{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0032,
      "grad_norm": NaN,
      "learning_rate": 9.999948122981575e-05,
      "loss": 1.0507,
      "step": 32
    },
    {
      "epoch": 0.0064,
      "grad_norm": 1.0899442434310913,
      "learning_rate": 9.999770471768777e-05,
      "loss": 1.016,
      "step": 64
    },
    {
      "epoch": 0.0096,
      "grad_norm": 2.957014799118042,
      "learning_rate": 9.999466495684926e-05,
      "loss": 0.9928,
      "step": 96
    },
    {
      "epoch": 0.0128,
      "grad_norm": 1.3575732707977295,
      "learning_rate": 9.999036202410325e-05,
      "loss": 0.8757,
      "step": 128
    },
    {
      "epoch": 0.016,
      "grad_norm": 1.0611408948898315,
      "learning_rate": 9.998498908285819e-05,
      "loss": 0.8615,
      "step": 160
    },
    {
      "epoch": 0.0192,
      "grad_norm": 1.382992148399353,
      "learning_rate": 9.997819962824957e-05,
      "loss": 0.8216,
      "step": 192
    },
    {
      "epoch": 0.0224,
      "grad_norm": 2.863276481628418,
      "learning_rate": 9.997014741774866e-05,
      "loss": 0.7406,
      "step": 224
    },
    {
      "epoch": 0.0256,
      "grad_norm": 1.1629211902618408,
      "learning_rate": 9.996083265480365e-05,
      "loss": 0.8171,
      "step": 256
    },
    {
      "epoch": 0.0288,
      "grad_norm": 2.2264232635498047,
      "learning_rate": 9.995025557476261e-05,
      "loss": 0.8835,
      "step": 288
    },
    {
      "epoch": 0.032,
      "grad_norm": 1.7896003723144531,
      "learning_rate": 9.993841644486747e-05,
      "loss": 0.7303,
      "step": 320
    },
    {
      "epoch": 0.0352,
      "grad_norm": 1.403260350227356,
      "learning_rate": 9.992531556424726e-05,
      "loss": 0.7384,
      "step": 352
    },
    {
      "epoch": 0.0384,
      "grad_norm": 2.308896780014038,
      "learning_rate": 9.99109532639106e-05,
      "loss": 0.8211,
      "step": 384
    },
    {
      "epoch": 0.0416,
      "grad_norm": 1.282929539680481,
      "learning_rate": 9.989532990673728e-05,
      "loss": 0.7211,
      "step": 416
    },
    {
      "epoch": 0.0448,
      "grad_norm": 2.4921414852142334,
      "learning_rate": 9.987844588746915e-05,
      "loss": 0.8204,
      "step": 448
    },
    {
      "epoch": 0.048,
      "grad_norm": 1.3490195274353027,
      "learning_rate": 9.986030163270011e-05,
      "loss": 0.7623,
      "step": 480
    },
    {
      "epoch": 0.0512,
      "grad_norm": 1.436516523361206,
      "learning_rate": 9.98408976008653e-05,
      "loss": 0.7981,
      "step": 512
    },
    {
      "epoch": 0.0544,
      "grad_norm": 2.3144304752349854,
      "learning_rate": 9.982023428222962e-05,
      "loss": 0.7422,
      "step": 544
    },
    {
      "epoch": 0.0576,
      "grad_norm": 1.2702479362487793,
      "learning_rate": 9.979831219887525e-05,
      "loss": 0.8107,
      "step": 576
    },
    {
      "epoch": 0.0608,
      "grad_norm": 3.110814332962036,
      "learning_rate": 9.977513190468848e-05,
      "loss": 0.8395,
      "step": 608
    },
    {
      "epoch": 0.064,
      "grad_norm": 4.934881687164307,
      "learning_rate": 9.975069398534574e-05,
      "loss": 0.8456,
      "step": 640
    },
    {
      "epoch": 0.0672,
      "grad_norm": 1.5248093605041504,
      "learning_rate": 9.972499905829875e-05,
      "loss": 0.7604,
      "step": 672
    },
    {
      "epoch": 0.0704,
      "grad_norm": 1.5269616842269897,
      "learning_rate": 9.9698047772759e-05,
      "loss": 0.7557,
      "step": 704
    },
    {
      "epoch": 0.0736,
      "grad_norm": 1.523474097251892,
      "learning_rate": 9.966984080968128e-05,
      "loss": 0.7622,
      "step": 736
    },
    {
      "epoch": 0.0768,
      "grad_norm": 1.3121402263641357,
      "learning_rate": 9.96403788817465e-05,
      "loss": 0.6912,
      "step": 768
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.9180154800415039,
      "learning_rate": 9.96096627333437e-05,
      "loss": 0.8783,
      "step": 800
    },
    {
      "epoch": 0.0832,
      "grad_norm": 2.254473924636841,
      "learning_rate": 9.957769314055117e-05,
      "loss": 0.7987,
      "step": 832
    },
    {
      "epoch": 0.0864,
      "grad_norm": 1.9398365020751953,
      "learning_rate": 9.954447091111694e-05,
      "loss": 0.7703,
      "step": 864
    },
    {
      "epoch": 0.0896,
      "grad_norm": 1.4880696535110474,
      "learning_rate": 9.950999688443833e-05,
      "loss": 0.7258,
      "step": 896
    },
    {
      "epoch": 0.0928,
      "grad_norm": 1.8427962064743042,
      "learning_rate": 9.947427193154071e-05,
      "loss": 0.6981,
      "step": 928
    },
    {
      "epoch": 0.096,
      "grad_norm": 3.3647401332855225,
      "learning_rate": 9.943729695505552e-05,
      "loss": 0.7862,
      "step": 960
    },
    {
      "epoch": 0.0992,
      "grad_norm": 1.852992296218872,
      "learning_rate": 9.939907288919747e-05,
      "loss": 0.8016,
      "step": 992
    },
    {
      "epoch": 0.1024,
      "grad_norm": 1.2231330871582031,
      "learning_rate": 9.935960069974096e-05,
      "loss": 0.8001,
      "step": 1024
    },
    {
      "epoch": 0.1056,
      "grad_norm": 1.2329598665237427,
      "learning_rate": 9.931888138399561e-05,
      "loss": 0.7656,
      "step": 1056
    },
    {
      "epoch": 0.1088,
      "grad_norm": 1.4887111186981201,
      "learning_rate": 9.927691597078108e-05,
      "loss": 0.7772,
      "step": 1088
    },
    {
      "epoch": 0.112,
      "grad_norm": 1.1879202127456665,
      "learning_rate": 9.923370552040116e-05,
      "loss": 0.7368,
      "step": 1120
    },
    {
      "epoch": 0.1152,
      "grad_norm": 1.4578642845153809,
      "learning_rate": 9.918925112461688e-05,
      "loss": 0.7226,
      "step": 1152
    },
    {
      "epoch": 0.1184,
      "grad_norm": 3.8356716632843018,
      "learning_rate": 9.914355390661896e-05,
      "loss": 0.7468,
      "step": 1184
    },
    {
      "epoch": 0.1216,
      "grad_norm": 3.390878200531006,
      "learning_rate": 9.909661502099943e-05,
      "loss": 0.7163,
      "step": 1216
    },
    {
      "epoch": 0.1248,
      "grad_norm": 2.217479944229126,
      "learning_rate": 9.904843565372248e-05,
      "loss": 0.7805,
      "step": 1248
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.7309045195579529,
