{
  "best_metric": 1.9713504314422607,
  "best_model_checkpoint": "./outputs/checkpoint-4000",
  "epoch": 2.9143897996357016,
  "eval_steps": 100,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 2.7194,
      "step": 100
    },
    {
      "epoch": 0.07,
      "eval_loss": 2.617887258529663,
      "eval_runtime": 205.5207,
      "eval_samples_per_second": 30.527,
      "eval_steps_per_second": 3.82,
      "step": 100
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 2.5817,
      "step": 200
    },
    {
      "epoch": 0.15,
      "eval_loss": 2.5672028064727783,
      "eval_runtime": 205.4308,
      "eval_samples_per_second": 30.541,
      "eval_steps_per_second": 3.821,
      "step": 200
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 2.54,
      "step": 300
    },
    {
      "epoch": 0.22,
      "eval_loss": 2.5315635204315186,
      "eval_runtime": 205.1897,
      "eval_samples_per_second": 30.577,
      "eval_steps_per_second": 3.826,
      "step": 300
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 2.5136,
      "step": 400
    },
    {
      "epoch": 0.29,
      "eval_loss": 2.4977593421936035,
      "eval_runtime": 204.833,
      "eval_samples_per_second": 30.63,
      "eval_steps_per_second": 3.832,
      "step": 400
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 2.4701,
      "step": 500
    },
    {
      "epoch": 0.36,
      "eval_loss": 2.472073554992676,
      "eval_runtime": 205.144,
      "eval_samples_per_second": 30.583,
      "eval_steps_per_second": 3.827,
      "step": 500
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 2.4445,
      "step": 600
    },
    {
      "epoch": 0.44,
      "eval_loss": 2.4433655738830566,
      "eval_runtime": 204.9978,
      "eval_samples_per_second": 30.605,
      "eval_steps_per_second": 3.829,
      "step": 600
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 2.4276,
      "step": 700
    },
    {
      "epoch": 0.51,
      "eval_loss": 2.4212682247161865,
      "eval_runtime": 204.9092,
      "eval_samples_per_second": 30.618,
      "eval_steps_per_second": 3.831,
      "step": 700
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 2.4062,
      "step": 800
    },
    {
      "epoch": 0.58,
      "eval_loss": 2.4029204845428467,
      "eval_runtime": 205.0316,
      "eval_samples_per_second": 30.6,
      "eval_steps_per_second": 3.829,
      "step": 800
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 2.3706,
      "step": 900
    },
    {
      "epoch": 0.66,
      "eval_loss": 2.3802640438079834,
      "eval_runtime": 205.3365,
      "eval_samples_per_second": 30.555,
      "eval_steps_per_second": 3.823,
      "step": 900
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 2.3666,
      "step": 1000
    },
    {
      "epoch": 0.73,
      "eval_loss": 2.361499547958374,
      "eval_runtime": 205.1533,
      "eval_samples_per_second": 30.582,
      "eval_steps_per_second": 3.826,
      "step": 1000
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 2.3602,
      "step": 1100
    },
    {
      "epoch": 0.8,
      "eval_loss": 2.340102434158325,
      "eval_runtime": 205.0301,
      "eval_samples_per_second": 30.6,
      "eval_steps_per_second": 3.829,
      "step": 1100
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 2.3174,
      "step": 1200
    },
    {
      "epoch": 0.87,
      "eval_loss": 2.3213117122650146,
      "eval_runtime": 205.0977,
      "eval_samples_per_second": 30.59,
      "eval_steps_per_second": 3.827,
      "step": 1200
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 2.3283,
      "step": 1300
    },
    {
      "epoch": 0.95,
      "eval_loss": 2.3045783042907715,
      "eval_runtime": 205.3162,
      "eval_samples_per_second": 30.558,
      "eval_steps_per_second": 3.823,
      "step": 1300
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 2.2798,
      "step": 1400
    },
    {
      "epoch": 1.02,
      "eval_loss": 2.280726671218872,
