{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 40.0,
  "global_step": 49040,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.41,
      "learning_rate": 1.9878787878787878e-05,
      "loss": 6.2879,
      "step": 500
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.9674037674037675e-05,
      "loss": 3.3765,
      "step": 1000
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.9469696969696972e-05,
      "loss": 2.5852,
      "step": 1500
    },
    {
      "epoch": 1.63,
      "learning_rate": 1.9264946764946766e-05,
      "loss": 1.663,
      "step": 2000
    },
    {
      "epoch": 2.04,
      "learning_rate": 1.9060196560196563e-05,
      "loss": 1.3474,
      "step": 2500
    },
    {
      "epoch": 2.45,
      "learning_rate": 1.8855446355446356e-05,
      "loss": 1.1403,
      "step": 3000
    },
    {
      "epoch": 2.85,
      "learning_rate": 1.8650696150696153e-05,
      "loss": 1.0676,
      "step": 3500
    },
    {
      "epoch": 3.26,
      "learning_rate": 1.8445945945945947e-05,
      "loss": 0.9461,
      "step": 4000
    },
    {
      "epoch": 3.67,
      "learning_rate": 1.8241195741195744e-05,
      "loss": 0.8768,
      "step": 4500
    },
    {
      "epoch": 4.08,
      "learning_rate": 1.8036445536445537e-05,
      "loss": 0.8472,
      "step": 5000
    },
    {
      "epoch": 4.49,
      "learning_rate": 1.7831695331695334e-05,
      "loss": 0.7818,
      "step": 5500
    },
    {
      "epoch": 4.89,
      "learning_rate": 1.7627354627354628e-05,
      "loss": 0.759,
      "step": 6000
    },
    {
      "epoch": 5.3,
      "learning_rate": 1.7422604422604425e-05,
      "loss": 0.6963,
      "step": 6500
    },
    {
      "epoch": 5.71,
      "learning_rate": 1.721785421785422e-05,
      "loss": 0.6911,
      "step": 7000
    },
    {
      "epoch": 6.12,
      "learning_rate": 1.7013104013104012e-05,
      "loss": 0.6695,
      "step": 7500
    },
    {
      "epoch": 6.53,
      "learning_rate": 1.680835380835381e-05,
      "loss": 0.6427,
      "step": 8000
    },
    {
      "epoch": 6.93,
      "learning_rate": 1.6603603603603603e-05,
      "loss": 0.6228,
      "step": 8500
    },
    {
      "epoch": 7.34,
      "learning_rate": 1.63992628992629e-05,
      "loss": 0.5752,
      "step": 9000
    },
    {
      "epoch": 7.75,
      "learning_rate": 1.6194512694512697e-05,
      "loss": 0.5751,
      "step": 9500
    },
    {
      "epoch": 8.16,
      "learning_rate": 1.598976248976249e-05,
      "loss": 0.5551,
      "step": 10000
    },
    {
      "epoch": 8.56,
      "learning_rate": 1.5785421785421787e-05,
      "loss": 0.5282,
      "step": 10500
    },
    {
      "epoch": 8.97,
      "learning_rate": 1.5580671580671584e-05,
      "loss": 0.5259,
      "step": 11000
    },
    {
      "epoch": 9.38,
      "learning_rate": 1.5375921375921378e-05,
      "loss": 0.4886,
      "step": 11500
    },
    {
      "epoch": 9.79,
      "learning_rate": 1.5171171171171173e-05,
      "loss": 0.489,
      "step": 12000
    },
    {
      "epoch": 10.2,
      "learning_rate": 1.4966830466830468e-05,
      "loss": 0.476,
      "step": 12500
    },
    {
      "epoch": 10.6,
      "learning_rate": 1.4762080262080263e-05,
      "loss": 0.4426,
      "step": 13000
    },
    {
      "epoch": 11.01,
      "learning_rate": 1.4557330057330059e-05,
      "loss": 0.4613,
      "step": 13500
    },
    {
      "epoch": 11.42,
      "learning_rate": 1.4352989352989356e-05,
      "loss": 0.4205,
      "step": 14000
    },
    {
      "epoch": 11.83,
      "learning_rate": 1.414823914823915e-05,
      "loss": 0.4196,
      "step": 14500
    },
    {
      "epoch": 12.23,
      "learning_rate": 1.3943488943488944e-05,
      "loss": 0.4189,
      "step": 15000
    },
    {
      "epoch": 12.64,
      "learning_rate": 1.373873873873874e-05,
      "loss": 0.3959,
      "step": 15500
    },
    {
      "epoch": 13.05,
      "learning_rate": 1.3533988533988535e-05,
      "loss": 0.3857,
      "step": 16000
    },
    {
      "epoch": 13.46,
      "learning_rate": 1.3329647829647832e-05,
      "loss": 0.3672,
      "step": 16500
    },
    {
      "epoch": 13.87,