      "learning_rate": 9.899901702209445e-05,
      "loss": 0.6929,
      "step": 1280
    },
    {
      "epoch": 0.1312,
      "grad_norm": 1.173700213432312,
      "learning_rate": 9.89483603747331e-05,
      "loss": 0.726,
      "step": 1312
    },
    {
      "epoch": 0.1344,
      "grad_norm": 1.4089820384979248,
      "learning_rate": 9.88964669915361e-05,
      "loss": 0.8606,
      "step": 1344
    },
    {
      "epoch": 0.1376,
      "grad_norm": 1.0375796556472778,
      "learning_rate": 9.884333818364861e-05,
      "loss": 0.721,
      "step": 1376
    },
    {
      "epoch": 0.1408,
      "grad_norm": 2.082084894180298,
      "learning_rate": 9.878897529343023e-05,
      "loss": 0.7884,
      "step": 1408
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.7961512804031372,
      "learning_rate": 9.873337969442101e-05,
      "loss": 0.774,
      "step": 1440
    },
    {
      "epoch": 0.1472,
      "grad_norm": 1.3074238300323486,
      "learning_rate": 9.867655279130683e-05,
      "loss": 0.7392,
      "step": 1472
    },
    {
      "epoch": 0.1504,
      "grad_norm": 1.5205963850021362,
      "learning_rate": 9.861849601988383e-05,
      "loss": 0.7731,
      "step": 1504
    },
    {
      "epoch": 0.1536,
      "grad_norm": 1.4995771646499634,
      "learning_rate": 9.855921084702219e-05,
      "loss": 0.8281,
      "step": 1536
    },
    {
      "epoch": 0.1568,
      "grad_norm": 1.0279921293258667,
      "learning_rate": 9.849869877062902e-05,
      "loss": 0.6942,
      "step": 1568
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.8020853996276855,
      "learning_rate": 9.843696131961058e-05,
      "loss": 0.7389,
      "step": 1600
    },
    {
      "epoch": 0.1632,
      "grad_norm": 2.99129056930542,
      "learning_rate": 9.837400005383354e-05,
      "loss": 0.7483,
      "step": 1632
    },
    {
      "epoch": 0.1664,
      "grad_norm": 2.325167179107666,
      "learning_rate": 9.830981656408574e-05,
      "loss": 0.7483,
      "step": 1664
    },
    {
      "epoch": 0.1696,
      "grad_norm": 0.7245140671730042,
      "learning_rate": 9.824441247203579e-05,
      "loss": 0.7633,
      "step": 1696
    },
    {
      "epoch": 0.1728,
      "grad_norm": 2.7938778400421143,
      "learning_rate": 9.817778943019228e-05,
      "loss": 0.7812,
      "step": 1728
    },
    {
      "epoch": 0.176,
      "grad_norm": 1.2263625860214233,
      "learning_rate": 9.810994912186189e-05,
      "loss": 0.7712,
      "step": 1760
    },
    {
      "epoch": 0.1792,
      "grad_norm": 1.2694672346115112,
      "learning_rate": 9.804089326110697e-05,
      "loss": 0.7297,
      "step": 1792
    },
    {
      "epoch": 0.1824,
      "grad_norm": 1.255414366722107,
      "learning_rate": 9.797062359270215e-05,
      "loss": 0.735,
      "step": 1824
    },
    {
      "epoch": 0.1856,
      "grad_norm": 1.3175591230392456,
      "learning_rate": 9.789914189209029e-05,
      "loss": 0.7633,
      "step": 1856
    },
    {
      "epoch": 0.1888,
      "grad_norm": 1.0326446294784546,
      "learning_rate": 9.78264499653376e-05,
      "loss": 0.7955,
      "step": 1888
    },
    {
      "epoch": 0.192,
      "grad_norm": 1.093620777130127,
      "learning_rate": 9.775254964908807e-05,
      "loss": 0.766,
      "step": 1920
    },
    {
      "epoch": 0.1952,
      "grad_norm": 1.4234970808029175,
      "learning_rate": 9.767744281051701e-05,
      "loss": 0.6725,
      "step": 1952
    },
    {
      "epoch": 0.1984,
      "grad_norm": 0.7571769952774048,
      "learning_rate": 9.760113134728384e-05,
      "loss": 0.6953,
      "step": 1984
    },
    {
      "epoch": 0.2016,
      "grad_norm": 1.7207865715026855,
      "learning_rate": 9.752361718748423e-05,
      "loss": 0.7356,
      "step": 2016
    },
    {
      "epoch": 0.2048,
      "grad_norm": 2.240748882293701,
      "learning_rate": 9.744490228960138e-05,
      "loss": 0.8067,
      "step": 2048
    },
    {
      "epoch": 0.208,
      "grad_norm": 1.2544214725494385,
      "learning_rate": 9.736498864245638e-05,
      "loss": 0.7618,
      "step": 2080
    },
    {
      "epoch": 0.2112,
      "grad_norm": 5.976646900177002,
      "learning_rate": 9.728387826515819e-05,
      "loss": 0.6825,
      "step": 2112
    },
    {
      "epoch": 0.2144,
      "grad_norm": 4.557011127471924,
      "learning_rate": 9.72015732070525e-05,
      "loss": 0.7623,
      "step": 2144
    },
    {
      "epoch": 0.2176,
      "grad_norm": 0.8000884056091309,
      "learning_rate": 9.71180755476699e-05,
      "loss": 0.7719,
      "step": 2176
    },
    {
      "epoch": 0.2208,
      "grad_norm": 1.115488052368164,
      "learning_rate": 9.703338739667346e-05,
      "loss": 0.7913,
      "step": 2208
    },
    {
      "epoch": 0.224,
      "grad_norm": 1.3180317878723145,
      "learning_rate": 9.694751089380536e-05,
      "loss": 0.7452,
      "step": 2240
    },
    {
      "epoch": 0.2272,
      "grad_norm": 2.9995932579040527,
      "learning_rate": 9.686044820883285e-05,
      "loss": 0.7962,
      "step": 2272
    },
    {
      "epoch": 0.2304,
      "grad_norm": 1.234027624130249,
      "learning_rate": 9.677220154149336e-05,
      "loss": 0.828,
      "step": 2304
    },
    {
      "epoch": 0.2336,
      "grad_norm": 1.6579309701919556,
      "learning_rate": 9.668277312143907e-05,
      "loss": 0.7569,
      "step": 2336
    },
    {
      "epoch": 0.2368,
      "grad_norm": 1.5580084323883057,
      "learning_rate": 9.65921652081804e-05,
      "loss": 0.7947,
      "step": 2368
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.6711795330047607,
      "learning_rate": 9.650038009102905e-05,
      "loss": 0.7461,
      "step": 2400
    },
    {
      "epoch": 0.2432,
      "grad_norm": 1.2285038232803345,
      "learning_rate": 9.640742008904005e-05,
      "loss": 0.6587,
      "step": 2432
    },
    {
      "epoch": 0.2464,
      "grad_norm": 0.7901808619499207,
      "learning_rate": 9.631328755095334e-05,
      "loss": 0.7182,
      "step": 2464
    },
    {
      "epoch": 0.2496,
      "grad_norm": 0.6125284433364868,
      "learning_rate": 9.62179848551342e-05,
      "loss": 0.709,
      "step": 2496
    },
    {
      "epoch": 0.2528,
      "grad_norm": 1.1602981090545654,