      "eval_runtime": 205.3248,
      "eval_samples_per_second": 30.556,
      "eval_steps_per_second": 3.823,
      "step": 1400
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 2.2397,
      "step": 1500
    },
    {
      "epoch": 1.09,
      "eval_loss": 2.26517915725708,
      "eval_runtime": 205.5768,
      "eval_samples_per_second": 30.519,
      "eval_steps_per_second": 3.819,
      "step": 1500
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 2.2329,
      "step": 1600
    },
    {
      "epoch": 1.17,
      "eval_loss": 2.2466039657592773,
      "eval_runtime": 221.8099,
      "eval_samples_per_second": 28.285,
      "eval_steps_per_second": 3.539,
      "step": 1600
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 2.2321,
      "step": 1700
    },
    {
      "epoch": 1.24,
      "eval_loss": 2.2312474250793457,
      "eval_runtime": 221.2202,
      "eval_samples_per_second": 28.361,
      "eval_steps_per_second": 3.549,
      "step": 1700
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 2.1942,
      "step": 1800
    },
    {
      "epoch": 1.31,
      "eval_loss": 2.2154178619384766,
      "eval_runtime": 205.4645,
      "eval_samples_per_second": 30.536,
      "eval_steps_per_second": 3.821,
      "step": 1800
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 2.1794,
      "step": 1900
    },
    {
      "epoch": 1.38,
      "eval_loss": 2.2015323638916016,
      "eval_runtime": 205.5536,
      "eval_samples_per_second": 30.522,
      "eval_steps_per_second": 3.819,
      "step": 1900
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 2.1733,
      "step": 2000
    },
    {
      "epoch": 1.46,
      "eval_loss": 2.188199758529663,
      "eval_runtime": 205.7545,
      "eval_samples_per_second": 30.493,
      "eval_steps_per_second": 3.815,
      "step": 2000
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 2.158,
      "step": 2100
    },
    {
      "epoch": 1.53,
      "eval_loss": 2.175708770751953,
      "eval_runtime": 206.3248,
      "eval_samples_per_second": 30.408,
      "eval_steps_per_second": 3.805,
      "step": 2100
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 2.14,
      "step": 2200
    },
    {
      "epoch": 1.6,
      "eval_loss": 2.158797025680542,
      "eval_runtime": 205.208,
      "eval_samples_per_second": 30.574,
      "eval_steps_per_second": 3.825,
      "step": 2200
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 2.13,
      "step": 2300
    },
    {
      "epoch": 1.68,
      "eval_loss": 2.147514581680298,
      "eval_runtime": 205.8056,
      "eval_samples_per_second": 30.485,
      "eval_steps_per_second": 3.814,
      "step": 2300
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 2.1364,
      "step": 2400
    },
    {
      "epoch": 1.75,
      "eval_loss": 2.135024309158325,
      "eval_runtime": 272.6766,
      "eval_samples_per_second": 23.009,
      "eval_steps_per_second": 2.879,
      "step": 2400
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 2.1042,
      "step": 2500
    },
    {
      "epoch": 1.82,
      "eval_loss": 2.120384454727173,
      "eval_runtime": 205.6423,
      "eval_samples_per_second": 30.509,
      "eval_steps_per_second": 3.817,
      "step": 2500
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 2.1049,
      "step": 2600
    },
    {
      "epoch": 1.89,
      "eval_loss": 2.1099581718444824,
      "eval_runtime": 212.4604,
      "eval_samples_per_second": 29.53,
      "eval_steps_per_second": 3.695,
      "step": 2600
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 2.0887,
      "step": 2700
    },
    {
      "epoch": 1.97,
      "eval_loss": 2.09414005279541,
      "eval_runtime": 205.8822,
      "eval_samples_per_second": 30.474,
      "eval_steps_per_second": 3.813,
      "step": 2700
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 2.0444,
      "step": 2800
    },
    {
      "epoch": 2.04,
      "eval_loss": 2.0865635871887207,
      "eval_runtime": 205.997,
      "eval_samples_per_second": 30.457,
      "eval_steps_per_second": 3.811,
      "step": 2800
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 2.0172,
      "step": 2900
    },
    {
      "epoch": 2.11,
      "eval_loss": 2.0748651027679443,
      "eval_runtime": 206.0085,
      "eval_samples_per_second": 30.455,
      "eval_steps_per_second": 3.811,
      "step": 2900
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 2.0389,
      "step": 3000
    },
    {
      "epoch": 2.19,
      "eval_loss": 2.0672781467437744,
      "eval_runtime": 206.2433,
      "eval_samples_per_second": 30.42,
      "eval_steps_per_second": 3.806,
      "step": 3000
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 2.009,
      "step": 3100
    },
    {
      "epoch": 2.26,
      "eval_loss": 2.0537331104278564,
      "eval_runtime": 205.4748,
      "eval_samples_per_second": 30.534,
      "eval_steps_per_second": 3.82,
      "step": 3100
    },
    {
      "epoch": 2.33,
      "learning_rate": 0.0002,
      "loss": 2.0242,
      "step": 3200
    },
    {
      "epoch": 2.33,
      "eval_loss": 2.0447185039520264,
      "eval_runtime": 205.62,
      "eval_samples_per_second": 30.513,
      "eval_steps_per_second": 3.818,
      "step": 3200
    },
    {
      "epoch": 2.4,
      "learning_rate": 0.0002,
      "loss": 1.9883,
      "step": 3300
    },
    {
      "epoch": 2.4,
      "eval_loss": 2.0326316356658936,
      "eval_runtime": 205.8972,
      "eval_samples_per_second": 30.472,
      "eval_steps_per_second": 3.813,
      "step": 3300
    },
    {
      "epoch": 2.48,
      "learning_rate": 0.0002,
      "loss": 1.9854,
      "step": 3400
    },
    {
      "epoch": 2.48,
      "eval_loss": 2.0283098220825195,
      "eval_runtime": 205.6153,
      "eval_samples_per_second": 30.513,
      "eval_steps_per_second": 3.818,
      "step": 3400
    },
    {
      "epoch": 2.55,
      "learning_rate": 0.0002,
      "loss": 1.9806,
      "step": 3500
    },
    {
      "epoch": 2.55,
      "eval_loss": 2.0158398151397705,
      "eval_runtime": 206.095,
      "eval_samples_per_second": 30.442,
      "eval_steps_per_second": 3.809,
      "step": 3500
    },
    {
      "epoch": 2.62,
      "learning_rate": 0.0002,
      "loss": 1.9797,
      "step": 3600
    },
    {
      "epoch": 2.62,
      "eval_loss": 2.0079221725463867,
      "eval_runtime": 205.9217,
      "eval_samples_per_second": 30.468,
      "eval_steps_per_second": 3.812,
      "step": 3600
    },
    {
      "epoch": 2.7,
      "learning_rate": 0.0002,
      "loss": 1.9575,
      "step": 3700
    },
    {
      "epoch": 2.7,
      "eval_loss": 1.996368646621704,
      "eval_runtime": 206.2331,
      "eval_samples_per_second": 30.422,
      "eval_steps_per_second": 3.806,
      "step": 3700
    },
    {
      "epoch": 2.77,
      "learning_rate": 0.0002,
      "loss": 1.9523,
      "step": 3800
    },
    {
      "epoch": 2.77,
      "eval_loss": 1.9896615743637085,
      "eval_runtime": 206.6251,
      "eval_samples_per_second": 30.364,
      "eval_steps_per_second": 3.799,
      "step": 3800
    },
    {
      "epoch": 2.84,
      "learning_rate": 0.0002,
      "loss": 1.9386,
      "step": 3900
    },
    {
      "epoch": 2.84,
      "eval_loss": 1.9760409593582153,
      "eval_runtime": 206.2709,
      "eval_samples_per_second": 30.416,
      "eval_steps_per_second": 3.806,
      "step": 3900
    },
    {
      "epoch": 2.91,
      "learning_rate": 0.0002,
      "loss": 1.9401,
      "step": 4000
    },
    {
      "epoch": 2.91,
      "eval_loss": 1.9713504314422607,
      "eval_runtime": 206.4883,
      "eval_samples_per_second": 30.384,
      "eval_steps_per_second": 3.802,
      "step": 4000
    }
  ],
  "logging_steps": 100,
  "max_steps": 4116,
  "num_train_epochs": 3,
  "save_steps": 100,
  "total_flos": 1.1744177835220992e+17,
  "trial_name": null,
  "trial_params": null
}