{
  "best_metric": 1.9740982055664062,
  "best_model_checkpoint": "./outputs/checkpoint-3900",
  "epoch": 2.841530054644809,
  "eval_steps": 100,
  "global_step": 3900,
  "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": 2.7201,
      "step": 100
    },
    {
      "epoch": 0.07,
      "eval_loss": 2.618009567260742,
      "eval_runtime": 1145.9752,
      "eval_samples_per_second": 5.475,
      "eval_steps_per_second": 0.685,
      "step": 100
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 2.5807,
      "step": 200
    },
    {
      "epoch": 0.15,
      "eval_loss": 2.5655324459075928,
      "eval_runtime": 1107.0292,
      "eval_samples_per_second": 5.667,
      "eval_steps_per_second": 0.709,
      "step": 200
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 2.5387,
      "step": 300
    },
    {
      "epoch": 0.22,
      "eval_loss": 2.530600070953369,
      "eval_runtime": 1174.5089,
      "eval_samples_per_second": 5.342,
      "eval_steps_per_second": 0.668,
      "step": 300
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 2.5124,
      "step": 400
    },
    {
      "epoch": 0.29,
      "eval_loss": 2.496591806411743,
      "eval_runtime": 1111.3152,
      "eval_samples_per_second": 5.646,
      "eval_steps_per_second": 0.706,
      "step": 400
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 2.4699,
      "step": 500
    },
    {
      "epoch": 0.36,
      "eval_loss": 2.470156192779541,
      "eval_runtime": 1148.7316,
      "eval_samples_per_second": 5.462,
      "eval_steps_per_second": 0.683,
      "step": 500
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 2.4445,
      "step": 600
    },
    {
      "epoch": 0.44,
      "eval_loss": 2.4447057247161865,
      "eval_runtime": 1011.8783,
      "eval_samples_per_second": 6.2,
      "eval_steps_per_second": 0.776,
      "step": 600
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 2.4269,
      "step": 700
    },
    {
      "epoch": 0.51,
      "eval_loss": 2.4192888736724854,
      "eval_runtime": 1320.0264,
      "eval_samples_per_second": 4.753,
      "eval_steps_per_second": 0.595,
      "step": 700
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 2.4054,
      "step": 800
    },
    {
      "epoch": 0.58,
      "eval_loss": 2.3991127014160156,
      "eval_runtime": 1042.8122,
      "eval_samples_per_second": 6.016,
      "eval_steps_per_second": 0.753,
      "step": 800
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 2.3697,
      "step": 900
    },
    {
      "epoch": 0.66,
      "eval_loss": 2.3775434494018555,
      "eval_runtime": 1384.392,
      "eval_samples_per_second": 4.532,
      "eval_steps_per_second": 0.567,
      "step": 900
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 2.3663,
      "step": 1000
    },
    {
      "epoch": 0.73,
      "eval_loss": 2.360269784927368,
      "eval_runtime": 205.5954,
      "eval_samples_per_second": 30.516,
      "eval_steps_per_second": 3.818,
      "step": 1000
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 2.3591,
      "step": 1100
    },
    {
      "epoch": 0.8,
      "eval_loss": 2.3368232250213623,
      "eval_runtime": 205.7042,
      "eval_samples_per_second": 30.5,
      "eval_steps_per_second": 3.816,
      "step": 1100
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 2.3156,
      "step": 1200
    },
    {
      "epoch": 0.87,
      "eval_loss": 2.3195290565490723,
      "eval_runtime": 205.8525,
      "eval_samples_per_second": 30.478,
      "eval_steps_per_second": 3.813,
      "step": 1200
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 2.3264,
      "step": 1300
    },
    {
      "epoch": 0.95,
      "eval_loss": 2.3028838634490967,
      "eval_runtime": 205.7996,
      "eval_samples_per_second": 30.486,
      "eval_steps_per_second": 3.814,
      "step": 1300
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 2.2775,
      "step": 1400
    },
    {
      "epoch": 1.02,
      "eval_loss": 2.279730796813965,
      "eval_runtime": 206.0898,
      "eval_samples_per_second": 30.443,
      "eval_steps_per_second": 3.809,
      "step": 1400
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 2.2381,
      "step": 1500
    },
    {
      "epoch": 1.09,
      "eval_loss": 2.2633275985717773,
      "eval_runtime": 205.9739,
      "eval_samples_per_second": 30.46,
      "eval_steps_per_second": 3.811,
      "step": 1500
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 2.23,
      "step": 1600
    },
    {
      "epoch": 1.17,
      "eval_loss": 2.2453196048736572,
      "eval_runtime": 206.3463,
      "eval_samples_per_second": 30.405,
      "eval_steps_per_second": 3.804,
      "step": 1600
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 2.2304,
      "step": 1700
    },
    {
      "epoch": 1.24,
      "eval_loss": 2.2294440269470215,
      "eval_runtime": 205.8333,
      "eval_samples_per_second": 30.481,
      "eval_steps_per_second": 3.814,
      "step": 1700
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 2.1919,
      "step": 1800
    },
    {
      "epoch": 1.31,
      "eval_loss": 2.2156779766082764,
      "eval_runtime": 205.9059,
      "eval_samples_per_second": 30.47,
      "eval_steps_per_second": 3.812,
      "step": 1800
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 2.1769,
      "step": 1900
    },
    {
      "epoch": 1.38,
      "eval_loss": 2.1997225284576416,
      "eval_runtime": 206.368,