      "learning_rate": 9.612151440951334e-05,
      "loss": 0.7039,
      "step": 2528
    },
    {
      "epoch": 0.256,
      "grad_norm": 2.366184711456299,
      "learning_rate": 9.602387865152597e-05,
      "loss": 0.8669,
      "step": 2560
    },
    {
      "epoch": 0.2592,
      "grad_norm": 2.583352565765381,
      "learning_rate": 9.592508004805023e-05,
      "loss": 0.7258,
      "step": 2592
    },
    {
      "epoch": 0.2624,
      "grad_norm": 2.132749557495117,
      "learning_rate": 9.58251210953449e-05,
      "loss": 0.6971,
      "step": 2624
    },
    {
      "epoch": 0.2656,
      "grad_norm": 1.4479436874389648,
      "learning_rate": 9.572400431898627e-05,
      "loss": 0.8086,
      "step": 2656
    },
    {
      "epoch": 0.2688,
      "grad_norm": 1.2764617204666138,
      "learning_rate": 9.562173227380436e-05,
      "loss": 0.7426,
      "step": 2688
    },
    {
      "epoch": 0.272,
      "grad_norm": 3.4120121002197266,
      "learning_rate": 9.55183075438184e-05,
      "loss": 0.7382,
      "step": 2720
    },
    {
      "epoch": 0.2752,
      "grad_norm": 1.9773039817810059,
      "learning_rate": 9.541373274217145e-05,
      "loss": 0.7903,
      "step": 2752
    },
    {
      "epoch": 0.2784,
      "grad_norm": 1.4097728729248047,
      "learning_rate": 9.530801051106449e-05,
      "loss": 0.7713,
      "step": 2784
    },
    {
      "epoch": 0.2816,
      "grad_norm": 1.0817668437957764,
      "learning_rate": 9.520114352168958e-05,
      "loss": 0.7275,
      "step": 2816
    },
    {
      "epoch": 0.2848,
      "grad_norm": 1.2667794227600098,
      "learning_rate": 9.509313447416242e-05,
      "loss": 0.6648,
      "step": 2848
    },
    {
      "epoch": 0.288,
      "grad_norm": 1.8679159879684448,
      "learning_rate": 9.498398609745405e-05,
      "loss": 0.7445,
      "step": 2880
    },
    {
      "epoch": 0.2912,
      "grad_norm": 2.8598544597625732,
      "learning_rate": 9.487370114932202e-05,
      "loss": 0.733,
      "step": 2912
    },
    {
      "epoch": 0.2944,
      "grad_norm": 0.9554559588432312,
      "learning_rate": 9.476228241624059e-05,
      "loss": 0.7487,
      "step": 2944
    },
    {
      "epoch": 0.2976,
      "grad_norm": 1.926672101020813,
      "learning_rate": 9.464973271333042e-05,
      "loss": 0.8864,
      "step": 2976
    },
    {
      "epoch": 0.3008,
      "grad_norm": 0.8425309658050537,
      "learning_rate": 9.45360548842874e-05,
      "loss": 0.7295,
      "step": 3008
    },
    {
      "epoch": 0.304,
      "grad_norm": 1.3110431432724,
      "learning_rate": 9.442125180131078e-05,
      "loss": 0.7547,
      "step": 3040
    },
    {
      "epoch": 0.3072,
      "grad_norm": 0.9774306416511536,
      "learning_rate": 9.430532636503068e-05,
      "loss": 0.7099,
      "step": 3072
    },
    {
      "epoch": 0.3104,
      "grad_norm": 0.6718234419822693,
      "learning_rate": 9.418828150443469e-05,
      "loss": 0.7636,
      "step": 3104
    },
    {
      "epoch": 0.3136,
      "grad_norm": 1.2758376598358154,
      "learning_rate": 9.407012017679393e-05,
      "loss": 0.7066,
      "step": 3136
    },
    {
      "epoch": 0.3168,
      "grad_norm": 1.3185311555862427,
      "learning_rate": 9.395084536758838e-05,
      "loss": 0.6785,
      "step": 3168
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.550795078277588,
      "learning_rate": 9.383046009043134e-05,
      "loss": 0.7451,
      "step": 3200
    },
    {
      "epoch": 0.3232,
      "grad_norm": 1.1686354875564575,
      "learning_rate": 9.370896738699339e-05,
      "loss": 0.6652,
      "step": 3232
    },
    {
      "epoch": 0.3264,
      "grad_norm": 0.848976194858551,
      "learning_rate": 9.358637032692545e-05,
      "loss": 0.7705,
      "step": 3264
    },
    {
      "epoch": 0.3296,
      "grad_norm": 1.3812384605407715,
      "learning_rate": 9.346267200778126e-05,
      "loss": 0.7168,
      "step": 3296
    },
    {
      "epoch": 0.3328,
      "grad_norm": 1.008135199546814,
      "learning_rate": 9.333787555493914e-05,
      "loss": 0.7352,
      "step": 3328
    },
    {
      "epoch": 0.336,
      "grad_norm": 1.2273484468460083,
      "learning_rate": 9.321198412152301e-05,
      "loss": 0.7979,
      "step": 3360
    },
    {
      "epoch": 0.3392,
      "grad_norm": 0.8740741610527039,
      "learning_rate": 9.308500088832272e-05,
      "loss": 0.6846,
      "step": 3392
    },
    {
      "epoch": 0.3424,
      "grad_norm": 1.3684589862823486,
      "learning_rate": 9.295692906371363e-05,
      "loss": 0.7758,
      "step": 3424
    },
    {
      "epoch": 0.3456,
      "grad_norm": 1.2861257791519165,
      "learning_rate": 9.282777188357565e-05,
      "loss": 0.6581,
      "step": 3456
    },
    {
      "epoch": 0.3488,
      "grad_norm": 0.8915108442306519,
      "learning_rate": 9.269753261121138e-05,
      "loss": 0.7935,
      "step": 3488
    },
    {
      "epoch": 0.352,
      "grad_norm": 1.1308799982070923,
      "learning_rate": 9.256621453726379e-05,
      "loss": 0.7759,
      "step": 3520
    },
    {
      "epoch": 0.3552,
      "grad_norm": 1.0886152982711792,
      "learning_rate": 9.243382097963291e-05,
      "loss": 0.7207,
      "step": 3552
    },
    {
      "epoch": 0.3584,
      "grad_norm": 0.675757110118866,
      "learning_rate": 9.230035528339211e-05,
      "loss": 0.6876,
      "step": 3584
    },
    {
      "epoch": 0.3616,
      "grad_norm": 3.258622884750366,
      "learning_rate": 9.216582082070358e-05,
      "loss": 0.7498,
      "step": 3616
    },
    {
      "epoch": 0.3648,
      "grad_norm": 3.8826818466186523,
      "learning_rate": 9.203022099073309e-05,
      "loss": 0.7993,
      "step": 3648
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.9782927632331848,
      "learning_rate": 9.189355921956412e-05,
      "loss": 0.7005,
      "step": 3680
    },
    {
      "epoch": 0.3712,
      "grad_norm": 1.1662654876708984,
      "learning_rate": 9.175583896011131e-05,
      "loss": 0.6732,
      "step": 3712
    },
    {
      "epoch": 0.3744,
      "grad_norm": 1.038501501083374,
      "learning_rate": 9.161706369203317e-05,
      "loss": 0.7414,
      "step": 3744
    },
    {
      "epoch": 0.3776,
      "grad_norm": 2.3218936920166016,