      "learning_rate": 1.3124897624897626e-05,
      "loss": 0.3936,
      "step": 17000
    },
    {
      "epoch": 14.27,
      "learning_rate": 1.2920147420147423e-05,
      "loss": 0.3475,
      "step": 17500
    },
    {
      "epoch": 14.68,
      "learning_rate": 1.2715397215397216e-05,
      "loss": 0.3541,
      "step": 18000
    },
    {
      "epoch": 15.09,
      "learning_rate": 1.2510647010647013e-05,
      "loss": 0.3613,
      "step": 18500
    },
    {
      "epoch": 15.5,
      "learning_rate": 1.2305896805896807e-05,
      "loss": 0.3399,
      "step": 19000
    },
    {
      "epoch": 15.91,
      "learning_rate": 1.2101146601146602e-05,
      "loss": 0.3298,
      "step": 19500
    },
    {
      "epoch": 16.31,
      "learning_rate": 1.1896396396396397e-05,
      "loss": 0.3221,
      "step": 20000
    },
    {
      "epoch": 16.72,
      "learning_rate": 1.1691646191646192e-05,
      "loss": 0.3166,
      "step": 20500
    },
    {
      "epoch": 17.13,
      "learning_rate": 1.1486895986895988e-05,
      "loss": 0.3352,
      "step": 21000
    },
    {
      "epoch": 17.54,
      "learning_rate": 1.1282145782145783e-05,
      "loss": 0.3161,
      "step": 21500
    },
    {
      "epoch": 17.94,
      "learning_rate": 1.1077395577395578e-05,
      "loss": 0.3143,
      "step": 22000
    },
    {
      "epoch": 18.35,
      "learning_rate": 1.0872645372645374e-05,
      "loss": 0.3005,
      "step": 22500
    },
    {
      "epoch": 18.76,
      "learning_rate": 1.0667895167895169e-05,
      "loss": 0.2967,
      "step": 23000
    },
    {
      "epoch": 19.17,
      "learning_rate": 1.0463144963144964e-05,
      "loss": 0.2963,
      "step": 23500
    },
    {
      "epoch": 19.58,
      "learning_rate": 1.0258394758394758e-05,
      "loss": 0.2882,
      "step": 24000
    },
    {
      "epoch": 19.98,
      "learning_rate": 1.0054054054054055e-05,
      "loss": 0.282,
      "step": 24500
    },
    {
      "epoch": 20.39,
      "learning_rate": 9.84930384930385e-06,
      "loss": 0.2895,
      "step": 25000
    },
    {
      "epoch": 20.8,
      "learning_rate": 9.644963144963145e-06,
      "loss": 0.2699,
      "step": 25500
    },
    {
      "epoch": 21.21,
      "learning_rate": 9.44021294021294e-06,
      "loss": 0.2781,
      "step": 26000
    },
    {
      "epoch": 21.62,
      "learning_rate": 9.235462735462736e-06,
      "loss": 0.2748,
      "step": 26500
    },
    {
      "epoch": 22.02,
      "learning_rate": 9.031122031122031e-06,
      "loss": 0.2679,
      "step": 27000
    },
    {
      "epoch": 22.43,
      "learning_rate": 8.826371826371826e-06,
      "loss": 0.2581,
      "step": 27500
    },
    {
      "epoch": 22.84,
      "learning_rate": 8.621621621621622e-06,
      "loss": 0.2643,
      "step": 28000
    },
    {
      "epoch": 23.25,
      "learning_rate": 8.416871416871417e-06,
      "loss": 0.2547,
      "step": 28500
    },
    {
      "epoch": 23.65,
      "learning_rate": 8.212121212121212e-06,
      "loss": 0.253,
      "step": 29000
    },
    {
      "epoch": 24.06,
      "learning_rate": 8.007371007371007e-06,
      "loss": 0.2576,
      "step": 29500
    },
    {
      "epoch": 24.47,
      "learning_rate": 7.802620802620803e-06,
      "loss": 0.2373,
      "step": 30000
    },
    {
      "epoch": 24.88,
      "learning_rate": 7.597870597870598e-06,
      "loss": 0.2571,
      "step": 30500
    },
    {
      "epoch": 25.29,
      "learning_rate": 7.393120393120393e-06,
      "loss": 0.2461,
      "step": 31000
    },
    {
      "epoch": 25.69,
      "learning_rate": 7.1883701883701885e-06,
      "loss": 0.243,
      "step": 31500
    },
    {
      "epoch": 26.1,
      "learning_rate": 6.9840294840294846e-06,
      "loss": 0.2399,
      "step": 32000
    },
    {
      "epoch": 26.51,
      "learning_rate": 6.77927927927928e-06,
      "loss": 0.2395,
      "step": 32500
    },
    {
      "epoch": 26.92,
      "learning_rate": 6.574529074529074e-06,
      "loss": 0.2387,
      "step": 33000
    },
    {
      "epoch": 27.32,
      "learning_rate": 6.3697788697788695e-06,
      "loss": 0.2298,
      "step": 33500
    },
    {
      "epoch": 27.73,
