{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 137964,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 4.9818793308399294e-05,
      "loss": 3.0214,
      "step": 500
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.9637586616798586e-05,
      "loss": 2.1747,
      "step": 1000
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.9456379925197885e-05,
      "loss": 1.8966,
      "step": 1500
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.927517323359717e-05,
      "loss": 1.7367,
      "step": 2000
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.909396654199647e-05,
      "loss": 1.6498,
      "step": 2500
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.8912759850395754e-05,
      "loss": 1.5595,
      "step": 3000
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.873155315879505e-05,
      "loss": 1.4976,
      "step": 3500
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.855034646719434e-05,
      "loss": 1.4493,
      "step": 4000
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.836913977559364e-05,
      "loss": 1.4386,
      "step": 4500
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.818793308399292e-05,
      "loss": 1.4008,
      "step": 5000
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.800672639239222e-05,
      "loss": 1.3648,
      "step": 5500
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.782551970079151e-05,
      "loss": 1.332,
      "step": 6000
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.7644313009190805e-05,
      "loss": 1.31,
      "step": 6500
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.74631063175901e-05,
      "loss": 1.3129,
      "step": 7000
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.728189962598939e-05,
      "loss": 1.2762,
      "step": 7500
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.710069293438869e-05,
      "loss": 1.2457,
      "step": 8000
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.691948624278797e-05,
      "loss": 1.2464,
      "step": 8500
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.673827955118727e-05,
      "loss": 1.2334,
      "step": 9000
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.655707285958656e-05,
      "loss": 1.2099,
      "step": 9500
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.6375866167985856e-05,
      "loss": 1.1947,
      "step": 10000
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.619465947638514e-05,
      "loss": 1.1837,
      "step": 10500
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.601345278478444e-05,
      "loss": 1.1655,
      "step": 11000
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.583224609318373e-05,
      "loss": 1.1704,
      "step": 11500
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.5651039401583024e-05,
      "loss": 1.1593,
      "step": 12000
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.5469832709982316e-05,
      "loss": 1.1309,
      "step": 12500
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.528862601838161e-05,
      "loss": 1.1268,
      "step": 13000
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.5107419326780906e-05,
      "loss": 1.1334,
      "step": 13500
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.492621263518019e-05,
      "loss": 1.1136,
      "step": 14000
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.474500594357949e-05,
      "loss": 1.1129,
      "step": 14500
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.4563799251978775e-05,
      "loss": 1.0964,
      "step": 15000
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.4382592560378074e-05,
      "loss": 1.0964,
      "step": 15500
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.420138586877736e-05,
      "loss": 1.0841,
      "step": 16000
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.402017917717666e-05,
      "loss": 1.0873,
      "step": 16500
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.383897248557595e-05,
      "loss": 1.0711,
      "step": 17000
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.365776579397524e-05,
      "loss": 1.061,
      "step": 17500
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.3476559102374534e-05,
