{
  "best_metric": 0.939085176178107,
  "best_model_checkpoint": "w-attempt11_150M_200k_weighted_accuracy_3/checkpoint-68620",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 68620,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07,
      "learning_rate": 9.963567473039931e-05,
      "loss": -45291724.8,
      "step": 500
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.92713494607986e-05,
      "loss": -45805633.536,
      "step": 1000
    },
    {
      "epoch": 0.22,
      "learning_rate": 9.890702419119791e-05,
      "loss": -45895168.0,
      "step": 1500
    },
    {
      "epoch": 0.29,
      "learning_rate": 9.854269892159721e-05,
      "loss": -46820265.984,
      "step": 2000
    },
    {
      "epoch": 0.36,
      "learning_rate": 9.81783736519965e-05,
      "loss": -45716246.528,
      "step": 2500
    },
    {
      "epoch": 0.44,
      "learning_rate": 9.781404838239581e-05,
      "loss": -47373447.168,
      "step": 3000
    },
    {
      "epoch": 0.51,
      "learning_rate": 9.744972311279511e-05,
      "loss": -46621626.368,
      "step": 3500
    },
    {
      "epoch": 0.58,
      "learning_rate": 9.70853978431944e-05,
      "loss": -46436790.272,
      "step": 4000
    },
    {
      "epoch": 0.66,
      "learning_rate": 9.672107257359371e-05,
      "loss": -48448180.224,
      "step": 4500
    },
    {
      "epoch": 0.73,
      "learning_rate": 9.635674730399302e-05,
      "loss": -49018150.912,
      "step": 5000
    },
    {
      "epoch": 0.8,
      "learning_rate": 9.599242203439231e-05,
      "loss": -47903965.184,
      "step": 5500
    },
    {
      "epoch": 0.87,
      "learning_rate": 9.562809676479161e-05,
      "loss": -48040722.432,
      "step": 6000
    },
    {
      "epoch": 0.95,
      "learning_rate": 9.52637714951909e-05,
      "loss": -48426446.848,
      "step": 6500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9390846650576595,
      "eval_loss": -48866836.0,
      "eval_runtime": 140.4892,
      "eval_samples_per_second": 32.565,
      "eval_steps_per_second": 5.431,
      "step": 6862
    },
    {
      "epoch": 1.02,
      "learning_rate": 9.489944622559021e-05,
      "loss": -48707878.912,
      "step": 7000
    },
    {
      "epoch": 1.09,
      "learning_rate": 9.45351209559895e-05,
      "loss": -48791568.384,
      "step": 7500
    },
    {
      "epoch": 1.17,
      "learning_rate": 9.417079568638881e-05,
      "loss": -49470013.44,
      "step": 8000
    },
    {
      "epoch": 1.24,
      "learning_rate": 9.380647041678811e-05,
      "loss": -49581268.992,
      "step": 8500
    },
    {
      "epoch": 1.31,
      "learning_rate": 9.34421451471874e-05,
      "loss": -49812631.552,
      "step": 9000
    },
    {
      "epoch": 1.38,
      "learning_rate": 9.307781987758671e-05,
      "loss": -51050233.856,
      "step": 9500
    },
    {
      "epoch": 1.46,
      "learning_rate": 9.271349460798602e-05,
      "loss": -49380384.768,
      "step": 10000
    },
    {
      "epoch": 1.53,
      "learning_rate": 9.234916933838531e-05,
      "loss": -50849955.84,
      "step": 10500
    },
    {
      "epoch": 1.6,
      "learning_rate": 9.198484406878461e-05,
      "loss": -50895675.392,
      "step": 11000
    },
    {
      "epoch": 1.68,
      "learning_rate": 9.162051879918392e-05,
      "loss": -51525443.584,
      "step": 11500
    },
    {
      "epoch": 1.75,
      "learning_rate": 9.125619352958321e-05,
      "loss": -51623043.072,
      "step": 12000
    },
    {
      "epoch": 1.82,
      "learning_rate": 9.089186825998251e-05,
      "loss": -51431817.216,
      "step": 12500
    },
    {
      "epoch": 1.89,
      "learning_rate": 9.052754299038182e-05,
      "loss": -52294926.336,
      "step": 13000
    },
    {
      "epoch": 1.97,
      "learning_rate": 9.016321772078111e-05,
      "loss": -52310294.528,
      "step": 13500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9390846650576595,
      "eval_loss": -52481936.0,
      "eval_runtime": 140.6751,
      "eval_samples_per_second": 32.522,
