loss": 0.2749, "step": 6412 }, { "epoch": 2.8354402388235727, "grad_norm": 0.47189957269077487, "learning_rate": 9.225439361840394e-08, "loss": 0.2486, "step": 6413 }, { "epoch": 2.8358825046990748, "grad_norm": 0.3594458907230851, "learning_rate": 9.176322673006332e-08, "loss": 0.2783, "step": 6414 }, { "epoch": 2.836324770574577, "grad_norm": 0.34618755442416455, "learning_rate": 9.12733587228326e-08, "loss": 0.2977, "step": 6415 }, { "epoch": 2.8367670364500794, "grad_norm": 0.3284354900202754, "learning_rate": 9.078478972634808e-08, "loss": 0.2802, "step": 6416 }, { "epoch": 2.8372093023255816, "grad_norm": 0.3527977336809966, "learning_rate": 9.029751986989966e-08, "loss": 0.2356, "step": 6417 }, { "epoch": 2.8376515682010837, "grad_norm": 0.38025122331435474, "learning_rate": 8.981154928243473e-08, "loss": 0.2559, "step": 6418 }, { "epoch": 2.8380938340765858, "grad_norm": 0.33641526669335575, "learning_rate": 8.93268780925588e-08, "loss": 0.2755, "step": 6419 }, { "epoch": 2.838536099952088, "grad_norm": 0.28010091155686967, "learning_rate": 8.884350642852979e-08, "loss": 0.2693, "step": 6420 }, { "epoch": 2.83897836582759, "grad_norm": 0.3143753958568381, "learning_rate": 8.836143441826539e-08, "loss": 0.2645, "step": 6421 }, { "epoch": 2.839420631703092, "grad_norm": 0.4128868048779906, "learning_rate": 8.788066218933689e-08, "loss": 0.2498, "step": 6422 }, { "epoch": 2.8398628975785942, "grad_norm": 0.31630768614556964, "learning_rate": 8.740118986897306e-08, "loss": 0.229, "step": 6423 }, { "epoch": 2.8403051634540963, "grad_norm": 0.34108578097399467, "learning_rate": 8.692301758405796e-08, "loss": 0.2253, "step": 6424 }, { "epoch": 2.840747429329599, "grad_norm": 0.30391180814562574, "learning_rate": 8.644614546113095e-08, "loss": 0.2525, "step": 6425 }, { "epoch": 2.8411896952051006, "grad_norm": 0.36720777091213674, "learning_rate": 8.597057362638883e-08, "loss": 0.2823, "step": 6426 }, { "epoch": 2.841631961080603, "grad_norm": 0.38310270015284326, "learning_rate": 8.549630220568372e-08, "loss": 0.2796, "step": 6427 }, { "epoch": 2.8420742269561052, "grad_norm": 0.3942157674972773, "learning_rate": 8.502333132452301e-08, "loss": 0.261, "step": 6428 }, { "epoch": 2.8425164928316073, "grad_norm": 0.3695700880052864, "learning_rate": 8.455166110806989e-08, "loss": 0.2645, "step": 6429 }, { "epoch": 2.8429587587071095, "grad_norm": 0.266776554506529, "learning_rate": 8.40812916811451e-08, "loss": 0.247, "step": 6430 }, { "epoch": 2.8434010245826116, "grad_norm": 0.3777679631340089, "learning_rate": 8.361222316822292e-08, "loss": 0.2871, "step": 6431 }, { "epoch": 2.8438432904581137, "grad_norm": 0.29278113295268726, "learning_rate": 8.314445569343521e-08, "loss": 0.2394, "step": 6432 }, { "epoch": 2.844285556333616, "grad_norm": 0.3616574076947047, "learning_rate": 8.267798938056737e-08, "loss": 0.2746, "step": 6433 }, { "epoch": 2.844727822209118, "grad_norm": 0.3096185338687272, "learning_rate": 8.221282435306289e-08, "loss": 0.2753, "step": 6434 }, { "epoch": 2.84517008808462, "grad_norm": 0.31292704916333564, "learning_rate": 8.174896073401883e-08, "loss": 0.2532, "step": 6435 }, { "epoch": 2.8456123539601226, "grad_norm": 0.8850704538062786, "learning_rate": 8.128639864618925e-08, "loss": 0.2549, "step": 6436 }, { "epoch": 2.8460546198356242, "grad_norm": 0.29426745674871885, "learning_rate": 8.0825138211984e-08, "loss": 0.2481, "step": 6437 }, { "epoch": 2.846496885711127, "grad_norm": 0.29186758625118414, "learning_rate": 8.036517955346657e-08, "loss": 0.2422, "step": 6438 }, { "epoch": 2.846939151586629, "grad_norm": 0.35455938143111915, "learning_rate": 7.990652279235789e-08, "loss": 0.2636, "step": 6439 }, { "epoch": 2.847381417462131, "grad_norm": 0.33426406457057084, "learning_rate": 7.944916805003367e-08, "loss": 0.2383, "step": 6440 }, { "epoch": 2.847823683337633, "grad_norm": 0.35834400471798866, "learning_rate": 7.89931154475243e-08, "loss": 0.2301, "step": 6441 }, { "epoch": 2.8482659492131353, "grad_norm": 0.30828535318202066, "learning_rate": 7.853836510551771e-08, "loss": 0.2322, "step": 6442 }, { "epoch": 2.8487082150886374, "grad_norm": 0.29378542676623587, "learning_rate": 7.808491714435428e-08, "loss": 0.2651, "step": 6443 }, { "epoch": 2.8491504809641395, "grad_norm": 0.32258882664814514, "learning_rate": 7.763277168403249e-08, "loss": 0.2673, "step": 6444 }, { "epoch": 2.849592746839642, "grad_norm": 0.31546748078583614, "learning_rate": 7.718192884420383e-08, "loss": 0.2804, "step": 6445 }, { "epoch": 2.8500350127151437, "grad_norm": 0.29625993626968433, "learning_rate": 7.673238874417676e-08, "loss": 0.2707, "step": 6446 }, { "epoch": 2.8504772785906463, "grad_norm": 0.29161602977882906, "learning_rate": 7.628415150291501e-08, "loss": 0.2775, "step": 6447 }, { "epoch": 2.8509195444661484, "grad_norm": 0.3890152936126343, "learning_rate": 7.583721723903593e-08, "loss": 0.2612, "step": 6448 }, { "epoch": 2.8513618103416505, "grad_norm": 0.2993605247196615, "learning_rate": 7.539158607081321e-08, "loss": 0.2422, "step": 6449 }, { "epoch": 2.8518040762171526, "grad_norm": 0.25560941809123294, "learning_rate": 7.494725811617587e-08, "loss": 0.2591, "step": 6450 }, { "epoch": 2.8522463420926547, "grad_norm": 0.40689779660396963, "learning_rate": 7.450423349270763e-08, "loss": 0.2985, "step": 6451 }, { "epoch": 2.852688607968157, "grad_norm": 0.3297979050942822, "learning_rate": 7.406251231764639e-08, "loss": 0.2585, "step": 6452 }, { "epoch": 2.853130873843659, "grad_norm": 0.3384829933219542, "learning_rate": 7.362209470788806e-08, "loss": 0.2569, "step": 6453 }, { "epoch": 2.853573139719161, "grad_norm": 0.3770728811973308, "learning_rate": 7.318298077998054e-08, "loss": 0.2396, "step": 6454 }, { "epoch": 2.854015405594663, "grad_norm": 0.3879912677135586, "learning_rate": 7.27451706501281e-08, "loss": 0.2464, "step": 6455 }, { "epoch": 2.8544576714701657, "grad_norm": 0.32129913721065084, "learning_rate": 7.230866443418971e-08, "loss": 0.2405, "step": 6456 }, { "epoch": 2.8548999373456674, "grad_norm": 0.3662657361666361, "learning_rate": 7.187346224767855e-08, "loss": 0.271, "step": 6457 }, { "epoch": 2.85534220322117, "grad_norm": 0.38702085950477483, "learning_rate": 7.143956420576415e-08, "loss": 0.2656, "step": 6458 }, { "epoch": 2.855784469096672, "grad_norm": 0.41236923018269733, "learning_rate": 7.10069704232702e-08, "loss": 0.2797, "step": 6459 }, { "epoch": 2.856226734972174, "grad_norm": 0.3579317677731447, "learning_rate": 7.057568101467571e-08, "loss": 0.2425, "step": 6460 }, { "epoch": 2.8566690008476763, "grad_norm": 0.3475386348806716, "learning_rate": 7.01456960941127e-08, "loss": 0.2646, "step": 6461 }, { "epoch": 2.8571112667231784, "grad_norm": 0.3391546722647274, "learning_rate": 6.971701577537071e-08, "loss": 0.2625, "step": 6462 }, { "epoch": 2.8575535325986805, "grad_norm": 0.35174023903770185, "learning_rate": 6.928964017189232e-08, "loss": 0.3206, "step": 6463 }, { "epoch": 2.8579957984741826, "grad_norm": 0.35081379628738885, "learning_rate": 6.886356939677485e-08, "loss": 0.2583, "step": 6464 }, { "epoch": 2.8584380643496847, "grad_norm": 0.3340957622155315, "learning_rate": 6.843880356277033e-08, "loss": 0.2666, "step": 6465 }, { "epoch": 2.858880330225187, "grad_norm": 0.38606383928630783, "learning_rate": 6.801534278228661e-08, "loss": 0.3069, "step": 6466 }, { "epoch": 2.8593225961006894, "grad_norm": 0.3088863886973009, "learning_rate": 6.759318716738516e-08, "loss": 0.2422, "step": 6467 }, { "epoch": 2.8597648619761915, "grad_norm": 0.2899988206564555, "learning_rate": 6.717233682978164e-08, "loss": 0.3132, "step": 6468 }, { "epoch": 2.8602071278516936, "grad_norm": 0.33340439113435805, "learning_rate": 6.675279188084805e-08, "loss": 0.2736, "step": 6469 }, { "epoch": 2.8606493937271957, "grad_norm": 0.4020445460346603, "learning_rate": 6.633455243160891e-08, "loss": 0.2433, "step": 6470 }, { "epoch": 2.861091659602698, "grad_norm": 0.363872509618989, "learning_rate": 6.591761859274459e-08, "loss": 0.2497, "step": 6471 }, { "epoch": 2.8615339254782, "grad_norm": 0.39067284955017073, "learning_rate": 6.55019904745896e-08, "loss": 0.2902, "step": 6472 }, { "epoch": 2.861976191353702, "grad_norm": 0.3554400022828192, "learning_rate": 6.508766818713153e-08, "loss": 0.2828, "step": 6473 }, { "epoch": 2.862418457229204, "grad_norm": 0.27807376319901894, "learning_rate": 6.467465184001542e-08, "loss": 0.2609, "step": 6474 }, { "epoch": 2.8628607231047063, "grad_norm": 0.3725338939708873, "learning_rate": 6.426294154253831e-08, "loss": 0.2762, "step": 6475 }, { "epoch": 2.863302988980209, "grad_norm": 0.3678015917755088, "learning_rate": 6.385253740365249e-08, "loss": 0.2591, "step": 6476 }, { "epoch": 2.8637452548557105, "grad_norm": 0.29457492525811096, "learning_rate": 6.344343953196386e-08, "loss": 0.2601, "step": 6477 }, { "epoch": 2.864187520731213, "grad_norm": 0.34251990856171827, "learning_rate": 6.303564803573358e-08, "loss": 0.259, "step": 6478 }, { "epoch": 2.864629786606715, "grad_norm": 0.33911092624928235, "learning_rate": 6.262916302287703e-08, "loss": 0.2576, "step": 6479 }, { "epoch": 2.8650720524822173, "grad_norm": 0.347243002137254, "learning_rate": 6.222398460096257e-08, "loss": 0.2793, "step": 6480 }, { "epoch": 2.8655143183577194, "grad_norm": 0.3101426398797915, "learning_rate": 6.182011287721501e-08, "loss": 0.2466, "step": 6481 }, { "epoch": 2.8659565842332215, "grad_norm": 0.26293235973739804, "learning_rate": 6.141754795851107e-08, "loss": 0.2637, "step": 6482 }, { "epoch": 2.8663988501087236, "grad_norm": 0.37043996976782956, "learning_rate": 6.101628995138331e-08, "loss": 0.2674, "step": 6483 }, { "epoch": 2.8668411159842258, "grad_norm": 0.309328491615237, "learning_rate": 6.061633896201735e-08, "loss": 0.2621, "step": 6484 }, { "epoch": 2.867283381859728, "grad_norm": 0.3591035536810947, "learning_rate": 6.021769509625408e-08, "loss": 0.2914, "step": 6485 }, { "epoch": 2.86772564773523, "grad_norm": 0.2938075560221308, "learning_rate": 5.982035845958689e-08, "loss": 0.2798, "step": 6486 }, { "epoch": 2.8681679136107325, "grad_norm": 0.359635862332952, "learning_rate": 5.9424329157165004e-08, "loss": 0.2783, "step": 6487 }, { "epoch": 2.868610179486234, "grad_norm": 0.3224550618367726, "learning_rate": 5.9029607293790145e-08, "loss": 0.262, "step": 6488 }, { "epoch": 2.8690524453617368, "grad_norm": 0.29698188293107974, "learning_rate": 5.8636192973919315e-08, "loss": 0.2723, "step": 6489 }, { "epoch": 2.869494711237239, "grad_norm": 0.3312235133486596, "learning_rate": 5.8244086301662006e-08, "loss": 0.2932, "step": 6490 }, { "epoch": 2.869936977112741, "grad_norm": 0.3341651997335861, "learning_rate": 5.785328738078355e-08, "loss": 0.2636, "step": 6491 }, { "epoch": 2.870379242988243, "grad_norm": 0.3512816277236311, "learning_rate": 5.746379631470178e-08, "loss": 0.2634, "step": 6492 }, { "epoch": 2.870821508863745, "grad_norm": 0.320733316877436, "learning_rate": 5.707561320648924e-08, "loss": 0.2643, "step": 6493 }, { "epoch": 2.8712637747392473, "grad_norm": 0.4405057536554703, "learning_rate": 5.668873815887099e-08, "loss": 0.2773, "step": 6494 }, { "epoch": 2.8717060406147494, "grad_norm": 0.41381385573770546, "learning_rate": 5.6303171274227906e-08, "loss": 0.2607, "step": 6495 }, { "epoch": 2.872148306490252, "grad_norm": 0.3577412757475475, "learning_rate": 5.591891265459337e-08, "loss": 0.2889, "step": 6496 }, { "epoch": 2.8725905723657537, "grad_norm": 0.3998748497047009, "learning_rate": 5.553596240165438e-08, "loss": 0.2312, "step": 6497 }, { "epoch": 2.873032838241256, "grad_norm": 0.37362984615270495, "learning_rate": 5.515432061675263e-08, "loss": 0.2555, "step": 6498 }, { "epoch": 2.8734751041167583, "grad_norm": 0.3149363186657154, "learning_rate": 5.4773987400883466e-08, "loss": 0.269, "step": 6499 }, { "epoch": 2.8739173699922604, "grad_norm": 0.3329900781240375, "learning_rate": 5.43949628546947e-08, "loss": 0.2595, "step": 6500 }, { "epoch": 2.8743596358677626, "grad_norm": 0.3829660496424003, "learning_rate": 5.401724707848943e-08, "loss": 0.2846, "step": 6501 }, { "epoch": 2.8748019017432647, "grad_norm": 0.2731770208121049, "learning_rate": 5.364084017222382e-08, "loss": 0.2555, "step": 6502 }, { "epoch": 2.8752441676187668, "grad_norm": 0.3551861407724123, "learning_rate": 5.3265742235507624e-08, "loss": 0.2458, "step": 6503 }, { "epoch": 2.875686433494269, "grad_norm": 0.3232858439569926, "learning_rate": 5.2891953367602555e-08, "loss": 0.2495, "step": 6504 }, { "epoch": 2.876128699369771, "grad_norm": 0.3083455251393418, "learning_rate": 5.2519473667427264e-08, "loss": 0.2805, "step": 6505 }, { "epoch": 2.876570965245273, "grad_norm": 0.29038639613778694, "learning_rate": 5.214830323355124e-08, "loss": 0.2923, "step": 6506 }, { "epoch": 2.8770132311207757, "grad_norm": 