      "loss": 1.0518,
      "step": 18000
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.3295352410773826e-05,
      "loss": 1.0577,
      "step": 18500
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.311414571917312e-05,
      "loss": 1.0619,
      "step": 19000
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.293293902757241e-05,
      "loss": 1.061,
      "step": 19500
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.275173233597171e-05,
      "loss": 1.0464,
      "step": 20000
    },
    {
      "epoch": 0.45,
      "learning_rate": 4.2570525644370994e-05,
      "loss": 1.038,
      "step": 20500
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.238931895277029e-05,
      "loss": 1.027,
      "step": 21000
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.220811226116958e-05,
      "loss": 1.0309,
      "step": 21500
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.202690556956888e-05,
      "loss": 1.0258,
      "step": 22000
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.184569887796817e-05,
      "loss": 1.01,
      "step": 22500
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.166449218636746e-05,
      "loss": 1.0198,
      "step": 23000
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.148328549476675e-05,
      "loss": 1.0066,
      "step": 23500
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.1302078803166045e-05,
      "loss": 0.9983,
      "step": 24000
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.112087211156534e-05,
      "loss": 0.9934,
      "step": 24500
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.093966541996463e-05,
      "loss": 0.9902,
      "step": 25000
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.075845872836393e-05,
      "loss": 1.0013,
      "step": 25500
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.057725203676321e-05,
      "loss": 0.9806,
      "step": 26000
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.039604534516251e-05,
      "loss": 1.0136,
      "step": 26500
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.0214838653561797e-05,
      "loss": 0.9899,
      "step": 27000
    },
    {
      "epoch": 0.6,
      "learning_rate": 4.0033631961961095e-05,
      "loss": 0.9824,
      "step": 27500
    },
    {
      "epoch": 0.61,
      "learning_rate": 3.985242527036039e-05,
      "loss": 0.9735,
      "step": 28000
    },
    {
      "epoch": 0.62,
      "learning_rate": 3.967121857875968e-05,
      "loss": 0.9811,
      "step": 28500
    },
    {
      "epoch": 0.63,
      "learning_rate": 3.949001188715897e-05,
      "loss": 0.9708,
      "step": 29000
    },
    {
      "epoch": 0.64,
      "learning_rate": 3.930880519555826e-05,
      "loss": 0.9819,
      "step": 29500
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.9127598503957555e-05,
      "loss": 0.9623,
      "step": 30000
    },
    {
      "epoch": 0.66,
      "learning_rate": 3.894639181235685e-05,
      "loss": 0.9595,
      "step": 30500
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.876518512075614e-05,
      "loss": 0.965,
      "step": 31000
    },
    {
      "epoch": 0.68,
      "learning_rate": 3.858397842915543e-05,
      "loss": 0.9619,
      "step": 31500
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.840277173755473e-05,
      "loss": 0.9644,
      "step": 32000
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.8221565045954015e-05,
      "loss": 0.9442,
      "step": 32500
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.8040358354353314e-05,
      "loss": 0.9481,
      "step": 33000
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.7859151662752606e-05,
      "loss": 0.9554,
      "step": 33500
    },
    {
      "epoch": 0.74,
      "learning_rate": 3.76779449711519e-05,
      "loss": 0.9415,
      "step": 34000
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.749673827955119e-05,
      "loss": 0.9278,
      "step": 34500
    },
    {
      "epoch": 0.76,
      "learning_rate": 3.731553158795048e-05,
      "loss": 0.933,
      "step": 35000
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.7134324896349774e-05,
      "loss": 0.9278,
      "step": 35500