      "eval_steps_per_second": 5.424,
      "step": 13724
    },
    {
      "epoch": 2.04,
      "learning_rate": 8.979889245118042e-05,
      "loss": -52849963.008,
      "step": 14000
    },
    {
      "epoch": 2.11,
      "learning_rate": 8.943456718157972e-05,
      "loss": -51996053.504,
      "step": 14500
    },
    {
      "epoch": 2.19,
      "learning_rate": 8.907024191197903e-05,
      "loss": -52932702.208,
      "step": 15000
    },
    {
      "epoch": 2.26,
      "learning_rate": 8.870591664237832e-05,
      "loss": -53120417.792,
      "step": 15500
    },
    {
      "epoch": 2.33,
      "learning_rate": 8.834159137277762e-05,
      "loss": -53047812.096,
      "step": 16000
    },
    {
      "epoch": 2.4,
      "learning_rate": 8.797726610317693e-05,
      "loss": -53340852.224,
      "step": 16500
    },
    {
      "epoch": 2.48,
      "learning_rate": 8.761294083357622e-05,
      "loss": -54068551.68,
      "step": 17000
    },
    {
      "epoch": 2.55,
      "learning_rate": 8.724861556397553e-05,
      "loss": -54157238.272,
      "step": 17500
    },
    {
      "epoch": 2.62,
      "learning_rate": 8.688429029437483e-05,
      "loss": -55157616.64,
      "step": 18000
    },
    {
      "epoch": 2.7,
      "learning_rate": 8.651996502477412e-05,
      "loss": -54994378.752,
      "step": 18500
    },
    {
      "epoch": 2.77,
      "learning_rate": 8.615563975517343e-05,
      "loss": -55051157.504,
      "step": 19000
    },
    {
      "epoch": 2.84,
      "learning_rate": 8.579131448557272e-05,
      "loss": -55962931.2,
      "step": 19500
    },
    {
      "epoch": 2.91,
      "learning_rate": 8.542698921597203e-05,
      "loss": -55627776.0,
      "step": 20000
    },
    {
      "epoch": 2.99,
      "learning_rate": 8.506266394637132e-05,
      "loss": -56315912.192,
      "step": 20500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.925901846473729,
      "eval_loss": -56022420.0,
      "eval_runtime": 140.5361,
      "eval_samples_per_second": 32.554,
      "eval_steps_per_second": 5.429,
      "step": 20586
    },
    {
      "epoch": 3.06,
      "learning_rate": 8.469833867677062e-05,
      "loss": -54493945.856,
      "step": 21000
    },
    {
      "epoch": 3.13,
      "learning_rate": 8.433401340716991e-05,
      "loss": -56493727.744,
      "step": 21500
    },
    {
      "epoch": 3.21,
      "learning_rate": 8.396968813756922e-05,
      "loss": -55263518.72,
      "step": 22000
    },
    {
      "epoch": 3.28,
      "learning_rate": 8.360536286796853e-05,
      "loss": -57487265.792,
      "step": 22500
    },
    {
      "epoch": 3.35,
      "learning_rate": 8.324103759836782e-05,
      "loss": -56985743.36,
      "step": 23000
    },
    {
      "epoch": 3.42,
      "learning_rate": 8.287671232876712e-05,
      "loss": -58815537.152,
      "step": 23500
    },
    {
      "epoch": 3.5,
      "learning_rate": 8.251238705916643e-05,
      "loss": -57407537.152,
      "step": 24000
    },
    {
      "epoch": 3.57,
      "learning_rate": 8.214806178956573e-05,
      "loss": -57564954.624,
      "step": 24500
    },
    {
      "epoch": 3.64,
      "learning_rate": 8.178373651996502e-05,
      "loss": -58936123.392,
      "step": 25000
    },
    {
      "epoch": 3.72,
      "learning_rate": 8.141941125036433e-05,
      "loss": -57919283.2,
      "step": 25500
    },
    {
      "epoch": 3.79,
      "learning_rate": 8.105508598076364e-05,
      "loss": -59147018.24,
      "step": 26000
    },
    {
      "epoch": 3.86,
      "learning_rate": 8.069076071116293e-05,
      "loss": -60339425.28,
      "step": 26500
    },
    {
      "epoch": 3.93,
      "learning_rate": 8.032643544156223e-05,
      "loss": -59097784.32,
      "step": 27000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.2777975411017508,
      "eval_loss": -59458916.0,
      "eval_runtime": 140.4799,
      "eval_samples_per_second": 32.567,
      "eval_steps_per_second": 5.431,
      "step": 27448
    },
    {
      "epoch": 4.01,
      "learning_rate": 7.996211017196154e-05,
