{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 47275,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05,
      "learning_rate": 4.9471179270227394e-05,
      "loss": 3.4788,
      "step": 500
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.8942358540454786e-05,
      "loss": 2.8908,
      "step": 1000
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.841353781068218e-05,
      "loss": 2.6428,
      "step": 1500
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.7884717080909575e-05,
      "loss": 2.4772,
      "step": 2000
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.735589635113697e-05,
      "loss": 2.3782,
      "step": 2500
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.682707562136436e-05,
      "loss": 2.2953,
      "step": 3000
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.629825489159175e-05,
      "loss": 2.2185,
      "step": 3500
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.576943416181915e-05,
      "loss": 2.1603,
      "step": 4000
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.524167107350608e-05,
      "loss": 2.126,
      "step": 4500
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.471285034373348e-05,
      "loss": 2.083,
      "step": 5000
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.418402961396087e-05,
      "loss": 2.0452,
      "step": 5500
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.365520888418826e-05,
      "loss": 2.023,
      "step": 6000
    },
    {
      "epoch": 0.69,
      "learning_rate": 4.312638815441565e-05,
      "loss": 1.9837,
      "step": 6500
    },
    {
      "epoch": 0.74,
      "learning_rate": 4.259756742464305e-05,
      "loss": 1.9416,
      "step": 7000
    },
    {
      "epoch": 0.79,
      "learning_rate": 4.206874669487044e-05,
      "loss": 1.9265,
      "step": 7500
    },
    {
      "epoch": 0.85,
      "learning_rate": 4.1539925965097834e-05,
      "loss": 1.9048,
      "step": 8000
    },
    {
      "epoch": 0.9,
      "learning_rate": 4.101216287678477e-05,
      "loss": 1.8912,
      "step": 8500
    },
    {
      "epoch": 0.95,
      "learning_rate": 4.0483342147012164e-05,
      "loss": 1.8753,
      "step": 9000
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.995452141723956e-05,
      "loss": 1.8297,
      "step": 9500
    },
    {
      "epoch": 1.06,
      "learning_rate": 3.942570068746695e-05,
      "loss": 1.7432,
      "step": 10000
    },
    {
      "epoch": 1.11,
      "learning_rate": 3.889793759915389e-05,
      "loss": 1.7278,
      "step": 10500
    },
    {
      "epoch": 1.16,
      "learning_rate": 3.8369116869381283e-05,
      "loss": 1.716,
      "step": 11000
    },
    {
      "epoch": 1.22,
      "learning_rate": 3.7841353781068215e-05,
      "loss": 1.7175,
      "step": 11500
    },
    {
      "epoch": 1.27,
      "learning_rate": 3.731359069275516e-05,
      "loss": 1.7146,
      "step": 12000
    },
    {
      "epoch": 1.32,
      "learning_rate": 3.678476996298255e-05,
      "loss": 1.693,
      "step": 12500
    },
    {
      "epoch": 1.37,
      "learning_rate": 3.625594923320994e-05,
      "loss": 1.686,
      "step": 13000
    },
    {
      "epoch": 1.43,
      "learning_rate": 3.5727128503437334e-05,
      "loss": 1.6742,
      "step": 13500
    },
    {
      "epoch": 1.48,
      "learning_rate": 3.5198307773664726e-05,
      "loss": 1.6799,
      "step": 14000
    },
    {
      "epoch": 1.53,
      "learning_rate": 3.4669487043892124e-05,
      "loss": 1.6635,
      "step": 14500
    },
    {
      "epoch": 1.59,
      "learning_rate": 3.4140666314119516e-05,
      "loss": 1.6466,
      "step": 15000
    },
    {
      "epoch": 1.64,
      "learning_rate": 3.361184558434691e-05,
      "loss": 1.6462,
      "step": 15500
    },
    {
      "epoch": 1.69,
      "learning_rate": 3.30830248545743e-05,
      "loss": 1.6325,
      "step": 16000
    },
    {
      "epoch": 1.75,
      "learning_rate": 3.25542041248017e-05,