      "learning_rate": 9.147723692164427e-05,
      "loss": 0.8008,
      "step": 3776
    },
    {
      "epoch": 0.3808,
      "grad_norm": 2.1190292835235596,
      "learning_rate": 9.13363621818265e-05,
      "loss": 0.711,
      "step": 3808
    },
    {
      "epoch": 0.384,
      "grad_norm": 1.8653652667999268,
      "learning_rate": 9.119444303193996e-05,
      "loss": 0.7641,
      "step": 3840
    },
    {
      "epoch": 0.3872,
      "grad_norm": 1.760704517364502,
      "learning_rate": 9.10514830577329e-05,
      "loss": 0.7231,
      "step": 3872
    },
    {
      "epoch": 0.3904,
      "grad_norm": 0.8437080979347229,
      "learning_rate": 9.090748587125118e-05,
      "loss": 0.7089,
      "step": 3904
    },
    {
      "epoch": 0.3936,
      "grad_norm": 1.6417099237442017,
      "learning_rate": 9.076245511074703e-05,
      "loss": 0.7645,
      "step": 3936
    },
    {
      "epoch": 0.3968,
      "grad_norm": 1.0280499458312988,
      "learning_rate": 9.06163944405871e-05,
      "loss": 0.78,
      "step": 3968
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.645205020904541,
      "learning_rate": 9.046930755115985e-05,
      "loss": 0.7443,
      "step": 4000
    },
    {
      "epoch": 0.4032,
      "grad_norm": 1.4332572221755981,
      "learning_rate": 9.032119815878236e-05,
      "loss": 0.7138,
      "step": 4032
    },
    {
      "epoch": 0.4064,
      "grad_norm": 1.3062529563903809,
      "learning_rate": 9.017207000560639e-05,
      "loss": 0.6866,
      "step": 4064
    },
    {
      "epoch": 0.4096,
      "grad_norm": 1.559920072555542,
      "learning_rate": 9.002192685952385e-05,
      "loss": 0.7289,
      "step": 4096
    },
    {
      "epoch": 0.4128,
      "grad_norm": 2.111950635910034,
      "learning_rate": 8.987077251407158e-05,
      "loss": 0.7011,
      "step": 4128
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.8812033534049988,
      "learning_rate": 8.971861078833557e-05,
      "loss": 0.7469,
      "step": 4160
    },
    {
      "epoch": 0.4192,
      "grad_norm": 0.8479238748550415,
      "learning_rate": 8.956544552685437e-05,
      "loss": 0.7263,
      "step": 4192
    },
    {
      "epoch": 0.4224,
      "grad_norm": 1.0125929117202759,
      "learning_rate": 8.941128059952201e-05,
      "loss": 0.6762,
      "step": 4224
    },
    {
      "epoch": 0.4256,
      "grad_norm": 0.9122424721717834,
      "learning_rate": 8.925611990149021e-05,
      "loss": 0.7076,
      "step": 4256
    },
    {
      "epoch": 0.4288,
      "grad_norm": 1.814253330230713,
      "learning_rate": 8.909996735306996e-05,
      "loss": 0.7143,
      "step": 4288
    },
    {
      "epoch": 0.432,
      "grad_norm": 1.4890289306640625,
      "learning_rate": 8.894282689963251e-05,
      "loss": 0.6931,
      "step": 4320
    },
    {
      "epoch": 0.4352,
      "grad_norm": 1.908116340637207,
      "learning_rate": 8.878470251150959e-05,
      "loss": 0.701,
      "step": 4352
    },
    {
      "epoch": 0.4384,
      "grad_norm": 1.2831019163131714,
      "learning_rate": 8.862559818389322e-05,
      "loss": 0.7625,
      "step": 4384
    },
    {
      "epoch": 0.4416,
      "grad_norm": 0.923768162727356,
      "learning_rate": 8.846551793673467e-05,
      "loss": 0.7902,
      "step": 4416
    },
    {
      "epoch": 0.4448,
      "grad_norm": 1.6989527940750122,
      "learning_rate": 8.83044658146429e-05,
      "loss": 0.7006,
      "step": 4448
    },
    {
      "epoch": 0.448,
      "grad_norm": 1.203029990196228,
      "learning_rate": 8.814244588678245e-05,
      "loss": 0.7588,
      "step": 4480
    },
    {
      "epoch": 0.4512,
      "grad_norm": 1.8377019166946411,
      "learning_rate": 8.797946224677052e-05,
      "loss": 0.6975,
      "step": 4512
    },
    {
      "epoch": 0.4544,
      "grad_norm": 1.4714457988739014,
      "learning_rate": 8.78155190125736e-05,
      "loss": 0.6502,
      "step": 4544
    },
    {
      "epoch": 0.4576,
      "grad_norm": 1.6311497688293457,
      "learning_rate": 8.765062032640346e-05,
      "loss": 0.7536,
      "step": 4576
    },
    {
      "epoch": 0.4608,
      "grad_norm": 1.8238953351974487,
      "learning_rate": 8.748477035461238e-05,
      "loss": 0.7899,
      "step": 4608
    },
    {
      "epoch": 0.464,
      "grad_norm": 1.5541362762451172,
      "learning_rate": 8.7317973287588e-05,
      "loss": 0.6904,
      "step": 4640
    },
    {
      "epoch": 0.4672,
      "grad_norm": 1.032272219657898,
      "learning_rate": 8.715023333964736e-05,
      "loss": 0.7395,
      "step": 4672
    },
    {
      "epoch": 0.4704,
      "grad_norm": 1.3095510005950928,
      "learning_rate": 8.69815547489305e-05,
      "loss": 0.6854,
      "step": 4704
    },
    {
      "epoch": 0.4736,
      "grad_norm": 1.5274263620376587,
      "learning_rate": 8.681194177729327e-05,
      "loss": 0.7498,
      "step": 4736
    },
    {
      "epoch": 0.4768,
      "grad_norm": 1.4236122369766235,
      "learning_rate": 8.66413987101998e-05,
      "loss": 0.7356,
      "step": 4768
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.2118279933929443,
      "learning_rate": 8.646992985661404e-05,
      "loss": 0.7178,
      "step": 4800
    },
    {
      "epoch": 0.4832,
      "grad_norm": 3.3495805263519287,
      "learning_rate": 8.629753954889107e-05,
      "loss": 0.7326,
      "step": 4832
    },
    {
      "epoch": 0.4864,
      "grad_norm": 0.6829349398612976,
      "learning_rate": 8.612423214266749e-05,
      "loss": 0.7838,
      "step": 4864
    },
    {
      "epoch": 0.4896,
      "grad_norm": 0.8314148187637329,
      "learning_rate": 8.595001201675147e-05,
      "loss": 0.7007,
      "step": 4896
    },
    {
      "epoch": 0.4928,
      "grad_norm": 1.2672547101974487,
      "learning_rate": 8.577488357301209e-05,
      "loss": 0.7377,
      "step": 4928
    },
    {
      "epoch": 0.496,
      "grad_norm": 1.3968323469161987,
      "learning_rate": 8.559885123626807e-05,
      "loss": 0.6774,
      "step": 4960
    },
    {
      "epoch": 0.4992,
      "grad_norm": 1.2808008193969727,
      "learning_rate": 8.542191945417601e-05,
      "loss": 0.6807,
      "step": 4992
    },
    {
      "epoch": 0.5024,
      "grad_norm": 1.9290404319763184,
      "learning_rate": 8.524409269711807e-05,