      "loss": -59645280.256,
      "step": 27500
    },
    {
      "epoch": 4.08,
      "learning_rate": 7.959778490236083e-05,
      "loss": -59958251.52,
      "step": 28000
    },
    {
      "epoch": 4.15,
      "learning_rate": 7.923345963276013e-05,
      "loss": -59766607.872,
      "step": 28500
    },
    {
      "epoch": 4.23,
      "learning_rate": 7.886913436315944e-05,
      "loss": -59114577.92,
      "step": 29000
    },
    {
      "epoch": 4.3,
      "learning_rate": 7.850480909355873e-05,
      "loss": -60154269.696,
      "step": 29500
    },
    {
      "epoch": 4.37,
      "learning_rate": 7.814048382395804e-05,
      "loss": -60939288.576,
      "step": 30000
    },
    {
      "epoch": 4.44,
      "learning_rate": 7.777615855435734e-05,
      "loss": -61184118.784,
      "step": 30500
    },
    {
      "epoch": 4.52,
      "learning_rate": 7.741183328475663e-05,
      "loss": -60580569.088,
      "step": 31000
    },
    {
      "epoch": 4.59,
      "learning_rate": 7.704750801515594e-05,
      "loss": -61123616.768,
      "step": 31500
    },
    {
      "epoch": 4.66,
      "learning_rate": 7.668318274555523e-05,
      "loss": -61707681.792,
      "step": 32000
    },
    {
      "epoch": 4.74,
      "learning_rate": 7.631885747595454e-05,
      "loss": -62141902.848,
      "step": 32500
    },
    {
      "epoch": 4.81,
      "learning_rate": 7.595453220635383e-05,
      "loss": -62257491.968,
      "step": 33000
    },
    {
      "epoch": 4.88,
      "learning_rate": 7.559020693675313e-05,
      "loss": -61910138.88,
      "step": 33500
    },
    {
      "epoch": 4.95,
      "learning_rate": 7.522588166715244e-05,
      "loss": -62387458.048,
      "step": 34000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9374904803816638,
      "eval_loss": -62782156.0,
      "eval_runtime": 140.6016,
      "eval_samples_per_second": 32.539,
      "eval_steps_per_second": 5.427,
      "step": 34310
    },
    {
      "epoch": 5.03,
      "learning_rate": 7.486155639755173e-05,
      "loss": -62778101.76,
      "step": 34500
    },
    {
      "epoch": 5.1,
      "learning_rate": 7.449723112795104e-05,
      "loss": -61654474.752,
      "step": 35000
    },
    {
      "epoch": 5.17,
      "learning_rate": 7.413290585835034e-05,
      "loss": -63444623.36,
      "step": 35500
    },
    {
      "epoch": 5.25,
      "learning_rate": 7.376858058874963e-05,
      "loss": -63924170.752,
      "step": 36000
    },
    {
      "epoch": 5.32,
      "learning_rate": 7.340425531914894e-05,
      "loss": -64446738.432,
      "step": 36500
    },
    {
      "epoch": 5.39,
      "learning_rate": 7.303993004954824e-05,
      "loss": -63802613.76,
      "step": 37000
    },
    {
      "epoch": 5.46,
      "learning_rate": 7.267560477994753e-05,
      "loss": -64431120.384,
      "step": 37500
    },
    {
      "epoch": 5.54,
      "learning_rate": 7.231127951034684e-05,
      "loss": -64963448.832,
      "step": 38000
    },
    {
      "epoch": 5.61,
      "learning_rate": 7.194695424074615e-05,
      "loss": -65131409.408,
      "step": 38500
    },
    {
      "epoch": 5.68,
      "learning_rate": 7.158262897114544e-05,
      "loss": -65691656.192,
      "step": 39000
    },
    {
      "epoch": 5.76,
      "learning_rate": 7.121830370154474e-05,
      "loss": -64785408.0,
      "step": 39500
    },
    {
      "epoch": 5.83,
      "learning_rate": 7.085397843194405e-05,
      "loss": -65206128.64,
      "step": 40000
    },
    {
      "epoch": 5.9,
      "learning_rate": 7.048965316234334e-05,
      "loss": -66662268.928,
      "step": 40500
    },
    {
      "epoch": 5.97,
      "learning_rate": 7.012532789274264e-05,
      "loss": -66883559.424,
      "step": 41000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.0615205015522728,
      "eval_loss": -65968124.0,
      "eval_runtime": 140.5904,
      "eval_samples_per_second": 32.541,
      "eval_steps_per_second": 5.427,
      "step": 41172
    },
    {
      "epoch": 6.05,
      "learning_rate": 6.976100262314195e-05,