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.6953118204749066e-05,
      "loss": 0.9316,
      "step": 36000
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.677191151314836e-05,
      "loss": 0.9208,
      "step": 36500
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.659070482154765e-05,
      "loss": 0.9229,
      "step": 37000
    },
    {
      "epoch": 0.82,
      "learning_rate": 3.640949812994694e-05,
      "loss": 0.9311,
      "step": 37500
    },
    {
      "epoch": 0.83,
      "learning_rate": 3.6228291438346234e-05,
      "loss": 0.937,
      "step": 38000
    },
    {
      "epoch": 0.84,
      "learning_rate": 3.604708474674553e-05,
      "loss": 0.9227,
      "step": 38500
    },
    {
      "epoch": 0.85,
      "learning_rate": 3.5865878055144824e-05,
      "loss": 0.9231,
      "step": 39000
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.5684671363544116e-05,
      "loss": 0.9296,
      "step": 39500
    },
    {
      "epoch": 0.87,
      "learning_rate": 3.550346467194341e-05,
      "loss": 0.9091,
      "step": 40000
    },
    {
      "epoch": 0.88,
      "learning_rate": 3.53222579803427e-05,
      "loss": 0.901,
      "step": 40500
    },
    {
      "epoch": 0.89,
      "learning_rate": 3.514105128874199e-05,
      "loss": 0.8959,
      "step": 41000
    },
    {
      "epoch": 0.9,
      "learning_rate": 3.4959844597141284e-05,
      "loss": 0.9059,
      "step": 41500
    },
    {
      "epoch": 0.91,
      "learning_rate": 3.4778637905540576e-05,
      "loss": 0.8996,
      "step": 42000
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.459743121393987e-05,
      "loss": 0.8873,
      "step": 42500
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.441622452233916e-05,
      "loss": 0.8974,
      "step": 43000
    },
    {
      "epoch": 0.95,
      "learning_rate": 3.423501783073845e-05,
      "loss": 0.9052,
      "step": 43500
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.405381113913775e-05,
      "loss": 0.8852,
      "step": 44000
    },
    {
      "epoch": 0.97,
      "learning_rate": 3.387260444753704e-05,
      "loss": 0.9024,
      "step": 44500
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.3691397755936335e-05,
      "loss": 0.8941,
      "step": 45000
    },
    {
      "epoch": 0.99,
      "learning_rate": 3.351019106433563e-05,
      "loss": 0.9001,
      "step": 45500
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.332898437273492e-05,
      "loss": 0.8952,
      "step": 46000
    },
    {
      "epoch": 1.01,
      "learning_rate": 3.314777768113421e-05,
      "loss": 0.8079,
      "step": 46500
    },
    {
      "epoch": 1.02,
      "learning_rate": 3.29665709895335e-05,
      "loss": 0.8126,
      "step": 47000
    },
    {
      "epoch": 1.03,
      "learning_rate": 3.2785364297932795e-05,
      "loss": 0.8161,
      "step": 47500
    },
    {
      "epoch": 1.04,
      "learning_rate": 3.260415760633209e-05,
      "loss": 0.816,
      "step": 48000
    },
    {
      "epoch": 1.05,
      "learning_rate": 3.242295091473138e-05,
      "loss": 0.8252,
      "step": 48500
    },
    {
      "epoch": 1.07,
      "learning_rate": 3.224174422313067e-05,
      "loss": 0.8052,
      "step": 49000
    },
    {
      "epoch": 1.08,
      "learning_rate": 3.206053753152996e-05,
      "loss": 0.8165,
      "step": 49500
    },
    {
      "epoch": 1.09,
      "learning_rate": 3.187933083992926e-05,
      "loss": 0.8116,
      "step": 50000
    },
    {
      "epoch": 1.1,
      "learning_rate": 3.1698124148328554e-05,
      "loss": 0.8234,
      "step": 50500
    },
    {
      "epoch": 1.11,
      "learning_rate": 3.1516917456727846e-05,
      "loss": 0.8174,
      "step": 51000
    },
    {
      "epoch": 1.12,
      "learning_rate": 3.133571076512714e-05,
      "loss": 0.7963,
      "step": 51500
    },
    {
      "epoch": 1.13,
      "learning_rate": 3.115450407352643e-05,
      "loss": 0.807,
      "step": 52000
    },
    {
      "epoch": 1.14,
      "learning_rate": 3.097329738192572e-05,
      "loss": 0.8219,
      "step": 52500
    },
    {
      "epoch": 1.15,
      "learning_rate": 3.0792090690325014e-05,
      "loss": 0.8023,