0.33741109787106777, "learning_rate": 5.17784421641987e-08, "loss": 0.2929, "step": 6507 }, { "epoch": 2.8774554969962773, "grad_norm": 0.38383366044387085, "learning_rate": 5.1409890557246876e-08, "loss": 0.2761, "step": 6508 }, { "epoch": 2.87789776287178, "grad_norm": 0.5242543091527246, "learning_rate": 5.1042648510226664e-08, "loss": 0.2761, "step": 6509 }, { "epoch": 2.878340028747282, "grad_norm": 0.319828921575871, "learning_rate": 5.0676716120322545e-08, "loss": 0.2759, "step": 6510 }, { "epoch": 2.878782294622784, "grad_norm": 0.3504940954677107, "learning_rate": 5.03120934843726e-08, "loss": 0.2528, "step": 6511 }, { "epoch": 2.8792245604982862, "grad_norm": 0.3090863635983496, "learning_rate": 4.994878069886744e-08, "loss": 0.2622, "step": 6512 }, { "epoch": 2.8796668263737883, "grad_norm": 0.35199146322909847, "learning_rate": 4.9586777859951805e-08, "loss": 0.2852, "step": 6513 }, { "epoch": 2.8801090922492905, "grad_norm": 0.32541615006266844, "learning_rate": 4.922608506342408e-08, "loss": 0.2295, "step": 6514 }, { "epoch": 2.8805513581247926, "grad_norm": 0.325432530517888, "learning_rate": 4.8866702404734015e-08, "loss": 0.2498, "step": 6515 }, { "epoch": 2.8809936240002947, "grad_norm": 0.348968354276365, "learning_rate": 4.85086299789872e-08, "loss": 0.275, "step": 6516 }, { "epoch": 2.881435889875797, "grad_norm": 0.36079654229089103, "learning_rate": 4.815186788094173e-08, "loss": 0.2585, "step": 6517 }, { "epoch": 2.8818781557512994, "grad_norm": 0.43572479998773744, "learning_rate": 4.779641620500819e-08, "loss": 0.297, "step": 6518 }, { "epoch": 2.8823204216268015, "grad_norm": 0.3145648552010207, "learning_rate": 4.744227504525023e-08, "loss": 0.297, "step": 6519 }, { "epoch": 2.8827626875023036, "grad_norm": 0.3114123749271214, "learning_rate": 4.708944449538566e-08, "loss": 0.2581, "step": 6520 }, { "epoch": 2.8832049533778057, "grad_norm": 0.35110175852899195, "learning_rate": 4.673792464878535e-08, "loss": 0.2619, "step": 6521 }, { "epoch": 2.883647219253308, "grad_norm": 0.2763505941480592, "learning_rate": 4.638771559847322e-08, "loss": 0.2867, "step": 6522 }, { "epoch": 2.88408948512881, "grad_norm": 0.3836773676269336, "learning_rate": 4.603881743712513e-08, "loss": 0.3057, "step": 6523 }, { "epoch": 2.884531751004312, "grad_norm": 0.341283085346546, "learning_rate": 4.569123025707167e-08, "loss": 0.2697, "step": 6524 }, { "epoch": 2.884974016879814, "grad_norm": 0.35884523904558036, "learning_rate": 4.534495415029649e-08, "loss": 0.256, "step": 6525 }, { "epoch": 2.8854162827553163, "grad_norm": 0.3164860372697736, "learning_rate": 4.499998920843407e-08, "loss": 0.2631, "step": 6526 }, { "epoch": 2.885858548630819, "grad_norm": 0.34460848149738704, "learning_rate": 4.4656335522775266e-08, "loss": 0.2699, "step": 6527 }, { "epoch": 2.8863008145063205, "grad_norm": 0.3025657342583043, "learning_rate": 4.431399318426066e-08, "loss": 0.3116, "step": 6528 }, { "epoch": 2.886743080381823, "grad_norm": 0.3152039949861339, "learning_rate": 4.397296228348613e-08, "loss": 0.2802, "step": 6529 }, { "epoch": 2.887185346257325, "grad_norm": 0.4000061191494465, "learning_rate": 4.363324291069948e-08, "loss": 0.2449, "step": 6530 }, { "epoch": 2.8876276121328273, "grad_norm": 0.521813052021486, "learning_rate": 4.3294835155801575e-08, "loss": 0.2705, "step": 6531 }, { "epoch": 2.8880698780083294, "grad_norm": 0.26471610254133104, "learning_rate": 4.295773910834633e-08, "loss": 0.2548, "step": 6532 }, { "epoch": 2.8885121438838315, "grad_norm": 0.3563225806951841, "learning_rate": 4.262195485754072e-08, "loss": 0.2663, "step": 6533 }, { "epoch": 2.8889544097593336, "grad_norm": 0.3364439323390563, "learning_rate": 4.228748249224424e-08, "loss": 0.2773, "step": 6534 }, { "epoch": 2.8893966756348357, "grad_norm": 0.32900565973065793, "learning_rate": 4.195432210096828e-08, "loss": 0.2657, "step": 6535 }, { "epoch": 2.889838941510338, "grad_norm": 0.36870745457622495, "learning_rate": 4.1622473771879554e-08, "loss": 0.2536, "step": 6536 }, { "epoch": 2.89028120738584, "grad_norm": 0.36911749541122274, "learning_rate": 4.129193759279504e-08, "loss": 0.2569, "step": 6537 }, { "epoch": 2.8907234732613425, "grad_norm": 0.45102073713148616, "learning_rate": 4.096271365118643e-08, "loss": 0.304, "step": 6538 }, { "epoch": 2.891165739136844, "grad_norm": 0.3680764033893018, "learning_rate": 4.063480203417625e-08, "loss": 0.2656, "step": 6539 }, { "epoch": 2.8916080050123467, "grad_norm": 0.3562590062202638, "learning_rate": 4.0308202828541197e-08, "loss": 0.25, "step": 6540 }, { "epoch": 2.892050270887849, "grad_norm": 0.3742273701029764, "learning_rate": 3.9982916120709925e-08, "loss": 0.2612, "step": 6541 }, { "epoch": 2.892492536763351, "grad_norm": 0.3076660352149632, "learning_rate": 3.965894199676412e-08, "loss": 0.2265, "step": 6542 }, { "epoch": 2.892934802638853, "grad_norm": 0.35172252712138213, "learning_rate": 3.933628054243854e-08, "loss": 0.2704, "step": 6543 }, { "epoch": 2.893377068514355, "grad_norm": 0.29831067851463877, "learning_rate": 3.901493184311933e-08, "loss": 0.2891, "step": 6544 }, { "epoch": 2.8938193343898573, "grad_norm": 0.3541066738946007, "learning_rate": 3.869489598384624e-08, "loss": 0.2545, "step": 6545 }, { "epoch": 2.8942616002653594, "grad_norm": 0.33376535774713617, "learning_rate": 3.837617304931096e-08, "loss": 0.2749, "step": 6546 }, { "epoch": 2.894703866140862, "grad_norm": 0.3587126567782987, "learning_rate": 3.805876312385826e-08, "loss": 0.2356, "step": 6547 }, { "epoch": 2.8951461320163636, "grad_norm": 0.32838524613745496, "learning_rate": 3.7742666291485374e-08, "loss": 0.2458, "step": 6548 }, { "epoch": 2.895588397891866, "grad_norm": 0.4643058467537375, "learning_rate": 3.7427882635842624e-08, "loss": 0.2951, "step": 6549 }, { "epoch": 2.8960306637673683, "grad_norm": 0.3794913524411719, "learning_rate": 3.711441224023116e-08, "loss": 0.2835, "step": 6550 }, { "epoch": 2.8964729296428704, "grad_norm": 0.466226569862623, "learning_rate": 3.6802255187605165e-08, "loss": 0.256, "step": 6551 }, { "epoch": 2.8969151955183725, "grad_norm": 0.3444280019129499, "learning_rate": 3.649141156057301e-08, "loss": 0.2558, "step": 6552 }, { "epoch": 2.8973574613938746, "grad_norm": 0.3508725304639436, "learning_rate": 3.6181881441392784e-08, "loss": 0.2795, "step": 6553 }, { "epoch": 2.8977997272693767, "grad_norm": 