      "loss": -66262577.152,
      "step": 41500
    },
    {
      "epoch": 6.12,
      "learning_rate": 6.939667735354124e-05,
      "loss": -66138095.616,
      "step": 42000
    },
    {
      "epoch": 6.19,
      "learning_rate": 6.903235208394055e-05,
      "loss": -66000941.056,
      "step": 42500
    },
    {
      "epoch": 6.27,
      "learning_rate": 6.866802681433985e-05,
      "loss": -66043711.488,
      "step": 43000
    },
    {
      "epoch": 6.34,
      "learning_rate": 6.830370154473916e-05,
      "loss": -67294355.456,
      "step": 43500
    },
    {
      "epoch": 6.41,
      "learning_rate": 6.793937627513845e-05,
      "loss": -66696265.728,
      "step": 44000
    },
    {
      "epoch": 6.48,
      "learning_rate": 6.757505100553775e-05,
      "loss": -67892416.512,
      "step": 44500
    },
    {
      "epoch": 6.56,
      "learning_rate": 6.721072573593705e-05,
      "loss": -68610072.576,
      "step": 45000
    },
    {
      "epoch": 6.63,
      "learning_rate": 6.684640046633635e-05,
      "loss": -68491333.632,
      "step": 45500
    },
    {
      "epoch": 6.7,
      "learning_rate": 6.648207519673564e-05,
      "loss": -67069939.712,
      "step": 46000
    },
    {
      "epoch": 6.78,
      "learning_rate": 6.611774992713495e-05,
      "loss": -68588945.408,
      "step": 46500
    },
    {
      "epoch": 6.85,
      "learning_rate": 6.575342465753424e-05,
      "loss": -68691898.368,
      "step": 47000
    },
    {
      "epoch": 6.92,
      "learning_rate": 6.538909938793355e-05,
      "loss": -68925136.896,
      "step": 47500
    },
    {
      "epoch": 7.0,
      "learning_rate": 6.502477411833285e-05,
      "loss": -68470861.824,
      "step": 48000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.09869071386148431,
      "eval_loss": -69001016.0,
      "eval_runtime": 140.4928,
      "eval_samples_per_second": 32.564,
      "eval_steps_per_second": 5.431,
      "step": 48034
    },
    {
      "epoch": 7.07,
      "learning_rate": 6.466044884873214e-05,
      "loss": -69397372.928,
      "step": 48500
    },
    {
      "epoch": 7.14,
      "learning_rate": 6.429612357913145e-05,
      "loss": -68643381.248,
      "step": 49000
    },
    {
      "epoch": 7.21,
      "learning_rate": 6.393179830953075e-05,
      "loss": -69129224.192,
      "step": 49500
    },
    {
      "epoch": 7.29,
      "learning_rate": 6.356747303993004e-05,
      "loss": -68606349.312,
      "step": 50000
    },
    {
      "epoch": 7.36,
      "learning_rate": 6.320314777032935e-05,
      "loss": -69869944.832,
      "step": 50500
    },
    {
      "epoch": 7.43,
      "learning_rate": 6.283882250072866e-05,
      "loss": -69662130.176,
      "step": 51000
    },
    {
      "epoch": 7.51,
      "learning_rate": 6.247449723112795e-05,
      "loss": -70071443.456,
      "step": 51500
    },
    {
      "epoch": 7.58,
      "learning_rate": 6.211017196152725e-05,
      "loss": -71158382.592,
      "step": 52000
    },
    {
      "epoch": 7.65,
      "learning_rate": 6.174584669192656e-05,
      "loss": -71736705.024,
      "step": 52500
    },
    {
      "epoch": 7.72,
      "learning_rate": 6.138152142232586e-05,
      "loss": -71263109.12,
      "step": 53000
    },
    {
      "epoch": 7.8,
      "learning_rate": 6.1017196152725155e-05,
      "loss": -71382908.928,
      "step": 53500
    },
    {
      "epoch": 7.87,
      "learning_rate": 6.065287088312446e-05,
      "loss": -72212193.28,
      "step": 54000
    },
    {
      "epoch": 7.94,
      "learning_rate": 6.028854561352376e-05,
      "loss": -71681441.792,
      "step": 54500
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9390846650576595,
      "eval_loss": -71871320.0,
      "eval_runtime": 140.58,
      "eval_samples_per_second": 32.544,
      "eval_steps_per_second": 5.428,
      "step": 54896
    },
    {
      "epoch": 8.02,
      "learning_rate": 5.992422034392306e-05,
      "loss": -72663113.728,
      "step": 55000
    },
    {
      "epoch": 8.09,
      "learning_rate": 5.9559895074322355e-05,