      "loss": 0.7376,
      "step": 5024
    },
    {
      "epoch": 0.5056,
      "grad_norm": 1.3726913928985596,
      "learning_rate": 8.506537545808892e-05,
      "loss": 0.7402,
      "step": 5056
    },
    {
      "epoch": 0.5088,
      "grad_norm": 1.7894905805587769,
      "learning_rate": 8.48857722525823e-05,
      "loss": 0.6991,
      "step": 5088
    },
    {
      "epoch": 0.512,
      "grad_norm": 1.1462016105651855,
      "learning_rate": 8.470528761847684e-05,
      "loss": 0.7989,
      "step": 5120
    },
    {
      "epoch": 0.5152,
      "grad_norm": 0.7457314729690552,
      "learning_rate": 8.452392611592153e-05,
      "loss": 0.7616,
      "step": 5152
    },
    {
      "epoch": 0.5184,
      "grad_norm": 1.728968858718872,
      "learning_rate": 8.434169232722043e-05,
      "loss": 0.6324,
      "step": 5184
    },
    {
      "epoch": 0.5216,
      "grad_norm": 0.9103218913078308,
      "learning_rate": 8.415859085671683e-05,
      "loss": 0.7222,
      "step": 5216
    },
    {
      "epoch": 0.5248,
      "grad_norm": 1.602072834968567,
      "learning_rate": 8.397462633067704e-05,
      "loss": 0.7265,
      "step": 5248
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.9967379570007324,
      "learning_rate": 8.378980339717349e-05,
      "loss": 0.7042,
      "step": 5280
    },
    {
      "epoch": 0.5312,
      "grad_norm": 1.9905532598495483,
      "learning_rate": 8.360412672596712e-05,
      "loss": 0.8098,
      "step": 5312
    },
    {
      "epoch": 0.5344,
      "grad_norm": 1.1432929039001465,
      "learning_rate": 8.341760100838965e-05,
      "loss": 0.7591,
      "step": 5344
    },
    {
      "epoch": 0.5376,
      "grad_norm": 2.1654775142669678,
      "learning_rate": 8.323023095722486e-05,
      "loss": 0.8071,
      "step": 5376
    },
    {
      "epoch": 0.5408,
      "grad_norm": 1.2390097379684448,
      "learning_rate": 8.304202130658959e-05,
      "loss": 0.834,
      "step": 5408
    },
    {
      "epoch": 0.544,
      "grad_norm": 1.0290433168411255,
      "learning_rate": 8.285297681181408e-05,
      "loss": 0.8228,
      "step": 5440
    },
    {
      "epoch": 0.5472,
      "grad_norm": 1.299111008644104,
      "learning_rate": 8.26631022493219e-05,
      "loss": 0.7099,
      "step": 5472
    },
    {
      "epoch": 0.5504,
      "grad_norm": 0.8850242495536804,
      "learning_rate": 8.247240241650918e-05,
      "loss": 0.7646,
      "step": 5504
    },
    {
      "epoch": 0.5536,
      "grad_norm": 1.980812907218933,
      "learning_rate": 8.22808821316235e-05,
      "loss": 0.7312,
      "step": 5536
    },
    {
      "epoch": 0.5568,
      "grad_norm": 1.0378026962280273,
      "learning_rate": 8.208854623364202e-05,
      "loss": 0.7277,
      "step": 5568
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.6820452213287354,
      "learning_rate": 8.189539958214935e-05,
      "loss": 0.7654,
      "step": 5600
    },
    {
      "epoch": 0.5632,
      "grad_norm": 1.494661808013916,
      "learning_rate": 8.170144705721465e-05,
      "loss": 0.7208,
      "step": 5632
    },
    {
      "epoch": 0.5664,
      "grad_norm": 0.9761049747467041,
      "learning_rate": 8.150669355926846e-05,
      "loss": 0.6898,
      "step": 5664
    },
    {
      "epoch": 0.5696,
      "grad_norm": 1.3057583570480347,
      "learning_rate": 8.131114400897874e-05,
      "loss": 0.7887,
      "step": 5696
    },
    {
      "epoch": 0.5728,
      "grad_norm": 1.0025156736373901,
      "learning_rate": 8.111480334712665e-05,
      "loss": 0.6483,
      "step": 5728
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.9818746447563171,
      "learning_rate": 8.091767653448167e-05,
      "loss": 0.8385,
      "step": 5760
    },
    {
      "epoch": 0.5792,
      "grad_norm": 1.1921987533569336,
      "learning_rate": 8.071976855167629e-05,
      "loss": 0.6707,
      "step": 5792
    },
    {
      "epoch": 0.5824,
      "grad_norm": 1.5055749416351318,
      "learning_rate": 8.052108439908013e-05,
      "loss": 0.7086,
      "step": 5824
    },
    {
      "epoch": 0.5856,
      "grad_norm": 1.7581650018692017,
      "learning_rate": 8.032162909667362e-05,
      "loss": 0.6696,
      "step": 5856
    },
    {
      "epoch": 0.5888,
      "grad_norm": 1.8909873962402344,
      "learning_rate": 8.01214076839212e-05,
      "loss": 0.7471,
      "step": 5888
    },
    {
      "epoch": 0.592,
      "grad_norm": 1.3570644855499268,
      "learning_rate": 7.992042521964389e-05,
      "loss": 0.655,
      "step": 5920
    },
    {
      "epoch": 0.5952,
      "grad_norm": 0.6561287641525269,
      "learning_rate": 7.971868678189161e-05,
      "loss": 0.719,
      "step": 5952
    },
    {
      "epoch": 0.5984,
      "grad_norm": 1.3650476932525635,
      "learning_rate": 7.951619746781474e-05,
      "loss": 0.7405,
      "step": 5984
    },
    {
      "epoch": 0.6016,
      "grad_norm": 2.8344266414642334,
      "learning_rate": 7.931296239353544e-05,
      "loss": 0.7192,
      "step": 6016
    },
    {
      "epoch": 0.6048,
      "grad_norm": 1.6202623844146729,
      "learning_rate": 7.910898669401839e-05,
      "loss": 0.7671,
      "step": 6048
    },
    {
      "epoch": 0.608,
      "grad_norm": 1.1194038391113281,
      "learning_rate": 7.890427552294093e-05,
      "loss": 0.7915,
      "step": 6080
    },
    {
      "epoch": 0.6112,
      "grad_norm": 0.8267541527748108,
      "learning_rate": 7.869883405256295e-05,
      "loss": 0.7441,
      "step": 6112
    },
    {
      "epoch": 0.6144,
      "grad_norm": 1.229134202003479,
      "learning_rate": 7.849266747359619e-05,
      "loss": 0.6548,
      "step": 6144
    },
    {
      "epoch": 0.6176,
      "grad_norm": 1.151248812675476,
      "learning_rate": 7.828578099507308e-05,
      "loss": 0.6795,
      "step": 6176
    },
    {
      "epoch": 0.6208,
      "grad_norm": 1.620975375175476,
      "learning_rate": 7.80781798442151e-05,
      "loss": 0.6352,
      "step": 6208
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.9030219912528992,
      "learning_rate": 7.786986926630078e-05,
      "loss": 0.7185,
      "step": 6240
    },
    {
      "epoch": 0.6272,
      "grad_norm": 1.2997703552246094,
      "learning_rate": 7.766085452453312e-05,
      "loss": 0.6523,