0.35108436116331787, "learning_rate": 3.5873664911977855e-08, "loss": 0.2649, "step": 6554 }, { "epoch": 2.898241993144879, "grad_norm": 0.36225717698980936, "learning_rate": 3.5566762053891314e-08, "loss": 0.2773, "step": 6555 }, { "epoch": 2.898684259020381, "grad_norm": 0.3553649351625648, "learning_rate": 3.5261172948350405e-08, "loss": 0.2489, "step": 6556 }, { "epoch": 2.899126524895883, "grad_norm": 0.3358026278184951, "learning_rate": 3.495689767622323e-08, "loss": 0.2642, "step": 6557 }, { "epoch": 2.8995687907713856, "grad_norm": 0.31783048157247085, "learning_rate": 3.4653936318032046e-08, "loss": 0.2576, "step": 6558 }, { "epoch": 2.9000110566468873, "grad_norm": 0.3338241565438192, "learning_rate": 3.435228895394993e-08, "loss": 0.2762, "step": 6559 }, { "epoch": 2.90045332252239, "grad_norm": 0.3604300933350062, "learning_rate": 3.4051955663802484e-08, "loss": 0.2682, "step": 6560 }, { "epoch": 2.900895588397892, "grad_norm": 0.333867067937576, "learning_rate": 3.3752936527068347e-08, "loss": 0.26, "step": 6561 }, { "epoch": 2.901337854273394, "grad_norm": 0.2970996683218851, "learning_rate": 3.3455231622877e-08, "loss": 0.273, "step": 6562 }, { "epoch": 2.901780120148896, "grad_norm": 0.3527105014099231, "learning_rate": 3.31588410300121e-08, "loss": 0.2924, "step": 6563 }, { "epoch": 2.9022223860243983, "grad_norm": 0.36226637518174454, "learning_rate": 3.2863764826907563e-08, "loss": 0.3008, "step": 6564 }, { "epoch": 2.9026646518999004, "grad_norm": 0.41378274049790964, "learning_rate": 3.257000309165037e-08, "loss": 0.2473, "step": 6565 }, { "epoch": 2.9031069177754025, "grad_norm": 0.33945493567789015, "learning_rate": 3.227755590198001e-08, "loss": 0.248, "step": 6566 }, { "epoch": 2.9035491836509046, "grad_norm": 0.381327520884516, "learning_rate": 3.198642333528679e-08, "loss": 0.255, "step": 6567 }, { "epoch": 2.9039914495264068, "grad_norm": 0.28149430017819843, "learning_rate": 3.169660546861464e-08, "loss": 0.2839, "step": 6568 }, { "epoch": 2.9044337154019093, "grad_norm": 0.3329815432177287, "learning_rate": 3.140810237865888e-08, "loss": 0.2414, "step": 6569 }, { "epoch": 2.9048759812774114, "grad_norm": 0.26407033254246615, "learning_rate": 3.1120914141766214e-08, "loss": 0.2855, "step": 6570 }, { "epoch": 2.9053182471529135, "grad_norm": 0.48824348529856737, "learning_rate": 3.083504083393751e-08, "loss": 0.3346, "step": 6571 }, { "epoch": 2.9057605130284156, "grad_norm": 0.4937131843600548, "learning_rate": 3.055048253082282e-08, "loss": 0.2583, "step": 6572 }, { "epoch": 2.9062027789039178, "grad_norm": 0.3370753313421246, "learning_rate": 3.026723930772635e-08, "loss": 0.2774, "step": 6573 }, { "epoch": 2.90664504477942, "grad_norm": 0.5538303732753619, "learning_rate": 2.998531123960369e-08, "loss": 0.3062, "step": 6574 }, { "epoch": 2.907087310654922, "grad_norm": 0.33541921688495746, "learning_rate": 2.970469840106238e-08, "loss": 0.2677, "step": 6575 }, { "epoch": 2.907529576530424, "grad_norm": 0.3102829157074823, "learning_rate": 2.9425400866361365e-08, "loss": 0.2476, "step": 6576 }, { "epoch": 2.907971842405926, "grad_norm": 0.28947975769998274, "learning_rate": 2.9147418709412067e-08, "loss": 0.2822, "step": 6577 }, { "epoch": 2.9084141082814288, "grad_norm": 0.32612180486725334, "learning_rate": 2.8870752003777315e-08, "loss": 0.2554, "step": 6578 }, { "epoch": 2.9088563741569304, "grad_norm": 0.34076153787129837, "learning_rate": 2.8595400822672982e-08, "loss": 0.2812, "step": 6579 }, { "epoch": 2.909298640032433, "grad_norm": 0.33818797577452764, "learning_rate": 2.8321365238966338e-08, "loss": 0.2575, "step": 6580 }, { "epoch": 2.909740905907935, "grad_norm": 0.3569014476980577, "learning_rate": 2.8048645325174927e-08, "loss": 0.2882, "step": 6581 }, { "epoch": 2.910183171783437, "grad_norm": 0.2710868894204186, "learning_rate": 2.777724115347047e-08, "loss": 0.2342, "step": 6582 }, { "epoch": 2.9106254376589393, "grad_norm": 0.3312329921723016, "learning_rate": 2.7507152795675508e-08, "loss": 0.2519, "step": 6583 }, { "epoch": 2.9110677035344414, "grad_norm": 0.3438252994792464, "learning_rate": 2.7238380323263425e-08, "loss": 0.2744, "step": 6584 }, { "epoch": 2.9115099694099436, "grad_norm": 0.3524298225393348, "learning_rate": 2.6970923807361215e-08, "loss": 0.263, "step": 6585 }, { "epoch": 2.9119522352854457, "grad_norm": 0.3642970930354536, "learning_rate": 2.67047833187456e-08, "loss": 0.2554, "step": 6586 }, { "epoch": 2.912394501160948, "grad_norm": 0.34131240972266763, "learning_rate": 2.6439958927847455e-08, "loss": 0.2671, "step": 6587 }, { "epoch": 2.91283676703645, "grad_norm": 0.2988211164442172, "learning_rate": 2.617645070474739e-08, "loss": 0.2827, "step": 6588 }, { "epoch": 2.9132790329119524, "grad_norm": 0.40069511976049565, "learning_rate": 2.5914258719177965e-08, "loss": 0.3021, "step": 6589 }, { "epoch": 2.913721298787454, "grad_norm": 0.36418459289377575, "learning_rate": 2.5653383040524228e-08, "loss": 0.2737, "step": 6590 }, { "epoch": 2.9141635646629567, "grad_norm": 0.2529871186378897, "learning_rate": 2.539382373782262e-08, "loss": 0.2603, "step": 6591 }, { "epoch": 2.914605830538459, "grad_norm": 0.4384062896002929, "learning_rate": 2.513558087976098e-08, "loss": 0.2695, "step": 6592 }, { "epoch": 2.915048096413961, "grad_norm": 0.31259557333474347, "learning_rate": 2.4878654534678526e-08, "loss": 0.272, "step": 6593 }, { "epoch": 2.915490362289463, "grad_norm": 0.3213195075262928, "learning_rate": 2.4623044770566984e-08, "loss": 0.2659, "step": 6594 }, { "epoch": 2.915932628164965, "grad_norm": 0.2986921497769618, "learning_rate": 2.4368751655068356e-08, "loss": 0.2506, "step": 6595 }, { "epoch": 2.9163748940404672, "grad_norm": 0.3370244396761448, "learning_rate": 2.4115775255477704e-08, "loss": 0.2299, "step": 6596 }, { "epoch": 2.9168171599159693, "grad_norm": 0.3006425472527179, "learning_rate": 2.3864115638740913e-08, "loss": 0.2428, "step": 6597 }, { "epoch": 2.917259425791472, "grad_norm": 0.31766556836843346, "learning_rate": 2.3613772871454723e-08, "loss": 0.261, "step": 6598 }, { "epoch": 2.9177016916669736, "grad_norm": 0.29066220697019896, "learning_rate": 2.3364747019868905e-08, "loss": 0.2432, "step": 6599 }, { "epoch": 2.918143957542476, "grad_norm": 0.3437347096969442, "learning_rate": 2.3117038149882976e-08, "loss": 0.268, "step": 6600 }, { "epoch": 2.9185862234179782, "grad_norm": 0.3207585424020672, "learning_rate": 