      "loss": -71982637.056,
      "step": 55500
    },
    {
      "epoch": 8.16,
      "learning_rate": 5.919556980472166e-05,
      "loss": -72483135.488,
      "step": 56000
    },
    {
      "epoch": 8.23,
      "learning_rate": 5.883124453512095e-05,
      "loss": -72597225.472,
      "step": 56500
    },
    {
      "epoch": 8.31,
      "learning_rate": 5.846691926552026e-05,
      "loss": -74057515.008,
      "step": 57000
    },
    {
      "epoch": 8.38,
      "learning_rate": 5.810259399591956e-05,
      "loss": -72980668.416,
      "step": 57500
    },
    {
      "epoch": 8.45,
      "learning_rate": 5.7738268726318855e-05,
      "loss": -72482070.528,
      "step": 58000
    },
    {
      "epoch": 8.53,
      "learning_rate": 5.737394345671816e-05,
      "loss": -73420226.56,
      "step": 58500
    },
    {
      "epoch": 8.6,
      "learning_rate": 5.7009618187117465e-05,
      "loss": -73117335.552,
      "step": 59000
    },
    {
      "epoch": 8.67,
      "learning_rate": 5.664529291751676e-05,
      "loss": -75365900.288,
      "step": 59500
    },
    {
      "epoch": 8.74,
      "learning_rate": 5.628096764791606e-05,
      "loss": -73344704.512,
      "step": 60000
    },
    {
      "epoch": 8.82,
      "learning_rate": 5.591664237831537e-05,
      "loss": -74256801.792,
      "step": 60500
    },
    {
      "epoch": 8.89,
      "learning_rate": 5.555231710871466e-05,
      "loss": -74018381.824,
      "step": 61000
    },
    {
      "epoch": 8.96,
      "learning_rate": 5.5187991839113965e-05,
      "loss": -74225754.112,
      "step": 61500
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8926442611907266,
      "eval_loss": -74553696.0,
      "eval_runtime": 140.5145,
      "eval_samples_per_second": 32.559,
      "eval_steps_per_second": 5.43,
      "step": 61758
    },
    {
      "epoch": 9.04,
      "learning_rate": 5.482366656951326e-05,
      "loss": -74975141.888,
      "step": 62000
    },
    {
      "epoch": 9.11,
      "learning_rate": 5.445934129991257e-05,
      "loss": -75021574.144,
      "step": 62500
    },
    {
      "epoch": 9.18,
      "learning_rate": 5.409501603031186e-05,
      "loss": -74001276.928,
      "step": 63000
    },
    {
      "epoch": 9.25,
      "learning_rate": 5.3730690760711165e-05,
      "loss": -75256430.592,
      "step": 63500
    },
    {
      "epoch": 9.33,
      "learning_rate": 5.336636549111047e-05,
      "loss": -75693981.696,
      "step": 64000
    },
    {
      "epoch": 9.4,
      "learning_rate": 5.300204022150976e-05,
      "loss": -75368284.16,
      "step": 64500
    },
    {
      "epoch": 9.47,
      "learning_rate": 5.263771495190907e-05,
      "loss": -76256927.744,
      "step": 65000
    },
    {
      "epoch": 9.55,
      "learning_rate": 5.227338968230837e-05,
      "loss": -75650670.592,
      "step": 65500
    },
    {
      "epoch": 9.62,
      "learning_rate": 5.1909064412707665e-05,
      "loss": -75926216.704,
      "step": 66000
    },
    {
      "epoch": 9.69,
      "learning_rate": 5.154473914310697e-05,
      "loss": -74388004.864,
      "step": 66500
    },
    {
      "epoch": 9.76,
      "learning_rate": 5.1180413873506275e-05,
      "loss": -75555790.848,
      "step": 67000
    },
    {
      "epoch": 9.84,
      "learning_rate": 5.081608860390557e-05,
      "loss": -76560392.192,
      "step": 67500
    },
    {
      "epoch": 9.91,
      "learning_rate": 5.045176333430487e-05,
      "loss": -76913991.68,
      "step": 68000
    },
    {
      "epoch": 9.98,
      "learning_rate": 5.008743806470417e-05,
      "loss": -77898022.912,
      "step": 68500
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.939085176178107,
      "eval_loss": -77044920.0,
      "eval_runtime": 140.4212,
      "eval_samples_per_second": 32.581,
      "eval_steps_per_second": 5.434,
      "step": 68620
    }
  ],
  "logging_steps": 500,
  "max_steps": 137240,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 2.7475868033053613e+17,
  "train_batch_size": 6,
  "trial_name": null,
  "trial_params": null
}