      "step": 53000
    },
    {
      "epoch": 1.16,
      "learning_rate": 3.0610883998724306e-05,
      "loss": 0.784,
      "step": 53500
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.04296773071236e-05,
      "loss": 0.8099,
      "step": 54000
    },
    {
      "epoch": 1.19,
      "learning_rate": 3.024847061552289e-05,
      "loss": 0.8159,
      "step": 54500
    },
    {
      "epoch": 1.2,
      "learning_rate": 3.0067263923922185e-05,
      "loss": 0.7992,
      "step": 55000
    },
    {
      "epoch": 1.21,
      "learning_rate": 2.988605723232148e-05,
      "loss": 0.8047,
      "step": 55500
    },
    {
      "epoch": 1.22,
      "learning_rate": 2.970485054072077e-05,
      "loss": 0.8101,
      "step": 56000
    },
    {
      "epoch": 1.23,
      "learning_rate": 2.9523643849120064e-05,
      "loss": 0.8161,
      "step": 56500
    },
    {
      "epoch": 1.24,
      "learning_rate": 2.9342437157519353e-05,
      "loss": 0.7975,
      "step": 57000
    },
    {
      "epoch": 1.25,
      "learning_rate": 2.9161230465918648e-05,
      "loss": 0.8019,
      "step": 57500
    },
    {
      "epoch": 1.26,
      "learning_rate": 2.8980023774317937e-05,
      "loss": 0.7975,
      "step": 58000
    },
    {
      "epoch": 1.27,
      "learning_rate": 2.8798817082717232e-05,
      "loss": 0.7981,
      "step": 58500
    },
    {
      "epoch": 1.28,
      "learning_rate": 2.8617610391116524e-05,
      "loss": 0.7917,
      "step": 59000
    },
    {
      "epoch": 1.29,
      "learning_rate": 2.843640369951582e-05,
      "loss": 0.801,
      "step": 59500
    },
    {
      "epoch": 1.3,
      "learning_rate": 2.8255197007915108e-05,
      "loss": 0.8019,
      "step": 60000
    },
    {
      "epoch": 1.32,
      "learning_rate": 2.8073990316314403e-05,
      "loss": 0.7991,
      "step": 60500
    },
    {
      "epoch": 1.33,
      "learning_rate": 2.7892783624713692e-05,
      "loss": 0.7888,
      "step": 61000
    },
    {
      "epoch": 1.34,
      "learning_rate": 2.7711576933112987e-05,
      "loss": 0.7898,
      "step": 61500
    },
    {
      "epoch": 1.35,
      "learning_rate": 2.7530370241512283e-05,
      "loss": 0.7881,
      "step": 62000
    },
    {
      "epoch": 1.36,
      "learning_rate": 2.734916354991157e-05,
      "loss": 0.7897,
      "step": 62500
    },
    {
      "epoch": 1.37,
      "learning_rate": 2.7167956858310867e-05,
      "loss": 0.7816,
      "step": 63000
    },
    {
      "epoch": 1.38,
      "learning_rate": 2.6986750166710155e-05,
      "loss": 0.7877,
      "step": 63500
    },
    {
      "epoch": 1.39,
      "learning_rate": 2.680554347510945e-05,
      "loss": 0.7926,
      "step": 64000
    },
    {
      "epoch": 1.4,
      "learning_rate": 2.662433678350874e-05,
      "loss": 0.7904,
      "step": 64500
    },
    {
      "epoch": 1.41,
      "learning_rate": 2.6443130091908035e-05,
      "loss": 0.788,
      "step": 65000
    },
    {
      "epoch": 1.42,
      "learning_rate": 2.6261923400307327e-05,
      "loss": 0.7884,
      "step": 65500
    },
    {
      "epoch": 1.44,
      "learning_rate": 2.6080716708706622e-05,
      "loss": 0.7791,
      "step": 66000
    },
    {
      "epoch": 1.45,
      "learning_rate": 2.589951001710591e-05,
      "loss": 0.7844,
      "step": 66500
    },
    {
      "epoch": 1.46,
      "learning_rate": 2.5718303325505206e-05,
      "loss": 0.7844,
      "step": 67000
    },
    {
      "epoch": 1.47,
      "learning_rate": 2.55370966339045e-05,
      "loss": 0.7849,
      "step": 67500
    },
    {
      "epoch": 1.48,
      "learning_rate": 2.535588994230379e-05,
      "loss": 0.7804,
      "step": 68000
    },
    {
      "epoch": 1.49,
      "learning_rate": 2.5174683250703085e-05,
      "loss": 0.7773,
      "step": 68500
    },
    {
      "epoch": 1.5,
      "learning_rate": 2.4993476559102374e-05,
      "loss": 0.7851,
      "step": 69000
    },
    {
      "epoch": 1.51,
      "learning_rate": 2.481226986750167e-05,
      "loss": 0.7779,
      "step": 69500
    },
    {
      "epoch": 1.52,
      "learning_rate": 2.463106317590096e-05,
      "loss": 0.7679,
      "step": 70000
    },
    {