2.287064632705005e-08, "loss": 0.2716, "step": 6601 }, { "epoch": 2.9190284892934804, "grad_norm": 0.27982106391917155, "learning_rate": 2.2625571616572417e-08, "loss": 0.2591, "step": 6602 }, { "epoch": 2.9194707551689825, "grad_norm": 0.3411776242485417, "learning_rate": 2.238181408330542e-08, "loss": 0.2654, "step": 6603 }, { "epoch": 2.9199130210444846, "grad_norm": 0.35324608200349905, "learning_rate": 2.2139373791755237e-08, "loss": 0.2306, "step": 6604 }, { "epoch": 2.9203552869199867, "grad_norm": 0.3420291813823418, "learning_rate": 2.1898250806079992e-08, "loss": 0.2675, "step": 6605 }, { "epoch": 2.920797552795489, "grad_norm": 0.2883353311265066, "learning_rate": 2.1658445190088084e-08, "loss": 0.2275, "step": 6606 }, { "epoch": 2.921239818670991, "grad_norm": 0.3469950463166083, "learning_rate": 2.1419957007239867e-08, "loss": 0.2715, "step": 6607 }, { "epoch": 2.921682084546493, "grad_norm": 0.3402744903488134, "learning_rate": 2.1182786320648187e-08, "loss": 0.2805, "step": 6608 }, { "epoch": 2.9221243504219956, "grad_norm": 0.3455642435897468, "learning_rate": 2.094693319307506e-08, "loss": 0.2821, "step": 6609 }, { "epoch": 2.9225666162974973, "grad_norm": 0.6374178913761714, "learning_rate": 2.0712397686936115e-08, "loss": 0.311, "step": 6610 }, { "epoch": 2.923008882173, "grad_norm": 0.35842612026488135, "learning_rate": 2.047917986429615e-08, "loss": 0.266, "step": 6611 }, { "epoch": 2.923451148048502, "grad_norm": 0.35784771055315084, "learning_rate": 2.024727978687302e-08, "loss": 0.2661, "step": 6612 }, { "epoch": 2.923893413924004, "grad_norm": 0.2880671225656137, "learning_rate": 2.00166975160343e-08, "loss": 0.2722, "step": 6613 }, { "epoch": 2.924335679799506, "grad_norm": 0.3405172647488026, "learning_rate": 1.9787433112800624e-08, "loss": 0.288, "step": 6614 }, { "epoch": 2.9247779456750083, "grad_norm": 0.3852497455301218, "learning_rate": 1.955948663784235e-08, "loss": 0.2604, "step": 6615 }, { "epoch": 2.9252202115505104, "grad_norm": 0.3400533900041565, "learning_rate": 1.9332858151481228e-08, "loss": 0.2885, "step": 6616 }, { "epoch": 2.9256624774260125, "grad_norm": 0.344790845344811, "learning_rate": 1.9107547713691498e-08, "loss": 0.2621, "step": 6617 }, { "epoch": 2.9261047433015146, "grad_norm": 0.343412917127137, "learning_rate": 1.8883555384097697e-08, "loss": 0.2782, "step": 6618 }, { "epoch": 2.9265470091770167, "grad_norm": 0.38570537819392464, "learning_rate": 1.8660881221974625e-08, "loss": 0.2742, "step": 6619 }, { "epoch": 2.9269892750525193, "grad_norm": 0.29670657833339736, "learning_rate": 1.8439525286250704e-08, "loss": 0.2478, "step": 6620 }, { "epoch": 2.9274315409280214, "grad_norm": 0.31773367625040455, "learning_rate": 1.8219487635501853e-08, "loss": 0.2681, "step": 6621 }, { "epoch": 2.9278738068035235, "grad_norm": 0.29318242625041685, "learning_rate": 1.800076832795927e-08, "loss": 0.2514, "step": 6622 }, { "epoch": 2.9283160726790256, "grad_norm": 0.3096718948446534, "learning_rate": 1.778336742150222e-08, "loss": 0.281, "step": 6623 }, { "epoch": 2.9287583385545277, "grad_norm": 0.38947245652320417, "learning_rate": 1.7567284973663002e-08, "loss": 0.2633, "step": 6624 }, { "epoch": 2.92920060443003, "grad_norm": 0.35286370515974685, "learning_rate": 1.7352521041623106e-08, "loss": 0.2358, "step": 6625 }, { "epoch": 2.929642870305532, "grad_norm": 0.41413656101121915, "learning_rate": 1.713907568221651e-08, "loss": 0.2538, "step": 6626 }, { "epoch": 2.930085136181034, "grad_norm": 0.340909797092182, "learning_rate": 1.692694895192859e-08, "loss": 0.2703, "step": 6627 }, { "epoch": 2.930527402056536, "grad_norm": 0.3343967085222334, "learning_rate": 1.6716140906893885e-08, "loss": 0.2564, "step": 6628 }, { "epoch": 2.9309696679320387, "grad_norm": 0.30572214802800446, "learning_rate": 1.650665160289999e-08, "loss": 0.2805, "step": 6629 }, { "epoch": 2.9314119338075404, "grad_norm": 0.4277690819056976, "learning_rate": 1.629848109538479e-08, "loss": 0.2764, "step": 6630 }, { "epoch": 2.931854199683043, "grad_norm": 0.3863747908869061, "learning_rate": 1.609162943943643e-08, "loss": 0.2655, "step": 6631 }, { "epoch": 2.932296465558545, "grad_norm": 0.4598212006717347, "learning_rate": 1.588609668979446e-08, "loss": 0.2799, "step": 6632 }, { "epoch": 2.932738731434047, "grad_norm": 0.3038682693108277, "learning_rate": 1.5681882900850932e-08, "loss": 0.2595, "step": 6633 }, { "epoch": 2.9331809973095493, "grad_norm": 0.30135867428282287, "learning_rate": 1.5478988126645943e-08, "loss": 0.2586, "step": 6634 }, { "epoch": 2.9336232631850514, "grad_norm": 0.2854760749775688, "learning_rate": 1.527741242087377e-08, "loss": 0.2742, "step": 6635 }, { "epoch": 2.9340655290605535, "grad_norm": 0.30668962825373747, "learning_rate": 1.5077155836876744e-08, "loss": 0.2498, "step": 6636 }, { "epoch": 2.9345077949360556, "grad_norm": 0.36061233787927527, "learning_rate": 1.4878218427649138e-08, "loss": 0.266, "step": 6637 }, { "epoch": 2.9349500608115577, "grad_norm": 0.3677533122204607, "learning_rate": 1.4680600245837728e-08, "loss": 0.255, "step": 6638 }, { "epoch": 2.93539232668706, "grad_norm": 0.33631020162362774, "learning_rate": 1.448430134373735e-08, "loss": 0.263, "step": 6639 }, { "epoch": 2.9358345925625624, "grad_norm": 0.3863742902257911, "learning_rate": 1.4289321773296449e-08, "loss": 0.2728, "step": 6640 }, { "epoch": 2.936276858438064, "grad_norm": 0.382198512018477, "learning_rate": 1.4095661586112086e-08, "loss": 0.2516, "step": 6641 }, { "epoch": 2.9367191243135666, "grad_norm": 0.25304982091734307, "learning_rate": 1.3903320833433264e-08, "loss": 0.2442, "step": 6642 }, { "epoch": 2.9371613901890687, "grad_norm": 0.3573544828437987, "learning_rate": 1.371229956616038e-08, "loss": 0.2527, "step": 6643 }, { "epoch": 2.937603656064571, "grad_norm": 0.3469485367258278, "learning_rate": 1.3522597834842999e-08, "loss": 0.2632, "step": 6644 }, { "epoch": 2.938045921940073, "grad_norm": 0.28063612532160465, "learning_rate": 1.3334215689682627e-08, "loss": 0.2698, "step": 6645 }, { "epoch": 2.938488187815575, "grad_norm": 0.31459946184715964, "learning_rate": 1.3147153180532168e-08, "loss": 0.2803, "step": 6646 }, { "epoch": 2.938930453691077, "grad_norm": 0.31273239930661134, "learning_rate": 1.2961410356893689e-08, "loss": 0.2628, "step": 6647 }, { "epoch": 2.9393727195665793, "grad_norm": 0.33985237805321683, "learning_rate": 