      "learning_rate": 6.165028665028665e-06,
      "loss": 0.2285,
      "step": 34000
    },
    {
      "epoch": 28.14,
      "learning_rate": 5.960687960687961e-06,
      "loss": 0.2311,
      "step": 34500
    },
    {
      "epoch": 28.55,
      "learning_rate": 5.755937755937756e-06,
      "loss": 0.2307,
      "step": 35000
    },
    {
      "epoch": 28.96,
      "learning_rate": 5.5511875511875515e-06,
      "loss": 0.231,
      "step": 35500
    },
    {
      "epoch": 29.36,
      "learning_rate": 5.346437346437347e-06,
      "loss": 0.2229,
      "step": 36000
    },
    {
      "epoch": 29.77,
      "learning_rate": 5.141687141687142e-06,
      "loss": 0.2224,
      "step": 36500
    },
    {
      "epoch": 30.18,
      "learning_rate": 4.936936936936937e-06,
      "loss": 0.2209,
      "step": 37000
    },
    {
      "epoch": 30.59,
      "learning_rate": 4.7321867321867326e-06,
      "loss": 0.2277,
      "step": 37500
    },
    {
      "epoch": 31.0,
      "learning_rate": 4.527436527436528e-06,
      "loss": 0.222,
      "step": 38000
    },
    {
      "epoch": 31.4,
      "learning_rate": 4.323095823095824e-06,
      "loss": 0.2129,
      "step": 38500
    },
    {
      "epoch": 31.81,
      "learning_rate": 4.118345618345619e-06,
      "loss": 0.2287,
      "step": 39000
    },
    {
      "epoch": 32.22,
      "learning_rate": 3.9140049140049144e-06,
      "loss": 0.214,
      "step": 39500
    },
    {
      "epoch": 32.63,
      "learning_rate": 3.7092547092547093e-06,
      "loss": 0.211,
      "step": 40000
    },
    {
      "epoch": 33.03,
      "learning_rate": 3.5045045045045046e-06,
      "loss": 0.2204,
      "step": 40500
    },
    {
      "epoch": 33.44,
      "learning_rate": 3.2997542997543003e-06,
      "loss": 0.2155,
      "step": 41000
    },
    {
      "epoch": 33.85,
      "learning_rate": 3.0954135954135955e-06,
      "loss": 0.2163,
      "step": 41500
    },
    {
      "epoch": 34.26,
      "learning_rate": 2.890663390663391e-06,
      "loss": 0.2082,
      "step": 42000
    },
    {
      "epoch": 34.67,
      "learning_rate": 2.685913185913186e-06,
      "loss": 0.203,
      "step": 42500
    },
    {
      "epoch": 35.07,
      "learning_rate": 2.4811629811629813e-06,
      "loss": 0.2044,
      "step": 43000
    },
    {
      "epoch": 35.48,
      "learning_rate": 2.2764127764127766e-06,
      "loss": 0.2084,
      "step": 43500
    },
    {
      "epoch": 35.89,
      "learning_rate": 2.0720720720720723e-06,
      "loss": 0.2088,
      "step": 44000
    },
    {
      "epoch": 36.3,
      "learning_rate": 1.8673218673218673e-06,
      "loss": 0.2004,
      "step": 44500
    },
    {
      "epoch": 36.7,
      "learning_rate": 1.6625716625716626e-06,
      "loss": 0.2051,
      "step": 45000
    },
    {
      "epoch": 37.11,
      "learning_rate": 1.4578214578214581e-06,
      "loss": 0.2077,
      "step": 45500
    },
    {
      "epoch": 37.52,
      "learning_rate": 1.2530712530712532e-06,
      "loss": 0.2019,
      "step": 46000
    },
    {
      "epoch": 37.93,
      "learning_rate": 1.0483210483210485e-06,
      "loss": 0.2011,
      "step": 46500
    },
    {
      "epoch": 38.34,
      "learning_rate": 8.435708435708436e-07,
      "loss": 0.2092,
      "step": 47000
    },
    {
      "epoch": 38.74,
      "learning_rate": 6.392301392301394e-07,
      "loss": 0.1992,
      "step": 47500
    },
    {
      "epoch": 39.15,
      "learning_rate": 4.3447993447993446e-07,
      "loss": 0.2025,
      "step": 48000
    },
    {
      "epoch": 39.56,
      "learning_rate": 2.2972972972972977e-07,
      "loss": 0.2005,
      "step": 48500
    },
    {
      "epoch": 39.97,
      "learning_rate": 2.4979524979524984e-08,
      "loss": 0.2026,
      "step": 49000
    },
    {
      "epoch": 40.0,
      "step": 49040,
      "total_flos": 7.6145025653737e+19,
      "train_loss": 0.4991270453272597,
      "train_runtime": 30553.5264,
      "train_samples_per_second": 12.831,
      "train_steps_per_second": 1.605
    }
  ],
  "max_steps": 49040,
  "num_train_epochs": 40,
  "total_flos": 7.6145025653737e+19,
  "trial_name": null,
  "trial_params": null
}