      "loss": 1.6235,
      "step": 16500
    },
    {
      "epoch": 1.8,
      "learning_rate": 3.202538339502908e-05,
      "loss": 1.6238,
      "step": 17000
    },
    {
      "epoch": 1.85,
      "learning_rate": 3.149656266525648e-05,
      "loss": 1.6193,
      "step": 17500
    },
    {
      "epoch": 1.9,
      "learning_rate": 3.096774193548387e-05,
      "loss": 1.6095,
      "step": 18000
    },
    {
      "epoch": 1.96,
      "learning_rate": 3.0438921205711264e-05,
      "loss": 1.6055,
      "step": 18500
    },
    {
      "epoch": 2.01,
      "learning_rate": 2.991010047593866e-05,
      "loss": 1.5826,
      "step": 19000
    },
    {
      "epoch": 2.06,
      "learning_rate": 2.938127974616605e-05,
      "loss": 1.4867,
      "step": 19500
    },
    {
      "epoch": 2.12,
      "learning_rate": 2.8852459016393445e-05,
      "loss": 1.4935,
      "step": 20000
    },
    {
      "epoch": 2.17,
      "learning_rate": 2.832469592808038e-05,
      "loss": 1.4922,
      "step": 20500
    },
    {
      "epoch": 2.22,
      "learning_rate": 2.7795875198307775e-05,
      "loss": 1.4821,
      "step": 21000
    },
    {
      "epoch": 2.27,
      "learning_rate": 2.7268112109994716e-05,
      "loss": 1.4862,
      "step": 21500
    },
    {
      "epoch": 2.33,
      "learning_rate": 2.6739291380222104e-05,
      "loss": 1.4818,
      "step": 22000
    },
    {
      "epoch": 2.38,
      "learning_rate": 2.6211528291909042e-05,
      "loss": 1.4798,
      "step": 22500
    },
    {
      "epoch": 2.43,
      "learning_rate": 2.5682707562136437e-05,
      "loss": 1.4719,
      "step": 23000
    },
    {
      "epoch": 2.49,
      "learning_rate": 2.5153886832363832e-05,
      "loss": 1.4622,
      "step": 23500
    },
    {
      "epoch": 2.54,
      "learning_rate": 2.4625066102591224e-05,
      "loss": 1.4654,
      "step": 24000
    },
    {
      "epoch": 2.59,
      "learning_rate": 2.4096245372818615e-05,
      "loss": 1.4618,
      "step": 24500
    },
    {
      "epoch": 2.64,
      "learning_rate": 2.3567424643046007e-05,
      "loss": 1.4418,
      "step": 25000
    },
    {
      "epoch": 2.7,
      "learning_rate": 2.3038603913273402e-05,
      "loss": 1.4613,
      "step": 25500
    },
    {
      "epoch": 2.75,
      "learning_rate": 2.2509783183500794e-05,
      "loss": 1.4543,
      "step": 26000
    },
    {
      "epoch": 2.8,
      "learning_rate": 2.198096245372819e-05,
      "loss": 1.4444,
      "step": 26500
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.145214172395558e-05,
      "loss": 1.4456,
      "step": 27000
    },
    {
      "epoch": 2.91,
      "learning_rate": 2.0923320994182975e-05,
      "loss": 1.4435,
      "step": 27500
    },
    {
      "epoch": 2.96,
      "learning_rate": 2.0394500264410367e-05,
      "loss": 1.4228,
      "step": 28000
    },
    {
      "epoch": 3.01,
      "learning_rate": 1.986567953463776e-05,
      "loss": 1.4057,
      "step": 28500
    },
    {
      "epoch": 3.07,
      "learning_rate": 1.9337916446324696e-05,
      "loss": 1.3607,
      "step": 29000
    },
    {
      "epoch": 3.12,
      "learning_rate": 1.8810153358011634e-05,
      "loss": 1.3515,
      "step": 29500
    },
    {
      "epoch": 3.17,
      "learning_rate": 1.828133262823903e-05,
      "loss": 1.3416,
      "step": 30000
    },
    {
      "epoch": 3.23,
      "learning_rate": 1.7753569539925964e-05,
      "loss": 1.3454,
      "step": 30500
    },
    {
      "epoch": 3.28,
      "learning_rate": 1.722474881015336e-05,
      "loss": 1.3449,
      "step": 31000
    },
    {
      "epoch": 3.33,
      "learning_rate": 1.669592808038075e-05,
      "loss": 1.3358,
      "step": 31500
    },
    {
      "epoch": 3.38,
      "learning_rate": 1.6167107350608142e-05,
      "loss": 1.3363,
      "step": 32000
    },
    {
      "epoch": 3.44,
      "learning_rate": 1.5638286620835537e-05,
      "loss": 1.3372,
      "step": 32500
    },