1.2776987267921203e-08, "loss": 0.2586, "step": 6648 }, { "epoch": 2.939814985442082, "grad_norm": 0.32198409993026467, "learning_rate": 1.2593883962419007e-08, "loss": 0.2336, "step": 6649 }, { "epoch": 2.9402572513175835, "grad_norm": 0.3614346179888934, "learning_rate": 1.2412100488842226e-08, "loss": 0.2789, "step": 6650 }, { "epoch": 2.940699517193086, "grad_norm": 0.3877373255955751, "learning_rate": 1.2231636895297382e-08, "loss": 0.2643, "step": 6651 }, { "epoch": 2.941141783068588, "grad_norm": 0.4178650895873785, "learning_rate": 1.2052493229540163e-08, "loss": 0.2657, "step": 6652 }, { "epoch": 2.9415840489440903, "grad_norm": 0.3669771418849551, "learning_rate": 1.1874669538978757e-08, "loss": 0.2878, "step": 6653 }, { "epoch": 2.9420263148195924, "grad_norm": 0.33377742943116623, "learning_rate": 1.1698165870669964e-08, "loss": 0.2825, "step": 6654 }, { "epoch": 2.9424685806950945, "grad_norm": 0.39488632596104983, "learning_rate": 1.1522982271323646e-08, "loss": 0.2533, "step": 6655 }, { "epoch": 2.9429108465705967, "grad_norm": 0.30251878322214415, "learning_rate": 1.134911878729883e-08, "loss": 0.2628, "step": 6656 }, { "epoch": 2.9433531124460988, "grad_norm": 0.296896658152139, "learning_rate": 1.1176575464605377e-08, "loss": 0.2693, "step": 6657 }, { "epoch": 2.943795378321601, "grad_norm": 0.3751995783273434, "learning_rate": 1.1005352348903986e-08, "loss": 0.256, "step": 6658 }, { "epoch": 2.944237644197103, "grad_norm": 0.31684662082885723, "learning_rate": 1.0835449485505634e-08, "loss": 0.2729, "step": 6659 }, { "epoch": 2.9446799100726055, "grad_norm": 0.3340974397647608, "learning_rate": 1.0666866919372687e-08, "loss": 0.274, "step": 6660 }, { "epoch": 2.945122175948107, "grad_norm": 0.44581838298155235, "learning_rate": 1.0499604695117793e-08, "loss": 0.2548, "step": 6661 }, { "epoch": 2.9455644418236098, "grad_norm": 0.30553275678828834, "learning_rate": 1.0333662857003324e-08, "loss": 0.2641, "step": 6662 }, { "epoch": 2.946006707699112, "grad_norm": 0.33821875818171204, "learning_rate": 1.016904144894304e-08, "loss": 0.2406, "step": 6663 }, { "epoch": 2.946448973574614, "grad_norm": 0.6163746925448083, "learning_rate": 1.0005740514501539e-08, "loss": 0.2479, "step": 6664 }, { "epoch": 2.946891239450116, "grad_norm": 0.3516154741686719, "learning_rate": 9.843760096894251e-09, "loss": 0.2554, "step": 6665 }, { "epoch": 2.947333505325618, "grad_norm": 0.3201809583444193, "learning_rate": 9.683100238985221e-09, "loss": 0.2582, "step": 6666 }, { "epoch": 2.9477757712011203, "grad_norm": 0.4043160129371303, "learning_rate": 9.523760983291552e-09, "loss": 0.2716, "step": 6667 }, { "epoch": 2.9482180370766224, "grad_norm": 0.34941842910616966, "learning_rate": 9.365742371978403e-09, "loss": 0.273, "step": 6668 }, { "epoch": 2.9486603029521246, "grad_norm": 0.29691319192674437, "learning_rate": 9.209044446863991e-09, "loss": 0.2648, "step": 6669 }, { "epoch": 2.9491025688276267, "grad_norm": 0.2953918085945108, "learning_rate": 9.053667249414588e-09, "loss": 0.2734, "step": 6670 }, { "epoch": 2.9495448347031292, "grad_norm": 0.34053728945388806, "learning_rate": 8.899610820748972e-09, "loss": 0.2557, "step": 6671 }, { "epoch": 2.9499871005786313, "grad_norm": 0.28499049511568486, "learning_rate": 8.746875201635085e-09, "loss": 0.2755, "step": 6672 }, { "epoch": 2.9504293664541335, "grad_norm": 0.38275141470339547, "learning_rate": 8.595460432492265e-09, "loss": 0.2421, "step": 6673 }, { "epoch": 2.9508716323296356, "grad_norm": 0.337202604514784, "learning_rate": 8.445366553389011e-09, "loss": 0.233, "step": 6674 }, { "epoch": 2.9513138982051377, "grad_norm": 0.42649715550685696, "learning_rate": 8.296593604046332e-09, "loss": 0.2719, "step": 6675 }, { "epoch": 2.95175616408064, "grad_norm": 0.29903168473704944, "learning_rate": 8.14914162383329e-09, "loss": 0.2526, "step": 6676 }, { "epoch": 2.952198429956142, "grad_norm": 0.3130393222299482, "learning_rate": 8.003010651770893e-09, "loss": 0.2607, "step": 6677 }, { "epoch": 2.952640695831644, "grad_norm": 0.2984183040965231, "learning_rate": 7.85820072653043e-09, "loss": 0.2815, "step": 6678 }, { "epoch": 2.953082961707146, "grad_norm": 0.40731537450471256, "learning_rate": 7.714711886433468e-09, "loss": 0.3011, "step": 6679 }, { "epoch": 2.9535252275826487, "grad_norm": 0.3407298376928531, "learning_rate": 7.572544169451301e-09, "loss": 0.2561, "step": 6680 }, { "epoch": 2.9539674934581504, "grad_norm": 0.30676018491943713, "learning_rate": 7.431697613207167e-09, "loss": 0.2778, "step": 6681 }, { "epoch": 2.954409759333653, "grad_norm": 0.34927639021464796, "learning_rate": 7.2921722549729176e-09, "loss": 0.2718, "step": 6682 }, { "epoch": 2.954852025209155, "grad_norm": 0.32650563863487025, "learning_rate": 7.153968131671241e-09, "loss": 0.2504, "step": 6683 }, { "epoch": 2.955294291084657, "grad_norm": 0.4448080939432674, "learning_rate": 7.017085279876213e-09, "loss": 0.2779, "step": 6684 }, { "epoch": 2.9557365569601592, "grad_norm": 0.36110039869331667, "learning_rate": 6.881523735811635e-09, "loss": 0.2503, "step": 6685 }, { "epoch": 2.9561788228356614, "grad_norm": 0.3578057524272253, "learning_rate": 6.747283535351034e-09, "loss": 0.2849, "step": 6686 }, { "epoch": 2.9566210887111635, "grad_norm": 0.2965436104060527, "learning_rate": 6.614364714018773e-09, "loss": 0.2484, "step": 6687 }, { "epoch": 2.9570633545866656, "grad_norm": 0.3479154821186613, "learning_rate": 6.4827673069900454e-09, "loss": 0.2352, "step": 6688 }, { "epoch": 2.9575056204621677, "grad_norm": 0.3325292520802151, "learning_rate": 6.352491349089774e-09, "loss": 0.2575, "step": 6689 }, { "epoch": 2.95794788633767, "grad_norm": 0.31581981868081327, "learning_rate": 6.223536874792602e-09, "loss": 0.2725, "step": 6690 }, { "epoch": 2.9583901522131724, "grad_norm": 0.3212338277086899, "learning_rate": 6.095903918225121e-09, "loss": 0.2735, "step": 6691 }, { "epoch": 2.958832418088674, "grad_norm": 0.34421969848472195, "learning_rate": 5.969592513162536e-09, "loss": 0.2541, "step": 6692 }, { "epoch": 2.9592746839641766, "grad_norm": 0.2799193804187069, "learning_rate": 5.844602693031443e-09, "loss": 0.25, "step": 6693 }, { "epoch": 2.9597169498396787, "grad_norm": 0.31034326780915816, "learning_rate": 5.720934490907604e-09, "loss": 0.261, "step": 6694 }, { "epoch": 2.960159215715181, "grad_norm": 0.3056282525909562, "learning_rate": 5.5985879395181745e-09, "loss": 