      "epoch": 1.53,
      "learning_rate": 2.4449856484300253e-05,
      "loss": 0.7709,
      "step": 70500
    },
    {
      "epoch": 1.54,
      "learning_rate": 2.4268649792699545e-05,
      "loss": 0.7782,
      "step": 71000
    },
    {
      "epoch": 1.55,
      "learning_rate": 2.4087443101098837e-05,
      "loss": 0.7587,
      "step": 71500
    },
    {
      "epoch": 1.57,
      "learning_rate": 2.3906236409498133e-05,
      "loss": 0.7693,
      "step": 72000
    },
    {
      "epoch": 1.58,
      "learning_rate": 2.3725029717897425e-05,
      "loss": 0.7735,
      "step": 72500
    },
    {
      "epoch": 1.59,
      "learning_rate": 2.3543823026296717e-05,
      "loss": 0.7686,
      "step": 73000
    },
    {
      "epoch": 1.6,
      "learning_rate": 2.336261633469601e-05,
      "loss": 0.7604,
      "step": 73500
    },
    {
      "epoch": 1.61,
      "learning_rate": 2.31814096430953e-05,
      "loss": 0.7562,
      "step": 74000
    },
    {
      "epoch": 1.62,
      "learning_rate": 2.3000202951494593e-05,
      "loss": 0.763,
      "step": 74500
    },
    {
      "epoch": 1.63,
      "learning_rate": 2.2818996259893888e-05,
      "loss": 0.7717,
      "step": 75000
    },
    {
      "epoch": 1.64,
      "learning_rate": 2.263778956829318e-05,
      "loss": 0.7734,
      "step": 75500
    },
    {
      "epoch": 1.65,
      "learning_rate": 2.2456582876692472e-05,
      "loss": 0.7701,
      "step": 76000
    },
    {
      "epoch": 1.66,
      "learning_rate": 2.2275376185091764e-05,
      "loss": 0.771,
      "step": 76500
    },
    {
      "epoch": 1.67,
      "learning_rate": 2.2094169493491056e-05,
      "loss": 0.7608,
      "step": 77000
    },
    {
      "epoch": 1.69,
      "learning_rate": 2.1912962801890348e-05,
      "loss": 0.7674,
      "step": 77500
    },
    {
      "epoch": 1.7,
      "learning_rate": 2.1731756110289643e-05,
      "loss": 0.7656,
      "step": 78000
    },
    {
      "epoch": 1.71,
      "learning_rate": 2.1550549418688935e-05,
      "loss": 0.7623,
      "step": 78500
    },
    {
      "epoch": 1.72,
      "learning_rate": 2.1369342727088227e-05,
      "loss": 0.7578,
      "step": 79000
    },
    {
      "epoch": 1.73,
      "learning_rate": 2.118813603548752e-05,
      "loss": 0.7728,
      "step": 79500
    },
    {
      "epoch": 1.74,
      "learning_rate": 2.100692934388681e-05,
      "loss": 0.7581,
      "step": 80000
    },
    {
      "epoch": 1.75,
      "learning_rate": 2.0825722652286106e-05,
      "loss": 0.7535,
      "step": 80500
    },
    {
      "epoch": 1.76,
      "learning_rate": 2.06445159606854e-05,
      "loss": 0.7584,
      "step": 81000
    },
    {
      "epoch": 1.77,
      "learning_rate": 2.046330926908469e-05,
      "loss": 0.7595,
      "step": 81500
    },
    {
      "epoch": 1.78,
      "learning_rate": 2.0282102577483982e-05,
      "loss": 0.7605,
      "step": 82000
    },
    {
      "epoch": 1.79,
      "learning_rate": 2.0100895885883274e-05,
      "loss": 0.7453,
      "step": 82500
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.9919689194282566e-05,
      "loss": 0.7524,
      "step": 83000
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.973848250268186e-05,
      "loss": 0.7434,
      "step": 83500
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.9557275811081154e-05,
      "loss": 0.7459,
      "step": 84000
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.9376069119480446e-05,
      "loss": 0.752,
      "step": 84500
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.9194862427879738e-05,
      "loss": 0.7463,
      "step": 85000
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.901365573627903e-05,
      "loss": 0.7503,
      "step": 85500
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.8832449044678325e-05,
      "loss": 0.7455,
      "step": 86000
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.8651242353077617e-05,
      "loss": 0.7545,
      "step": 86500
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.847003566147691e-05,
      "loss": 0.7394,
      "step": 87000
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.82888289698762e-05,
      "loss": 0.7513,