    {
      "epoch": 3.49,
      "learning_rate": 1.5109465891062929e-05,
      "loss": 1.3279,
      "step": 33000
    },
    {
      "epoch": 3.54,
      "learning_rate": 1.4580645161290324e-05,
      "loss": 1.3446,
      "step": 33500
    },
    {
      "epoch": 3.6,
      "learning_rate": 1.4051824431517715e-05,
      "loss": 1.3296,
      "step": 34000
    },
    {
      "epoch": 3.65,
      "learning_rate": 1.352300370174511e-05,
      "loss": 1.3208,
      "step": 34500
    },
    {
      "epoch": 3.7,
      "learning_rate": 1.2994182971972502e-05,
      "loss": 1.3309,
      "step": 35000
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.2465362242199895e-05,
      "loss": 1.324,
      "step": 35500
    },
    {
      "epoch": 3.81,
      "learning_rate": 1.1936541512427288e-05,
      "loss": 1.321,
      "step": 36000
    },
    {
      "epoch": 3.86,
      "learning_rate": 1.1407720782654681e-05,
      "loss": 1.3155,
      "step": 36500
    },
    {
      "epoch": 3.91,
      "learning_rate": 1.0878900052882075e-05,
      "loss": 1.3266,
      "step": 37000
    },
    {
      "epoch": 3.97,
      "learning_rate": 1.0352194606028556e-05,
      "loss": 1.3113,
      "step": 37500
    },
    {
      "epoch": 4.02,
      "learning_rate": 9.823373876255949e-06,
      "loss": 1.2879,
      "step": 38000
    },
    {
      "epoch": 4.07,
      "learning_rate": 9.294553146483342e-06,
      "loss": 1.2644,
      "step": 38500
    },
    {
      "epoch": 4.12,
      "learning_rate": 8.765732416710736e-06,
      "loss": 1.259,
      "step": 39000
    },
    {
      "epoch": 4.18,
      "learning_rate": 8.236911686938129e-06,
      "loss": 1.2709,
      "step": 39500
    },
    {
      "epoch": 4.23,
      "learning_rate": 7.708090957165522e-06,
      "loss": 1.2528,
      "step": 40000
    },
    {
      "epoch": 4.28,
      "learning_rate": 7.18032786885246e-06,
      "loss": 1.2498,
      "step": 40500
    },
    {
      "epoch": 4.34,
      "learning_rate": 6.651507139079852e-06,
      "loss": 1.2522,
      "step": 41000
    },
    {
      "epoch": 4.39,
      "learning_rate": 6.122686409307245e-06,
      "loss": 1.2504,
      "step": 41500
    },
    {
      "epoch": 4.44,
      "learning_rate": 5.5938656795346375e-06,
      "loss": 1.2602,
      "step": 42000
    },
    {
      "epoch": 4.49,
      "learning_rate": 5.065044949762031e-06,
      "loss": 1.2431,
      "step": 42500
    },
    {
      "epoch": 4.55,
      "learning_rate": 4.536224219989424e-06,
      "loss": 1.2391,
      "step": 43000
    },
    {
      "epoch": 4.6,
      "learning_rate": 4.007403490216817e-06,
      "loss": 1.2426,
      "step": 43500
    },
    {
      "epoch": 4.65,
      "learning_rate": 3.4785827604442092e-06,
      "loss": 1.2421,
      "step": 44000
    },
    {
      "epoch": 4.71,
      "learning_rate": 2.9497620306716025e-06,
      "loss": 1.2411,
      "step": 44500
    },
    {
      "epoch": 4.76,
      "learning_rate": 2.4219989423585406e-06,
      "loss": 1.2403,
      "step": 45000
    },
    {
      "epoch": 4.81,
      "learning_rate": 1.8931782125859334e-06,
      "loss": 1.2524,
      "step": 45500
    },
    {
      "epoch": 4.87,
      "learning_rate": 1.3643574828133265e-06,
      "loss": 1.2433,
      "step": 46000
    },
    {
      "epoch": 4.92,
      "learning_rate": 8.355367530407193e-07,
      "loss": 1.2377,
      "step": 46500
    },
    {
      "epoch": 4.97,
      "learning_rate": 3.0671602326811214e-07,
      "loss": 1.2334,
      "step": 47000
    },
    {
      "epoch": 5.0,
      "step": 47275,
      "total_flos": 5.929143675153285e+17,
      "train_loss": 1.586575432610474,
      "train_runtime": 13457.6286,
      "train_samples_per_second": 224.823,
      "train_steps_per_second": 3.513
    }
  ],
  "logging_steps": 500,
  "max_steps": 47275,
  "num_train_epochs": 5,
  "save_steps": 10000000,
  "total_flos": 5.929143675153285e+17,
  "trial_name": null,
  "trial_params": null
}