      "step": 6272
    },
    {
      "epoch": 0.6304,
      "grad_norm": 1.208347201347351,
      "learning_rate": 7.74511408999066e-05,
      "loss": 0.6928,
      "step": 6304
    },
    {
      "epoch": 0.6336,
      "grad_norm": 0.723646879196167,
      "learning_rate": 7.724073369107376e-05,
      "loss": 0.6603,
      "step": 6336
    },
    {
      "epoch": 0.6368,
      "grad_norm": 1.125978946685791,
      "learning_rate": 7.702963821421133e-05,
      "loss": 0.7328,
      "step": 6368
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.039461135864258,
      "learning_rate": 7.6817859802886e-05,
      "loss": 0.7545,
      "step": 6400
    },
    {
      "epoch": 0.6432,
      "grad_norm": 1.3743586540222168,
      "learning_rate": 7.660540380791942e-05,
      "loss": 0.67,
      "step": 6432
    },
    {
      "epoch": 0.6464,
      "grad_norm": 1.402256727218628,
      "learning_rate": 7.639227559725332e-05,
      "loss": 0.636,
      "step": 6464
    },
    {
      "epoch": 0.6496,
      "grad_norm": 1.0240074396133423,
      "learning_rate": 7.617848055581361e-05,
      "loss": 0.8179,
      "step": 6496
    },
    {
      "epoch": 0.6528,
      "grad_norm": 0.8905365467071533,
      "learning_rate": 7.596402408537443e-05,
      "loss": 0.7542,
      "step": 6528
    },
    {
      "epoch": 0.656,
      "grad_norm": 1.8598270416259766,
      "learning_rate": 7.574891160442179e-05,
      "loss": 0.7266,
      "step": 6560
    },
    {
      "epoch": 0.6592,
      "grad_norm": 0.9146720170974731,
      "learning_rate": 7.553314854801641e-05,
      "loss": 0.7861,
      "step": 6592
    },
    {
      "epoch": 0.6624,
      "grad_norm": 1.8956897258758545,
      "learning_rate": 7.531674036765662e-05,
      "loss": 0.7113,
      "step": 6624
    },
    {
      "epoch": 0.6656,
      "grad_norm": 1.0353283882141113,
      "learning_rate": 7.509969253114055e-05,
      "loss": 0.6984,
      "step": 6656
    },
    {
      "epoch": 0.6688,
      "grad_norm": 1.890493631362915,
      "learning_rate": 7.488201052242789e-05,
      "loss": 0.6687,
      "step": 6688
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.9367122054100037,
      "learning_rate": 7.46636998415015e-05,
      "loss": 0.719,
      "step": 6720
    },
    {
      "epoch": 0.6752,
      "grad_norm": 1.1989344358444214,
      "learning_rate": 7.444476600422828e-05,
      "loss": 0.775,
      "step": 6752
    },
    {
      "epoch": 0.6784,
      "grad_norm": 0.8481733202934265,
      "learning_rate": 7.42252145422199e-05,
      "loss": 0.7667,
      "step": 6784
    },
    {
      "epoch": 0.6816,
      "grad_norm": 1.0271095037460327,
      "learning_rate": 7.400505100269307e-05,
      "loss": 0.653,
      "step": 6816
    },
    {
      "epoch": 0.6848,
      "grad_norm": 1.3998816013336182,
      "learning_rate": 7.378428094832931e-05,
      "loss": 0.6651,
      "step": 6848
    },
    {
      "epoch": 0.688,
      "grad_norm": 1.3338642120361328,
      "learning_rate": 7.356290995713437e-05,
      "loss": 0.6266,
      "step": 6880
    },
    {
      "epoch": 0.6912,
      "grad_norm": 0.8170168995857239,
      "learning_rate": 7.334094362229739e-05,
      "loss": 0.765,
      "step": 6912
    },
    {
      "epoch": 0.6944,
      "grad_norm": 1.4982614517211914,
      "learning_rate": 7.311838755204959e-05,
      "loss": 0.641,
      "step": 6944
    },
    {
      "epoch": 0.6976,
      "grad_norm": 1.623159646987915,
      "learning_rate": 7.290222928580347e-05,
      "loss": 0.6462,
      "step": 6976
    },
    {
      "epoch": 0.7008,
      "grad_norm": 1.169145941734314,
      "learning_rate": 7.267852862072673e-05,
      "loss": 0.7506,
      "step": 7008
    },
    {
      "epoch": 0.704,
      "grad_norm": 1.011816382408142,
      "learning_rate": 7.245425495690538e-05,
      "loss": 0.7183,
      "step": 7040
    },
    {
      "epoch": 0.7072,
      "grad_norm": 3.0435078144073486,
      "learning_rate": 7.222941396086789e-05,
      "loss": 0.7948,
      "step": 7072
    },
    {
      "epoch": 0.7104,
      "grad_norm": 0.802679717540741,
      "learning_rate": 7.2004011313477e-05,
      "loss": 0.8216,
      "step": 7104
    },
    {
      "epoch": 0.7136,
      "grad_norm": 0.7551457285881042,
      "learning_rate": 7.17780527097862e-05,
      "loss": 0.7823,
      "step": 7136
    },
    {
      "epoch": 0.7168,
      "grad_norm": 1.3118380308151245,
      "learning_rate": 7.155154385889589e-05,
      "loss": 0.7803,
      "step": 7168
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.1100643873214722,
      "learning_rate": 7.132449048380907e-05,
      "loss": 0.7425,
      "step": 7200
    },
    {
      "epoch": 0.7232,
      "grad_norm": 0.8792561888694763,
      "learning_rate": 7.109689832128673e-05,
      "loss": 0.7515,
      "step": 7232
    },
    {
      "epoch": 0.7264,
      "grad_norm": 0.8382082581520081,
      "learning_rate": 7.0868773121703e-05,
      "loss": 0.8134,
      "step": 7264
    },
    {
      "epoch": 0.7296,
      "grad_norm": 1.7332772016525269,
      "learning_rate": 7.064012064889971e-05,
      "loss": 0.6971,
      "step": 7296
    },
    {
      "epoch": 0.7328,
      "grad_norm": 1.4402042627334595,
      "learning_rate": 7.041094668004093e-05,
      "loss": 0.6845,
      "step": 7328
    },
    {
      "epoch": 0.736,
      "grad_norm": 1.1810777187347412,
      "learning_rate": 7.018125700546683e-05,
      "loss": 0.7472,
      "step": 7360
    },
    {
      "epoch": 0.7392,
      "grad_norm": 0.9390580058097839,
      "learning_rate": 6.995105742854759e-05,
      "loss": 0.8127,
      "step": 7392
    },
    {
      "epoch": 0.7424,
      "grad_norm": 1.570432186126709,
      "learning_rate": 6.972035376553656e-05,
      "loss": 0.7071,
      "step": 7424
    },
    {
      "epoch": 0.7456,
      "grad_norm": 1.168547511100769,
      "learning_rate": 6.94891518454234e-05,
      "loss": 0.7017,
      "step": 7456
    },
    {
      "epoch": 0.7488,
      "grad_norm": 1.1337932348251343,
      "learning_rate": 6.925745750978686e-05,
      "loss": 0.6738,
      "step": 7488
    },
    {
      "epoch": 0.752,
      "grad_norm": 1.351352334022522,
      "learning_rate": 6.902527661264701e-05,
      "loss": 0.7548,
      "step": 7520
    },