0.2832, "step": 6695 }, { "epoch": 2.960601481590683, "grad_norm": 0.3962082500084544, "learning_rate": 5.477563071240033e-09, "loss": 0.286, "step": 6696 }, { "epoch": 2.961043747466185, "grad_norm": 0.3901798464451184, "learning_rate": 5.357859918100339e-09, "loss": 0.2613, "step": 6697 }, { "epoch": 2.961486013341687, "grad_norm": 0.3368458416267354, "learning_rate": 5.23947851177653e-09, "loss": 0.2734, "step": 6698 }, { "epoch": 2.9619282792171893, "grad_norm": 0.390032480838634, "learning_rate": 5.12241888359577e-09, "loss": 0.2533, "step": 6699 }, { "epoch": 2.962370545092692, "grad_norm": 0.30287071165235535, "learning_rate": 5.006681064536611e-09, "loss": 0.275, "step": 6700 }, { "epoch": 2.9628128109681935, "grad_norm": 0.5200212377432121, "learning_rate": 4.892265085226777e-09, "loss": 0.2634, "step": 6701 }, { "epoch": 2.963255076843696, "grad_norm": 0.35598326148993287, "learning_rate": 4.779170975944824e-09, "loss": 0.2533, "step": 6702 }, { "epoch": 2.963697342719198, "grad_norm": 0.3483745416809276, "learning_rate": 4.667398766618481e-09, "loss": 0.2417, "step": 6703 }, { "epoch": 2.9641396085947003, "grad_norm": 0.461954331861641, "learning_rate": 4.556948486826862e-09, "loss": 0.2557, "step": 6704 }, { "epoch": 2.9645818744702024, "grad_norm": 0.33716127576341115, "learning_rate": 4.447820165798256e-09, "loss": 0.2556, "step": 6705 }, { "epoch": 2.9650241403457045, "grad_norm": 0.3629205354369802, "learning_rate": 4.3400138324123375e-09, "loss": 0.2632, "step": 6706 }, { "epoch": 2.9654664062212066, "grad_norm": 0.27214859170666117, "learning_rate": 4.233529515197954e-09, "loss": 0.2434, "step": 6707 }, { "epoch": 2.9659086720967087, "grad_norm": 0.31764571229320315, "learning_rate": 4.1283672423336754e-09, "loss": 0.2589, "step": 6708 }, { "epoch": 2.966350937972211, "grad_norm": 0.31735481620067185, "learning_rate": 4.024527041650017e-09, "loss": 0.2462, "step": 6709 }, { "epoch": 2.966793203847713, "grad_norm": 0.370931089463172, "learning_rate": 3.9220089406255545e-09, "loss": 0.251, "step": 6710 }, { "epoch": 2.9672354697232155, "grad_norm": 0.3568080901711006, "learning_rate": 3.820812966390808e-09, "loss": 0.2463, "step": 6711 }, { "epoch": 2.967677735598717, "grad_norm": 0.326121157679466, "learning_rate": 3.7209391457249113e-09, "loss": 0.2521, "step": 6712 }, { "epoch": 2.9681200014742197, "grad_norm": 0.49404943405998514, "learning_rate": 3.622387505058389e-09, "loss": 0.2914, "step": 6713 }, { "epoch": 2.968562267349722, "grad_norm": 0.3417292221674047, "learning_rate": 3.52515807047038e-09, "loss": 0.2421, "step": 6714 }, { "epoch": 2.969004533225224, "grad_norm": 0.336979796604691, "learning_rate": 3.4292508676914136e-09, "loss": 0.2521, "step": 6715 }, { "epoch": 2.969446799100726, "grad_norm": 0.3461554994037243, "learning_rate": 3.334665922101743e-09, "loss": 0.275, "step": 6716 }, { "epoch": 2.969889064976228, "grad_norm": 0.4190097646333429, "learning_rate": 3.241403258731901e-09, "loss": 0.2564, "step": 6717 }, { "epoch": 2.9703313308517303, "grad_norm": 0.32253639866375255, "learning_rate": 3.149462902262146e-09, "loss": 0.2505, "step": 6718 }, { "epoch": 2.9707735967272324, "grad_norm": 0.3936279327177362, "learning_rate": 3.05884487702246e-09, "loss": 0.2249, "step": 6719 }, { "epoch": 2.9712158626027345, "grad_norm": 0.3126482097878076, "learning_rate": 2.9695492069936604e-09, "loss": 0.2629, "step": 6720 }, { "epoch": 2.9716581284782366, "grad_norm": 0.38328355325908203, "learning_rate": 2.8815759158068448e-09, "loss": 0.2787, "step": 6721 }, { "epoch": 2.972100394353739, "grad_norm": 0.3504201676771564, "learning_rate": 2.794925026741724e-09, "loss": 0.259, "step": 6722 }, { "epoch": 2.9725426602292413, "grad_norm": 0.35207488289967337, "learning_rate": 2.7095965627299547e-09, "loss": 0.27, "step": 6723 }, { "epoch": 2.9729849261047434, "grad_norm": 0.2946484886967392, "learning_rate": 2.6255905463512534e-09, "loss": 0.2763, "step": 6724 }, { "epoch": 2.9734271919802455, "grad_norm": 0.3752006618792594, "learning_rate": 2.542906999836725e-09, "loss": 0.2738, "step": 6725 }, { "epoch": 2.9738694578557476, "grad_norm": 0.42127081941625505, "learning_rate": 2.461545945067756e-09, "loss": 0.2731, "step": 6726 }, { "epoch": 2.9743117237312497, "grad_norm": 0.32066567224791614, "learning_rate": 2.381507403574346e-09, "loss": 0.276, "step": 6727 }, { "epoch": 2.974753989606752, "grad_norm": 0.3977441887091466, "learning_rate": 2.3027913965373294e-09, "loss": 0.2723, "step": 6728 }, { "epoch": 2.975196255482254, "grad_norm": 0.30328229451215855, "learning_rate": 2.2253979447883766e-09, "loss": 0.2718, "step": 6729 }, { "epoch": 2.975638521357756, "grad_norm": 0.38030121819285345, "learning_rate": 2.149327068807772e-09, "loss": 0.2557, "step": 6730 }, { "epoch": 2.9760807872332586, "grad_norm": 0.2972095105063446, "learning_rate": 2.0745787887266335e-09, "loss": 0.2774, "step": 6731 }, { "epoch": 2.9765230531087603, "grad_norm": 0.4011562831532708, "learning_rate": 2.00115312432525e-09, "loss": 0.2792, "step": 6732 }, { "epoch": 2.976965318984263, "grad_norm": 0.3821123715563406, "learning_rate": 1.929050095035301e-09, "loss": 0.2779, "step": 6733 }, { "epoch": 2.977407584859765, "grad_norm": 0.3600299552392019, "learning_rate": 1.858269719937078e-09, "loss": 0.269, "step": 6734 }, { "epoch": 2.977849850735267, "grad_norm": 0.3292472498650254, "learning_rate": 1.7888120177611545e-09, "loss": 0.245, "step": 6735 }, { "epoch": 2.978292116610769, "grad_norm": 0.2979011567509205, "learning_rate": 1.7206770068889377e-09, "loss": 0.2496, "step": 6736 }, { "epoch": 2.9787343824862713, "grad_norm": 0.35646619862093276, "learning_rate": 1.6538647053510049e-09, "loss": 0.2421, "step": 6737 }, { "epoch": 2.9791766483617734, "grad_norm": 0.3657886560665196, "learning_rate": 1.5883751308282124e-09, "loss": 0.2575, "step": 6738 }, { "epoch": 2.9796189142372755, "grad_norm": 0.3354304553372704, "learning_rate": 1.524208300650587e-09, "loss": 0.269, "step": 6739 }, { "epoch": 2.9800611801127777, "grad_norm": 0.38827410377093624, "learning_rate": 1.4613642317995447e-09, "loss": 0.2715, "step": 6740 }, { "epoch": 2.9805034459882798, "grad_norm": 0.3671133927502311, "learning_rate": 1.3998429409056712e-09, "loss": 0.2882, "step": 6741 }, { "epoch": 2.9809457118637823, "grad_norm": 0.33585112926151495, "learning_rate": 1.3396444442487223e-09, "loss": 0.2613, "step": 