      "step": 87500
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.8107622278275493e-05,
      "loss": 0.7461,
      "step": 88000
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.7926415586674785e-05,
      "loss": 0.7406,
      "step": 88500
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.7745208895074077e-05,
      "loss": 0.7407,
      "step": 89000
    },
    {
      "epoch": 1.95,
      "learning_rate": 1.756400220347337e-05,
      "loss": 0.7425,
      "step": 89500
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.7382795511872664e-05,
      "loss": 0.7434,
      "step": 90000
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.7201588820271956e-05,
      "loss": 0.7428,
      "step": 90500
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.7020382128671248e-05,
      "loss": 0.7356,
      "step": 91000
    },
    {
      "epoch": 1.99,
      "learning_rate": 1.683917543707054e-05,
      "loss": 0.7433,
      "step": 91500
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.6657968745469836e-05,
      "loss": 0.7247,
      "step": 92000
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.6476762053869128e-05,
      "loss": 0.6601,
      "step": 92500
    },
    {
      "epoch": 2.02,
      "learning_rate": 1.629555536226842e-05,
      "loss": 0.6681,
      "step": 93000
    },
    {
      "epoch": 2.03,
      "learning_rate": 1.611434867066771e-05,
      "loss": 0.6627,
      "step": 93500
    },
    {
      "epoch": 2.04,
      "learning_rate": 1.5933141979067004e-05,
      "loss": 0.6547,
      "step": 94000
    },
    {
      "epoch": 2.05,
      "learning_rate": 1.5751935287466296e-05,
      "loss": 0.6589,
      "step": 94500
    },
    {
      "epoch": 2.07,
      "learning_rate": 1.5570728595865588e-05,
      "loss": 0.6623,
      "step": 95000
    },
    {
      "epoch": 2.08,
      "learning_rate": 1.538952190426488e-05,
      "loss": 0.6467,
      "step": 95500
    },
    {
      "epoch": 2.09,
      "learning_rate": 1.5208315212664173e-05,
      "loss": 0.6689,
      "step": 96000
    },
    {
      "epoch": 2.1,
      "learning_rate": 1.5027108521063465e-05,
      "loss": 0.6571,
      "step": 96500
    },
    {
      "epoch": 2.11,
      "learning_rate": 1.4845901829462757e-05,
      "loss": 0.6631,
      "step": 97000
    },
    {
      "epoch": 2.12,
      "learning_rate": 1.4664695137862053e-05,
      "loss": 0.673,
      "step": 97500
    },
    {
      "epoch": 2.13,
      "learning_rate": 1.4483488446261345e-05,
      "loss": 0.6565,
      "step": 98000
    },
    {
      "epoch": 2.14,
      "learning_rate": 1.4302281754660638e-05,
      "loss": 0.6649,
      "step": 98500
    },
    {
      "epoch": 2.15,
      "learning_rate": 1.412107506305993e-05,
      "loss": 0.658,
      "step": 99000
    },
    {
      "epoch": 2.16,
      "learning_rate": 1.3939868371459222e-05,
      "loss": 0.6716,
      "step": 99500
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.3758661679858514e-05,
      "loss": 0.6547,
      "step": 100000
    },
    {
      "epoch": 2.19,
      "learning_rate": 1.3577454988257806e-05,
      "loss": 0.6499,
      "step": 100500
    },
    {
      "epoch": 2.2,
      "learning_rate": 1.33962482966571e-05,
      "loss": 0.655,
      "step": 101000
    },
    {
      "epoch": 2.21,
      "learning_rate": 1.3215041605056392e-05,
      "loss": 0.6636,
      "step": 101500
    },
    {
      "epoch": 2.22,
      "learning_rate": 1.3033834913455684e-05,
      "loss": 0.6471,
      "step": 102000
    },
    {
      "epoch": 2.23,
      "learning_rate": 1.2852628221854976e-05,
      "loss": 0.6588,
      "step": 102500
    },
    {
      "epoch": 2.24,
      "learning_rate": 1.2671421530254271e-05,
      "loss": 0.6687,
      "step": 103000
    },
    {
      "epoch": 2.25,
      "learning_rate": 1.2490214838653561e-05,
      "loss": 0.6511,
      "step": 103500
    },
    {
      "epoch": 2.26,
      "learning_rate": 1.2309008147052855e-05,
      "loss": 0.657,
      "step": 104000
    },
    {
      "epoch": 2.27,
      "learning_rate": 1.2127801455452149e-05,
      "loss": 0.6578,
      "step": 104500
    },
    {
      "epoch": 2.28,
      "learning_rate": 1.194659476385144e-05,
      "loss": 0.648,
      "step": 105000
    },
    {
      "epoch": 2.29,
      "learning_rate": 1.1765388072250733e-05,
      "loss": 0.6599,
      "step": 105500
    },
    {
      "epoch": 2.3,
      "learning_rate": 1.1584181380650025e-05,
      "loss": 0.6415,
      "step": 106000
    },
    {
      "epoch": 2.32,
      "learning_rate": 1.1402974689049317e-05,
      "loss": 0.6565,
      "step": 106500
    },
    {
      "epoch": 2.33,
      "learning_rate": 1.122176799744861e-05,
      "loss": 0.6327,
      "step": 107000
    },
    {
      "epoch": 2.34,
      "learning_rate": 1.1040561305847904e-05,
      "loss": 0.6526,
      "step": 107500
    },
    {
      "epoch": 2.35,
      "learning_rate": 1.0859354614247196e-05,
      "loss": 0.6588,
      "step": 108000
    },
    {
      "epoch": 2.36,
      "learning_rate": 1.0678147922646488e-05,
      "loss": 0.6592,
      "step": 108500
    },
    {
      "epoch": 2.37,
      "learning_rate": 1.049694123104578e-05,
      "loss": 0.6426,
      "step": 109000
    },
    {
      "epoch": 2.38,
      "learning_rate": 1.0315734539445072e-05,
      "loss": 0.6492,
      "step": 109500
    },
    {
      "epoch": 2.39,
      "learning_rate": 1.0134527847844366e-05,
      "loss": 0.6459,
      "step": 110000
    },
    {
      "epoch": 2.4,
      "learning_rate": 9.953321156243658e-06,
      "loss": 0.6454,
      "step": 110500
    },
    {
      "epoch": 2.41,
      "learning_rate": 9.772114464642951e-06,
      "loss": 0.648,
      "step": 111000
    },
    {
      "epoch": 2.42,
      "learning_rate": 9.590907773042243e-06,
      "loss": 0.6466,
      "step": 111500
    },
    {
      "epoch": 2.44,
      "learning_rate": 9.409701081441535e-06,
      "loss": 0.6441,
      "step": 112000
    },
    {
      "epoch": 2.45,
      "learning_rate": 9.228494389840829e-06,
      "loss": 0.6498,
      "step": 112500
    },
    {
      "epoch": 2.46,
      "learning_rate": 9.047287698240121e-06,
      "loss": 0.6443,
      "step": 113000
    },
    {
      "epoch": 2.47,
      "learning_rate": 8.866081006639413e-06,
      "loss": 0.648,
      "step": 113500
    },
    {
      "epoch": 2.48,
      "learning_rate": 8.684874315038707e-06,
      "loss": 0.6532,
      "step": 114000
    },
    {
      "epoch": 2.49,
      "learning_rate": 8.503667623437999e-06,
      "loss": 0.6421,
      "step": 114500
    },
    {
      "epoch": 2.5,
      "learning_rate": 8.32246093183729e-06,
      "loss": 0.6498,
      "step": 115000
    },
    {
      "epoch": 2.51,
      "learning_rate": 8.141254240236584e-06,
      "loss": 0.6442,
      "step": 115500
    },
    {
      "epoch": 2.52,
      "learning_rate": 7.960047548635876e-06,
      "loss": 0.6465,
      "step": 116000
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.778840857035168e-06,
      "loss": 0.6421,
      "step": 116500
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.597634165434461e-06,
      "loss": 0.6498,
      "step": 117000
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.416427473833754e-06,
      "loss": 0.6403,
      "step": 117500
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.2352207822330476e-06,
      "loss": 0.6344,
      "step": 118000
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.0540140906323395e-06,
      "loss": 0.6552,
      "step": 118500
    },
    {
      "epoch": 2.59,
      "learning_rate": 6.8728073990316315e-06,
      "loss": 0.6366,
      "step": 119000
    },
    {
      "epoch": 2.6,
      "learning_rate": 6.691600707430924e-06,
      "loss": 0.6374,
      "step": 119500
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.510394015830216e-06,
      "loss": 0.6409,
      "step": 120000
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.329187324229509e-06,
      "loss": 0.6277,
      "step": 120500
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.147980632628802e-06,
      "loss": 0.6409,
      "step": 121000
    },
    {
      "epoch": 2.64,
      "learning_rate": 5.966773941028095e-06,
      "loss": 0.6385,
      "step": 121500
    },
    {
      "epoch": 2.65,
      "learning_rate": 5.785567249427387e-06,
      "loss": 0.6455,