    {
      "epoch": 0.7552,
      "grad_norm": 0.6679269671440125,
      "learning_rate": 6.87926150203176e-05,
      "loss": 0.7106,
      "step": 7552
    },
    {
      "epoch": 0.7584,
      "grad_norm": 1.3825992345809937,
      "learning_rate": 6.855947861125759e-05,
      "loss": 0.6443,
      "step": 7584
    },
    {
      "epoch": 0.7616,
      "grad_norm": 1.1650683879852295,
      "learning_rate": 6.832587327592275e-05,
      "loss": 0.7547,
      "step": 7616
    },
    {
      "epoch": 0.7648,
      "grad_norm": 1.5112355947494507,
      "learning_rate": 6.809180491661678e-05,
      "loss": 0.7076,
      "step": 7648
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.8795199990272522,
      "learning_rate": 6.785727944734228e-05,
      "loss": 0.7345,
      "step": 7680
    },
    {
      "epoch": 0.7712,
      "grad_norm": 1.6340776681900024,
      "learning_rate": 6.762230279365114e-05,
      "loss": 0.7517,
      "step": 7712
    },
    {
      "epoch": 0.7744,
      "grad_norm": 1.022924542427063,
      "learning_rate": 6.738688089249502e-05,
      "loss": 0.6874,
      "step": 7744
    },
    {
      "epoch": 0.7776,
      "grad_norm": 1.2930107116699219,
      "learning_rate": 6.715101969207525e-05,
      "loss": 0.7479,
      "step": 7776
    },
    {
      "epoch": 0.7808,
      "grad_norm": 1.9842311143875122,
      "learning_rate": 6.691472515169251e-05,
      "loss": 0.7479,
      "step": 7808
    },
    {
      "epoch": 0.784,
      "grad_norm": 1.5960675477981567,
      "learning_rate": 6.667800324159636e-05,
      "loss": 0.7928,
      "step": 7840
    },
    {
      "epoch": 0.7872,
      "grad_norm": 3.447913885116577,
      "learning_rate": 6.644085994283433e-05,
      "loss": 0.6924,
      "step": 7872
    },
    {
      "epoch": 0.7904,
      "grad_norm": 0.8809865713119507,
      "learning_rate": 6.620330124710077e-05,
      "loss": 0.7955,
      "step": 7904
    },
    {
      "epoch": 0.7936,
      "grad_norm": 1.3761461973190308,
      "learning_rate": 6.596533315658555e-05,
      "loss": 0.6842,
      "step": 7936
    },
    {
      "epoch": 0.7968,
      "grad_norm": 0.9557456374168396,
      "learning_rate": 6.572696168382235e-05,
      "loss": 0.7285,
      "step": 7968
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.7569695115089417,
      "learning_rate": 6.548819285153676e-05,
      "loss": 0.6431,
      "step": 8000
    },
    {
      "epoch": 0.8032,
      "grad_norm": 1.2884209156036377,
      "learning_rate": 6.524903269249411e-05,
      "loss": 0.739,
      "step": 8032
    },
    {
      "epoch": 0.8064,
      "grad_norm": 1.033050775527954,
      "learning_rate": 6.500948724934703e-05,
      "loss": 0.6759,
      "step": 8064
    },
    {
      "epoch": 0.8096,
      "grad_norm": 0.9404661655426025,
      "learning_rate": 6.47695625744828e-05,
      "loss": 0.696,
      "step": 8096
    },
    {
      "epoch": 0.8128,
      "grad_norm": 0.8363805413246155,
      "learning_rate": 6.452926472987044e-05,
      "loss": 0.7273,
      "step": 8128
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.7976164817810059,
      "learning_rate": 6.428859978690748e-05,
      "loss": 0.6671,
      "step": 8160
    },
    {
      "epoch": 0.8192,
      "grad_norm": 1.6969666481018066,
      "learning_rate": 6.404757382626669e-05,
      "loss": 0.6968,
      "step": 8192
    },
    {
      "epoch": 0.8224,
      "grad_norm": 1.061860203742981,
      "learning_rate": 6.380619293774223e-05,
      "loss": 0.7424,
      "step": 8224
    },
    {
      "epoch": 0.8256,
      "grad_norm": 1.2336043119430542,
      "learning_rate": 6.356446322009607e-05,
      "loss": 0.6786,
      "step": 8256
    },
    {
      "epoch": 0.8288,
      "grad_norm": 1.3530735969543457,
      "learning_rate": 6.332239078090358e-05,
      "loss": 0.7042,
      "step": 8288
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.9186837673187256,
      "learning_rate": 6.307998173639954e-05,
      "loss": 0.7433,
      "step": 8320
    },
    {
      "epoch": 0.8352,
      "grad_norm": 1.0583479404449463,
      "learning_rate": 6.283724221132333e-05,
      "loss": 0.6515,
      "step": 8352
    },
    {
      "epoch": 0.8384,
      "grad_norm": 1.468887209892273,
      "learning_rate": 6.259417833876432e-05,
      "loss": 0.7033,
      "step": 8384
    },
    {
      "epoch": 0.8416,
      "grad_norm": 0.7726921439170837,
      "learning_rate": 6.235079626000694e-05,
      "loss": 0.721,
      "step": 8416
    },
    {
      "epoch": 0.8448,
      "grad_norm": 1.8641211986541748,
      "learning_rate": 6.21071021243754e-05,
      "loss": 0.626,
      "step": 8448
    },
    {
      "epoch": 0.848,
      "grad_norm": 1.9702180624008179,
      "learning_rate": 6.186310208907839e-05,
      "loss": 0.6017,
      "step": 8480
    },
    {
      "epoch": 0.8512,
      "grad_norm": 2.057535171508789,
      "learning_rate": 6.161880231905354e-05,
      "loss": 0.7612,
      "step": 8512
    },
    {
      "epoch": 0.8544,
      "grad_norm": 2.2840230464935303,
      "learning_rate": 6.137420898681158e-05,
      "loss": 0.6609,
      "step": 8544
    },
    {
      "epoch": 0.8576,
      "grad_norm": 1.7856135368347168,
      "learning_rate": 6.112932827228044e-05,
      "loss": 0.7015,
      "step": 8576
    },
    {
      "epoch": 0.8608,
      "grad_norm": 1.0354335308074951,
      "learning_rate": 6.0884166362649075e-05,
      "loss": 0.6714,
      "step": 8608
    },
    {
      "epoch": 0.864,
      "grad_norm": 1.054237961769104,
      "learning_rate": 6.063872945221118e-05,
      "loss": 0.6928,
      "step": 8640
    },
    {
      "epoch": 0.8672,
      "grad_norm": 1.004862904548645,
      "learning_rate": 6.039302374220861e-05,
      "loss": 0.7676,
      "step": 8672
    },
    {
      "epoch": 0.8704,
      "grad_norm": 0.8693735003471375,
      "learning_rate": 6.0147055440674795e-05,
      "loss": 0.7562,
      "step": 8704
    },
    {
      "epoch": 0.8736,
      "grad_norm": 1.6824612617492676,
      "learning_rate": 5.990083076227782e-05,
      "loss": 0.6509,
      "step": 8736
    },
    {
      "epoch": 0.8768,
      "grad_norm": 3.1215667724609375,
      "learning_rate": 5.9654355928163416e-05,
      "loss": 0.7553,
      "step": 8768
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.4479137659072876,
      "learning_rate": 5.9407637165797793e-05,
      "loss": 0.8046,
      "step": 8800
    },
    {
      "epoch": 0.8832,
      "grad_norm": 2.769347906112671,
      "learning_rate": 5.916068070881026e-05,
      "loss": 0.6869,
      "step": 8832
    },
    {
      "epoch": 0.8864,
      "grad_norm": 1.338932752609253,
      "learning_rate": 5.891349279683578e-05,
      "loss": 0.6742,
      "step": 8864
    },
    {
      "epoch": 0.8896,
      "grad_norm": 1.15195631980896,
      "learning_rate": 5.8666079675357285e-05,
      "loss": 0.6972,
      "step": 8896
    },
    {
      "epoch": 0.8928,
      "grad_norm": 1.0247623920440674,
      "learning_rate": 5.841844759554787e-05,
      "loss": 0.7107,
      "step": 8928
    },
    {
      "epoch": 0.896,
      "grad_norm": 1.4130921363830566,
      "learning_rate": 5.817060281411284e-05,
      "loss": 0.7327,
      "step": 8960
    },
    {
      "epoch": 0.8992,
      "grad_norm": 0.6436507701873779,
      "learning_rate": 5.792255159313169e-05,
      "loss": 0.6418,
      "step": 8992
    },
    {
      "epoch": 0.9024,
      "grad_norm": 0.9555985331535339,
      "learning_rate": 5.7674300199899834e-05,
      "loss": 0.7157,
      "step": 9024
    },
    {
      "epoch": 0.9056,
      "grad_norm": 0.8774769306182861,
      "learning_rate": 5.742585490677024e-05,
      "loss": 0.6197,
      "step": 9056
    },
    {
      "epoch": 0.9088,
      "grad_norm": 0.9347734451293945,
      "learning_rate": 5.7177221990995e-05,
      "loss": 0.6672,
      "step": 9088
    },
    {
      "epoch": 0.912,
      "grad_norm": 1.2730952501296997,
      "learning_rate": 5.692840773456669e-05,
      "loss": 0.7524,
      "step": 9120
    },
    {
      "epoch": 0.9152,
      "grad_norm": 1.3449304103851318,
      "learning_rate": 5.667941842405968e-05,
      "loss": 0.7106,
      "step": 9152
    },
    {
      "epoch": 0.9184,
      "grad_norm": 2.288444757461548,
      "learning_rate": 5.643026035047128e-05,
      "loss": 0.7239,
      "step": 9184
    },
    {
      "epoch": 0.9216,
      "grad_norm": 1.1817107200622559,
      "learning_rate": 5.618093980906276e-05,
      "loss": 0.7342,
      "step": 9216
    },
    {
      "epoch": 0.9248,
      "grad_norm": 1.4276821613311768,
      "learning_rate": 5.5931463099200355e-05,
      "loss": 0.6198,
      "step": 9248
    },
    {
      "epoch": 0.928,
      "grad_norm": 1.0878974199295044,
      "learning_rate": 5.568183652419606e-05,
      "loss": 0.7204,
      "step": 9280
    },
    {
      "epoch": 0.9312,
      "grad_norm": 1.5497533082962036,
      "learning_rate": 5.54320663911484e-05,
      "loss": 0.7218,
      "step": 9312
    },
    {
      "epoch": 0.9344,
      "grad_norm": 0.5286266207695007,
      "learning_rate": 5.518215901078302e-05,
      "loss": 0.8243,
      "step": 9344
    },
    {
      "epoch": 0.9376,
      "grad_norm": 1.9889594316482544,
      "learning_rate": 5.493212069729332e-05,
      "loss": 0.6849,
      "step": 9376
    },
    {
      "epoch": 0.9408,
      "grad_norm": 1.6639822721481323,
      "learning_rate": 5.468195776818084e-05,
      "loss": 0.682,
      "step": 9408
    },
    {
      "epoch": 0.944,
      "grad_norm": 3.0651698112487793,
      "learning_rate": 5.4431676544095676e-05,
      "loss": 0.8112,
      "step": 9440
    },
    {
      "epoch": 0.9472,
      "grad_norm": 1.0381174087524414,
      "learning_rate": 5.4181283348676806e-05,
      "loss": 0.6497,
      "step": 9472
    },
    {
      "epoch": 0.9504,
      "grad_norm": 1.0353689193725586,
      "learning_rate": 5.393078450839228e-05,
      "loss": 0.6654,
      "step": 9504
    },
    {
      "epoch": 0.9536,
      "grad_norm": 1.6130503416061401,
      "learning_rate": 5.368018635237936e-05,
      "loss": 0.7351,
      "step": 9536
    },
    {
      "epoch": 0.9568,
      "grad_norm": 1.171970248222351,
      "learning_rate": 5.3429495212284665e-05,
      "loss": 0.7099,
      "step": 9568
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.937739610671997,
      "learning_rate": 5.3178717422104144e-05,
      "loss": 0.6366,
      "step": 9600
    },
    {
      "epoch": 0.9632,
      "grad_norm": 1.8911631107330322,
      "learning_rate": 5.2927859318023073e-05,
      "loss": 0.6813,
      "step": 9632
    },
    {
      "epoch": 0.9664,
      "grad_norm": 1.1599578857421875,
      "learning_rate": 5.2676927238255946e-05,
      "loss": 0.7155,
      "step": 9664
    },
    {
      "epoch": 0.9696,
      "grad_norm": 1.2809479236602783,
      "learning_rate": 5.242592752288632e-05,
      "loss": 0.7051,
      "step": 9696
    },
    {
      "epoch": 0.9728,
      "grad_norm": 2.0790278911590576,
      "learning_rate": 5.2174866513706646e-05,
      "loss": 0.7387,
      "step": 9728
    },
    {
      "epoch": 0.976,
      "grad_norm": 1.0074536800384521,
      "learning_rate": 5.1923750554058084e-05,
      "loss": 0.6751,
      "step": 9760
    },
    {
      "epoch": 0.9792,
      "grad_norm": 1.3937727212905884,
      "learning_rate": 5.16725859886701e-05,
      "loss": 0.6902,
      "step": 9792
    },
    {
      "epoch": 0.9824,
      "grad_norm": 0.8866567015647888,
      "learning_rate": 5.142137916350028e-05,
      "loss": 0.7443,
      "step": 9824
    },
    {
      "epoch": 0.9856,
      "grad_norm": 0.857765793800354,
      "learning_rate": 5.1170136425573956e-05,
      "loss": 0.7032,
      "step": 9856
    },
    {
      "epoch": 0.9888,
      "grad_norm": 0.6846195459365845,
      "learning_rate": 5.0918864122823816e-05,
      "loss": 0.6508,
      "step": 9888
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.9779634475708008,
      "learning_rate": 5.066756860392956e-05,
      "loss": 0.7161,
      "step": 9920
    },
    {
      "epoch": 0.9952,
      "grad_norm": 1.3198580741882324,
      "learning_rate": 5.0416256218157476e-05,
      "loss": 0.6885,
      "step": 9952
    },
    {
      "epoch": 0.9984,
      "grad_norm": 0.8396392464637756,
      "learning_rate": 5.0164933315199955e-05,
      "loss": 0.7511,
      "step": 9984
    }
  ],
  "logging_steps": 32,
  "max_steps": 20000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.0363220663791616e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}