6742 }, { "epoch": 2.981387977739284, "grad_norm": 0.27970978616638686, "learning_rate": 1.2807687577598427e-09, "loss": 0.2741, "step": 6743 }, { "epoch": 2.9818302436147865, "grad_norm": 0.3337129537538099, "learning_rate": 1.2232158970199025e-09, "loss": 0.2993, "step": 6744 }, { "epoch": 2.9822725094902887, "grad_norm": 0.3055900724113713, "learning_rate": 1.1669858772583863e-09, "loss": 0.2966, "step": 6745 }, { "epoch": 2.9827147753657908, "grad_norm": 0.40336350406533394, "learning_rate": 1.1120787133561683e-09, "loss": 0.2378, "step": 6746 }, { "epoch": 2.983157041241293, "grad_norm": 0.27057057994514133, "learning_rate": 1.0584944198432923e-09, "loss": 0.2495, "step": 6747 }, { "epoch": 2.983599307116795, "grad_norm": 0.338844772281409, "learning_rate": 1.0062330108995266e-09, "loss": 0.2548, "step": 6748 }, { "epoch": 2.984041572992297, "grad_norm": 0.3357407015976349, "learning_rate": 9.552945003560298e-10, "loss": 0.3126, "step": 6749 }, { "epoch": 2.984483838867799, "grad_norm": 0.36132074054584173, "learning_rate": 9.056789016920197e-10, "loss": 0.2852, "step": 6750 }, { "epoch": 2.9849261047433018, "grad_norm": 0.3169463887354926, "learning_rate": 8.573862280369937e-10, "loss": 0.2423, "step": 6751 }, { "epoch": 2.9853683706188034, "grad_norm": 0.30785759727576034, "learning_rate": 8.104164921718394e-10, "loss": 0.2799, "step": 6752 }, { "epoch": 2.985810636494306, "grad_norm": 0.3108102693738211, "learning_rate": 7.647697065260584e-10, "loss": 0.2762, "step": 6753 }, { "epoch": 2.986252902369808, "grad_norm": 0.30961317631645774, "learning_rate": 7.204458831788774e-10, "loss": 0.2684, "step": 6754 }, { "epoch": 2.9866951682453102, "grad_norm": 0.37534042313293176, "learning_rate": 6.774450338598026e-10, "loss": 0.2572, "step": 6755 }, { "epoch": 2.9871374341208123, "grad_norm": 0.39593173194599596, "learning_rate": 6.357671699486201e-10, "loss": 0.2581, "step": 6756 }, { "epoch": 2.9875796999963145, "grad_norm": 0.3030866314985845, "learning_rate": 5.954123024742853e-10, "loss": 0.2742, "step": 6757 }, { "epoch": 2.9880219658718166, "grad_norm": 0.43568594516252074, "learning_rate": 5.563804421165886e-10, "loss": 0.2579, "step": 6758 }, { "epoch": 2.9884642317473187, "grad_norm": 0.4596531017751775, "learning_rate": 5.1867159920449e-10, "loss": 0.2962, "step": 6759 }, { "epoch": 2.988906497622821, "grad_norm": 0.33515777546728437, "learning_rate": 4.822857837166739e-10, "loss": 0.2904, "step": 6760 }, { "epoch": 2.989348763498323, "grad_norm": 0.44771851429718657, "learning_rate": 4.472230052821047e-10, "loss": 0.2598, "step": 6761 }, { "epoch": 2.9897910293738255, "grad_norm": 0.32030699779335575, "learning_rate": 4.1348327317947133e-10, "loss": 0.2582, "step": 6762 }, { "epoch": 2.990233295249327, "grad_norm": 0.3042322817057445, "learning_rate": 3.810665963377425e-10, "loss": 0.2411, "step": 6763 }, { "epoch": 2.9906755611248297, "grad_norm": 0.35723466647613167, "learning_rate": 3.4997298333505623e-10, "loss": 0.2799, "step": 6764 }, { "epoch": 2.991117827000332, "grad_norm": 0.3173519011231033, "learning_rate": 3.202024424003858e-10, "loss": 0.2523, "step": 6765 }, { "epoch": 2.991560092875834, "grad_norm": 0.29690053524561927, "learning_rate": 2.917549814113185e-10, "loss": 0.2696, "step": 6766 }, { "epoch": 2.992002358751336, "grad_norm": 0.304666554982515, "learning_rate": 2.6463060789683195e-10, "loss": 0.2822, "step": 6767 }, { "epoch": 2.992444624626838, "grad_norm": 0.3385833174856182, "learning_rate": 2.388293290339627e-10, "loss": 0.2805, "step": 6768 }, { "epoch": 2.9928868905023402, "grad_norm": 0.31474032211311354, "learning_rate": 2.1435115165113762e-10, "loss": 0.2442, "step": 6769 }, { "epoch": 2.9933291563778424, "grad_norm": 0.31739404581938613, "learning_rate": 1.9119608222595289e-10, "loss": 0.251, "step": 6770 }, { "epoch": 2.9937714222533445, "grad_norm": 0.33902877720463837, "learning_rate": 1.6936412688572935e-10, "loss": 0.2655, "step": 6771 }, { "epoch": 2.9942136881288466, "grad_norm": 0.3227093120025983, "learning_rate": 1.4885529140862277e-10, "loss": 0.2713, "step": 6772 }, { "epoch": 2.994655954004349, "grad_norm": 0.3738051908436808, "learning_rate": 1.2966958122140328e-10, "loss": 0.2309, "step": 6773 }, { "epoch": 2.9950982198798513, "grad_norm": 0.34260696630407184, "learning_rate": 1.1180700140112077e-10, "loss": 0.2543, "step": 6774 }, { "epoch": 2.9955404857553534, "grad_norm": 0.33424487992386437, "learning_rate": 9.52675566751049e-11, "loss": 0.2673, "step": 6775 }, { "epoch": 2.9959827516308555, "grad_norm": 0.3121472994031435, "learning_rate": 8.005125142040992e-11, "loss": 0.2397, "step": 6776 }, { "epoch": 2.9964250175063576, "grad_norm": 0.3582939827304444, "learning_rate": 6.615808966325965e-11, "loss": 0.254, "step": 6777 }, { "epoch": 2.9968672833818597, "grad_norm": 0.39139257747347717, "learning_rate": 5.358807508071273e-11, "loss": 0.2976, "step": 6778 }, { "epoch": 2.997309549257362, "grad_norm": 0.38434552986920734, "learning_rate": 4.234121099899735e-11, "loss": 0.2641, "step": 6779 }, { "epoch": 2.997751815132864, "grad_norm": 0.3497755002024062, "learning_rate": 3.241750039462144e-11, "loss": 0.2715, "step": 6780 }, { "epoch": 2.998194081008366, "grad_norm": 0.34824252090740904, "learning_rate": 2.3816945893262445e-11, "loss": 0.2675, "step": 6781 }, { "epoch": 2.9986363468838686, "grad_norm": 0.33663982209279864, "learning_rate": 1.6539549771432685e-11, "loss": 0.2796, "step": 6782 }, { "epoch": 2.9990786127593703, "grad_norm": 0.46890385334285983, "learning_rate": 1.0585313954813991e-11, "loss": 0.2673, "step": 6783 }, { "epoch": 2.999520878634873, "grad_norm": 0.3609659153239682, "learning_rate": 5.95424001881284e-12, "loss": 0.2707, "step": 6784 }, { "epoch": 2.999963144510375, "grad_norm": 0.3327035684702898, "learning_rate": 2.6463291891154486e-12, "loss": 0.2917, "step": 6785 }, { "epoch": 3.0, "grad_norm": 0.3327035684702898, "learning_rate": 6.615823411326716e-13, "loss": 0.2967, "step": 6786 }, { "epoch": 3.0, "step": 6786, "total_flos": 2307529338208256.0, "train_loss": 0.3301271531536556, "train_runtime": 285378.9111, "train_samples_per_second": 1.141, "train_steps_per_second": 0.024 } ], "logging_steps": 1, "max_steps": 6786, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 5000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 2307529338208256.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }