{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.6144764723689078,
  "eval_steps": 500,
  "global_step": 12000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 4.9997567688496474e-05,
      "loss": 1.6283,
      "step": 100
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.9990172715142793e-05,
      "loss": 0.068,
      "step": 200
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.997781629993153e-05,
      "loss": 0.0404,
      "step": 300
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.9960500896052476e-05,
      "loss": 0.0626,
      "step": 400
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.993822994123172e-05,
      "loss": 0.0428,
      "step": 500
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.9911007857049264e-05,
      "loss": 0.0265,
      "step": 600
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.987884004806111e-05,
      "loss": 0.0263,
      "step": 700
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.984173290072626e-05,
      "loss": 0.03,
      "step": 800
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.979969378213884e-05,
      "loss": 0.0253,
      "step": 900
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.975273103856537e-05,
      "loss": 0.038,
      "step": 1000
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.970085399378785e-05,
      "loss": 0.0322,
      "step": 1100
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.964407294725254e-05,
      "loss": 0.0296,
      "step": 1200
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.958239917202523e-05,
      "loss": 0.0295,
      "step": 1300
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.9515844912553106e-05,
      "loss": 0.0171,
      "step": 1400
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.944442338223378e-05,
      "loss": 0.0205,
      "step": 1500
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.9368148760792e-05,
      "loss": 0.029,
      "step": 1600
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.9287036191464414e-05,
      "loss": 0.0172,
      "step": 1700
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.9201101777993116e-05,
      "loss": 0.0235,
      "step": 1800
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.911036258142853e-05,
      "loss": 0.0168,
      "step": 1900
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.9014836616742065e-05,
      "loss": 0.0227,
      "step": 2000
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.891454284924965e-05,
      "loss": 0.0227,
      "step": 2100
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.8810575041794e-05,
      "loss": 0.0173,
      "step": 2200
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.8700853511049656e-05,
      "loss": 0.0227,
      "step": 2300
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.858642651436035e-05,
      "loss": 0.0144,
      "step": 2400
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.846731676957191e-05,
      "loss": 0.03,
      "step": 2500
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.834354792422293e-05,
      "loss": 0.0189,
      "step": 2600
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.821514455084985e-05,
      "loss": 0.0111,
      "step": 2700
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.8082132142108465e-05,
      "loss": 0.0152,
      "step": 2800
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.794453710571272e-05,
      "loss": 0.0182,
      "step": 2900
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.780238675919182e-05,
      "loss": 0.0111,
      "step": 3000
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.765570932446672e-05,
      "loss": 0.0138,
      "step": 3100
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.75045339222471e-05,
      "loss": 0.0256,
      "step": 3200
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.734889056624983e-05,
      "loss": 0.0199,
      "step": 3300
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.718881015724017e-05,
      "loss": 0.0159,
      "step": 3400
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.702432447689692e-05,
      "loss": 0.0121,
      "step": 3500
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.6855466181502544e-05,
      "loss": 0.0181,
      "step": 3600
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.66822687954598e-05,
      "loss": 0.0148,
      "step": 3700
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.65047667046359e-05,
      "loss": 0.0152,
      "step": 3800
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.632299514953571e-05,
      "loss": 0.0103,
      "step": 3900
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.613699021830524e-05,
      "loss": 0.0217,
      "step": 4000
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.59467888395669e-05,
      "loss": 0.0142,
      "step": 4100