      "step": 122000
    },
    {
      "epoch": 2.66,
      "learning_rate": 5.60436055782668e-06,
      "loss": 0.6401,
      "step": 122500
    },
    {
      "epoch": 2.67,
      "learning_rate": 5.4231538662259725e-06,
      "loss": 0.6365,
      "step": 123000
    },
    {
      "epoch": 2.69,
      "learning_rate": 5.2419471746252644e-06,
      "loss": 0.6337,
      "step": 123500
    },
    {
      "epoch": 2.7,
      "learning_rate": 5.060740483024557e-06,
      "loss": 0.6316,
      "step": 124000
    },
    {
      "epoch": 2.71,
      "learning_rate": 4.87953379142385e-06,
      "loss": 0.6378,
      "step": 124500
    },
    {
      "epoch": 2.72,
      "learning_rate": 4.698327099823142e-06,
      "loss": 0.6311,
      "step": 125000
    },
    {
      "epoch": 2.73,
      "learning_rate": 4.517120408222436e-06,
      "loss": 0.6373,
      "step": 125500
    },
    {
      "epoch": 2.74,
      "learning_rate": 4.335913716621728e-06,
      "loss": 0.6405,
      "step": 126000
    },
    {
      "epoch": 2.75,
      "learning_rate": 4.15470702502102e-06,
      "loss": 0.6318,
      "step": 126500
    },
    {
      "epoch": 2.76,
      "learning_rate": 3.973500333420313e-06,
      "loss": 0.63,
      "step": 127000
    },
    {
      "epoch": 2.77,
      "learning_rate": 3.7922936418196054e-06,
      "loss": 0.6439,
      "step": 127500
    },
    {
      "epoch": 2.78,
      "learning_rate": 3.611086950218898e-06,
      "loss": 0.6343,
      "step": 128000
    },
    {
      "epoch": 2.79,
      "learning_rate": 3.4298802586181906e-06,
      "loss": 0.6426,
      "step": 128500
    },
    {
      "epoch": 2.81,
      "learning_rate": 3.248673567017483e-06,
      "loss": 0.6273,
      "step": 129000
    },
    {
      "epoch": 2.82,
      "learning_rate": 3.0674668754167754e-06,
      "loss": 0.6253,
      "step": 129500
    },
    {
      "epoch": 2.83,
      "learning_rate": 2.886260183816068e-06,
      "loss": 0.6427,
      "step": 130000
    },
    {
      "epoch": 2.84,
      "learning_rate": 2.7050534922153607e-06,
      "loss": 0.6235,
      "step": 130500
    },
    {
      "epoch": 2.85,
      "learning_rate": 2.523846800614653e-06,
      "loss": 0.6254,
      "step": 131000
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.342640109013946e-06,
      "loss": 0.6308,
      "step": 131500
    },
    {
      "epoch": 2.87,
      "learning_rate": 2.1614334174132383e-06,
      "loss": 0.629,
      "step": 132000
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.9802267258125307e-06,
      "loss": 0.6336,
      "step": 132500
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.7990200342118236e-06,
      "loss": 0.6313,
      "step": 133000
    },
    {
      "epoch": 2.9,
      "learning_rate": 1.6178133426111162e-06,
      "loss": 0.6339,
      "step": 133500
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.4366066510104086e-06,
      "loss": 0.6363,
      "step": 134000
    },
    {
      "epoch": 2.92,
      "learning_rate": 1.2553999594097012e-06,
      "loss": 0.6319,
      "step": 134500
    },
    {
      "epoch": 2.94,
      "learning_rate": 1.0741932678089936e-06,
      "loss": 0.6242,
      "step": 135000
    },
    {
      "epoch": 2.95,
      "learning_rate": 8.929865762082863e-07,
      "loss": 0.6283,
      "step": 135500
    },
    {
      "epoch": 2.96,
      "learning_rate": 7.117798846075788e-07,
      "loss": 0.6442,
      "step": 136000
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.305731930068714e-07,
      "loss": 0.6303,
      "step": 136500
    },
    {
      "epoch": 2.98,
      "learning_rate": 3.4936650140616397e-07,
      "loss": 0.6224,
      "step": 137000
    },
    {
      "epoch": 2.99,
      "learning_rate": 1.681598098054565e-07,
      "loss": 0.6341,
      "step": 137500
    },
    {
      "epoch": 3.0,
      "step": 137964,
      "total_flos": 1.394478551542825e+17,
      "train_loss": 0.8443687578099376,
      "train_runtime": 46828.2642,
      "train_samples_per_second": 29.461,
      "train_steps_per_second": 2.946
    }
  ],
  "max_steps": 137964,
  "num_train_epochs": 3,
  "total_flos": 1.394478551542825e+17,
  "trial_name": null,
  "trial_params": null
}