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.575242877508777e-05,
      "loss": 0.0151,
      "step": 4200
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.5553948612282607e-05,
      "loss": 0.0164,
      "step": 4300
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.5353433433160075e-05,
      "loss": 0.0095,
      "step": 4400
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.514687230313633e-05,
      "loss": 0.015,
      "step": 4500
    },
    {
      "epoch": 0.62,
      "learning_rate": 4.49363112993749e-05,
      "loss": 0.015,
      "step": 4600
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.472179222575561e-05,
      "loss": 0.0181,
      "step": 4700
    },
    {
      "epoch": 0.65,
      "learning_rate": 4.4503357671976574e-05,
      "loss": 0.0175,
      "step": 4800
    },
    {
      "epoch": 0.66,
      "learning_rate": 4.428105100509852e-05,
      "loss": 0.0069,
      "step": 4900
    },
    {
      "epoch": 0.67,
      "learning_rate": 4.4054916360934957e-05,
      "loss": 0.0108,
      "step": 5000
    },
    {
      "epoch": 0.69,
      "learning_rate": 4.3824998635289594e-05,
      "loss": 0.0204,
      "step": 5100
    },
    {
      "epoch": 0.7,
      "learning_rate": 4.3591343475042946e-05,
      "loss": 0.0192,
      "step": 5200
    },
    {
      "epoch": 0.71,
      "learning_rate": 4.3353997269089774e-05,
      "loss": 0.0086,
      "step": 5300
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.311300713912926e-05,
      "loss": 0.0179,
      "step": 5400
    },
    {
      "epoch": 0.74,
      "learning_rate": 4.286842093030963e-05,
      "loss": 0.0142,
      "step": 5500
    },
    {
      "epoch": 0.75,
      "learning_rate": 4.262028720172921e-05,
      "loss": 0.0187,
      "step": 5600
    },
    {
      "epoch": 0.77,
      "learning_rate": 4.2368655216795735e-05,
      "loss": 0.0176,
      "step": 5700
    },
    {
      "epoch": 0.78,
      "learning_rate": 4.211357493344573e-05,
      "loss": 0.0145,
      "step": 5800
    },
    {
      "epoch": 0.79,
      "learning_rate": 4.185509699422615e-05,
      "loss": 0.0087,
      "step": 5900
    },
    {
      "epoch": 0.81,
      "learning_rate": 4.1593272716239985e-05,
      "loss": 0.0136,
      "step": 6000
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.132815408095797e-05,
      "loss": 0.0137,
      "step": 6100
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.105979372389847e-05,
      "loss": 0.011,
      "step": 6200
    },
    {
      "epoch": 0.85,
      "learning_rate": 4.0788244924177365e-05,
      "loss": 0.024,
      "step": 6300
    },
    {
      "epoch": 0.86,
      "learning_rate": 4.0513561593930325e-05,
      "loss": 0.0067,
      "step": 6400
    },
    {
      "epoch": 0.87,
      "learning_rate": 4.023579826760924e-05,
      "loss": 0.0102,
      "step": 6500
    },
    {
      "epoch": 0.89,
      "learning_rate": 3.995501009115527e-05,
      "loss": 0.013,
      "step": 6600
    },
    {
      "epoch": 0.9,
      "learning_rate": 3.9674104896418544e-05,
      "loss": 0.0133,
      "step": 6700
    },
    {
      "epoch": 0.91,
      "learning_rate": 3.9387463695523555e-05,
      "loss": 0.0087,
      "step": 6800
    },
    {
      "epoch": 0.93,
      "learning_rate": 3.9097966069208193e-05,
      "loss": 0.01,
      "step": 6900
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.880566949309125e-05,
      "loss": 0.0113,
      "step": 7000
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.8513595749088484e-05,
      "loss": 0.0101,
      "step": 7100
    },
    {
      "epoch": 0.97,
      "learning_rate": 3.821590244319273e-05,
      "loss": 0.0092,
      "step": 7200
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.791558530862982e-05,
      "loss": 0.0176,
      "step": 7300
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.76127039690774e-05,
      "loss": 0.0114,
      "step": 7400
    },
    {
      "epoch": 1.01,
      "learning_rate": 3.7307318557299355e-05,
      "loss": 0.0083,
      "step": 7500
    },
    {
      "epoch": 1.02,
      "learning_rate": 3.699948970320738e-05,
      "loss": 0.0067,
      "step": 7600
    },
    {
      "epoch": 1.04,
      "learning_rate": 3.668927852182374e-05,
      "loss": 0.0021,
      "step": 7700
    },
    {
      "epoch": 1.05,
      "learning_rate": 3.637674660114774e-05,
      "loss": 0.0087,
      "step": 7800
    },
    {
      "epoch": 1.06,
      "learning_rate": 3.606195598992832e-05,
      "loss": 0.005,
      "step": 7900
    },
    {
      "epoch": 1.08,
      "learning_rate": 3.574496918534519e-05,
      "loss": 0.014,
      "step": 8000
    },
    {
      "epoch": 1.09,
      "learning_rate": 3.542584912060087e-05,
      "loss": 0.0073,
      "step": 8100
    },
    {
      "epoch": 1.1,
      "learning_rate": 3.510465915242623e-05,
      "loss": 0.0073,
      "step": 8200
    },
    {
      "epoch": 1.12,
      "learning_rate": 3.478146304850188e-05,
      "loss": 0.0106,
      "step": 8300
    },
    {
      "epoch": 1.13,
      "learning_rate": 3.4456324974798025e-05,
      "loss": 0.0106,
      "step": 8400
    },
    {
      "epoch": 1.14,
      "learning_rate": 3.412930948283528e-05,
      "loss": 0.0036,
      "step": 8500
    },
    {
      "epoch": 1.16,
      "learning_rate": 3.380048149686877e-05,
      "loss": 0.0051,
      "step": 8600
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.3469906300998476e-05,
      "loss": 0.0047,
      "step": 8700
    },
    {
      "epoch": 1.18,
      "learning_rate": 3.313764952620792e-05,
      "loss": 0.0035,
      "step": 8800
    },
    {
      "epoch": 1.2,
      "learning_rate": 3.280377713733412e-05,
      "loss": 0.0094,
      "step": 8900
    },
    {
      "epoch": 1.21,
      "learning_rate": 3.246835541997116e-05,
      "loss": 0.0077,
      "step": 9000
    },
    {
      "epoch": 1.22,
      "learning_rate": 3.213145096731021e-05,
      "loss": 0.0076,
      "step": 9100
    },
    {
      "epoch": 1.24,
      "learning_rate": 3.1793130666918324e-05,
      "loss": 0.0074,
      "step": 9200
    },
    {
      "epoch": 1.25,
      "learning_rate": 3.145346168745891e-05,
      "loss": 0.0072,
      "step": 9300
    },
    {
      "epoch": 1.26,
      "learning_rate": 3.111251146535634e-05,
      "loss": 0.0043,
      "step": 9400
    },
    {
      "epoch": 1.28,
      "learning_rate": 3.077377511336779e-05,
      "loss": 0.0071,
      "step": 9500
    },
    {
      "epoch": 1.29,
      "learning_rate": 3.043047683848379e-05,
      "loss": 0.0102,
      "step": 9600
    },
    {
      "epoch": 1.31,
      "learning_rate": 3.0086100419987084e-05,
      "loss": 0.005,
      "step": 9700
    },
    {
      "epoch": 1.32,
      "learning_rate": 2.974071422889685e-05,
      "loss": 0.0155,
      "step": 9800
    },
    {
      "epoch": 1.33,
      "learning_rate": 2.9394386836708154e-05,
      "loss": 0.0086,
      "step": 9900
    },
    {
      "epoch": 1.35,
      "learning_rate": 2.9047187001778132e-05,
      "loss": 0.0048,
      "step": 10000
    },
    {
      "epoch": 1.36,
      "learning_rate": 2.8699183655674938e-05,
      "loss": 0.0085,
      "step": 10100
    },
    {
      "epoch": 1.37,
      "learning_rate": 2.835044588949241e-05,
      "loss": 0.0091,
      "step": 10200
    },
    {
      "epoch": 1.39,
      "learning_rate": 2.8001042940132995e-05,
      "loss": 0.0089,
      "step": 10300
    },
    {
      "epoch": 1.4,
      "learning_rate": 2.7651044176561748e-05,
      "loss": 0.0093,
      "step": 10400
    },
    {
      "epoch": 1.41,
      "learning_rate": 2.7300519086034166e-05,
      "loss": 0.0073,
      "step": 10500
    },
    {
      "epoch": 1.43,
      "learning_rate": 2.6949537260300395e-05,
      "loss": 0.0029,
      "step": 10600
    },
    {
      "epoch": 1.44,
      "learning_rate": 2.6598168381788864e-05,
      "loss": 0.0027,
      "step": 10700
    },
    {
      "epoch": 1.45,
      "learning_rate": 2.6246482209771783e-05,
      "loss": 0.011,
      "step": 10800
    },
    {
      "epoch": 1.47,
      "learning_rate": 2.5894548566515485e-05,
      "loss": 0.0108,
      "step": 10900
    },
    {
      "epoch": 1.48,
      "learning_rate": 2.554243732341816e-05,
      "loss": 0.0033,
      "step": 11000
    },
    {
      "epoch": 1.49,
      "learning_rate": 2.519021838713791e-05,
      "loss": 0.0048,
      "step": 11100
    },
    {
      "epoch": 1.51,
      "learning_rate": 2.483796168571378e-05,
      "loss": 0.0015,
      "step": 11200
    },
    {
      "epoch": 1.52,
      "learning_rate": 2.4485737154682554e-05,
      "loss": 0.0053,
      "step": 11300
    },
    {
      "epoch": 1.53,
      "learning_rate": 2.4133614723194047e-05,
      "loss": 0.0043,
      "step": 11400
    },
    {
      "epoch": 1.55,
      "learning_rate": 2.378166430012766e-05,
      "loss": 0.0086,
      "step": 11500
    },
    {
      "epoch": 1.56,
      "learning_rate": 2.3429955760212966e-05,
      "loss": 0.0077,
      "step": 11600
    },
    {
      "epoch": 1.57,
      "learning_rate": 2.307855893015712e-05,
      "loss": 0.0066,
      "step": 11700
    },
    {
      "epoch": 1.59,
      "learning_rate": 2.272754357478172e-05,
      "loss": 0.0055,
      "step": 11800
    },
    {
      "epoch": 1.6,
      "learning_rate": 2.237697938317202e-05,
      "loss": 0.0067,
      "step": 11900
    },
    {
      "epoch": 1.61,
      "learning_rate": 2.202693595484113e-05,
      "loss": 0.0082,
      "step": 12000
    }
  ],
  "logging_steps": 100,
  "max_steps": 22296,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 3000,
  "total_flos": 3.353181220065116e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}