      "eval_samples_per_second": 30.402,
      "eval_steps_per_second": 3.804,
      "step": 1900
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 2.1705,
      "step": 2000
    },
    {
      "epoch": 1.46,
      "eval_loss": 2.1834230422973633,
      "eval_runtime": 206.1218,
      "eval_samples_per_second": 30.438,
      "eval_steps_per_second": 3.808,
      "step": 2000
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 2.1549,
      "step": 2100
    },
    {
      "epoch": 1.53,
      "eval_loss": 2.171769618988037,
      "eval_runtime": 206.0843,
      "eval_samples_per_second": 30.444,
      "eval_steps_per_second": 3.809,
      "step": 2100
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 2.1373,
      "step": 2200
    },
    {
      "epoch": 1.6,
      "eval_loss": 2.1567013263702393,
      "eval_runtime": 206.3812,
      "eval_samples_per_second": 30.4,
      "eval_steps_per_second": 3.804,
      "step": 2200
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 2.1272,
      "step": 2300
    },
    {
      "epoch": 1.68,
      "eval_loss": 2.144631862640381,
      "eval_runtime": 206.1489,
      "eval_samples_per_second": 30.434,
      "eval_steps_per_second": 3.808,
      "step": 2300
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 2.1339,
      "step": 2400
    },
    {
      "epoch": 1.75,
      "eval_loss": 2.1313953399658203,
      "eval_runtime": 206.1249,
      "eval_samples_per_second": 30.438,
      "eval_steps_per_second": 3.808,
      "step": 2400
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 2.1011,
      "step": 2500
    },
    {
      "epoch": 1.82,
      "eval_loss": 2.1183180809020996,
      "eval_runtime": 206.302,
      "eval_samples_per_second": 30.412,
      "eval_steps_per_second": 3.805,
      "step": 2500
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 2.1031,
      "step": 2600
    },
    {
      "epoch": 1.89,
      "eval_loss": 2.107461929321289,
      "eval_runtime": 206.4059,
      "eval_samples_per_second": 30.396,
      "eval_steps_per_second": 3.803,
      "step": 2600
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 2.0873,
      "step": 2700
    },
    {
      "epoch": 1.97,
      "eval_loss": 2.09187388420105,
      "eval_runtime": 206.3333,
      "eval_samples_per_second": 30.407,
      "eval_steps_per_second": 3.805,
      "step": 2700
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 2.0424,
      "step": 2800
    },
    {
      "epoch": 2.04,
      "eval_loss": 2.0847010612487793,
      "eval_runtime": 206.387,
      "eval_samples_per_second": 30.399,
      "eval_steps_per_second": 3.804,
      "step": 2800
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 2.0159,
      "step": 2900
    },
    {
      "epoch": 2.11,
      "eval_loss": 2.07328462600708,
      "eval_runtime": 205.9829,
      "eval_samples_per_second": 30.459,
      "eval_steps_per_second": 3.811,
      "step": 2900
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 2.0375,
      "step": 3000
    },
    {
      "epoch": 2.19,
      "eval_loss": 2.063817024230957,
      "eval_runtime": 206.2209,
      "eval_samples_per_second": 30.424,
      "eval_steps_per_second": 3.807,
      "step": 3000
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 2.0054,
      "step": 3100
    },
    {
      "epoch": 2.26,
      "eval_loss": 2.0530624389648438,
      "eval_runtime": 206.2013,
      "eval_samples_per_second": 30.427,
      "eval_steps_per_second": 3.807,
      "step": 3100
    },
    {
      "epoch": 2.33,
      "learning_rate": 0.0002,
      "loss": 2.0202,
      "step": 3200
    },
    {
      "epoch": 2.33,
      "eval_loss": 2.041797637939453,
      "eval_runtime": 206.1064,
      "eval_samples_per_second": 30.441,
      "eval_steps_per_second": 3.809,
      "step": 3200
    },
    {
      "epoch": 2.4,
      "learning_rate": 0.0002,
      "loss": 1.9871,
      "step": 3300
    },
    {
      "epoch": 2.4,
      "eval_loss": 2.0300114154815674,
      "eval_runtime": 205.777,
      "eval_samples_per_second": 30.489,
      "eval_steps_per_second": 3.815,
      "step": 3300
    },
    {
      "epoch": 2.48,
      "learning_rate": 0.0002,
      "loss": 1.9829,
      "step": 3400
    },
    {
      "epoch": 2.48,
      "eval_loss": 2.0251846313476562,
      "eval_runtime": 206.4461,
      "eval_samples_per_second": 30.39,
      "eval_steps_per_second": 3.802,
      "step": 3400
    },
    {
      "epoch": 2.55,
      "learning_rate": 0.0002,
      "loss": 1.9787,
      "step": 3500
    },
    {
      "epoch": 2.55,
      "eval_loss": 2.0121688842773438,
      "eval_runtime": 206.1187,
      "eval_samples_per_second": 30.439,
      "eval_steps_per_second": 3.808,
      "step": 3500
    },
    {
      "epoch": 2.62,
      "learning_rate": 0.0002,
      "loss": 1.9764,
      "step": 3600
    },
    {
      "epoch": 2.62,
      "eval_loss": 2.0046775341033936,
      "eval_runtime": 206.1405,
      "eval_samples_per_second": 30.436,
      "eval_steps_per_second": 3.808,
      "step": 3600
    },
    {
      "epoch": 2.7,
      "learning_rate": 0.0002,
      "loss": 1.9565,
      "step": 3700
    },
    {
      "epoch": 2.7,
      "eval_loss": 1.9959838390350342,
      "eval_runtime": 214.3631,
      "eval_samples_per_second": 29.268,
      "eval_steps_per_second": 3.662,
      "step": 3700
    },
    {
      "epoch": 2.77,
      "learning_rate": 0.0002,
      "loss": 1.9494,
      "step": 3800
    },
    {
      "epoch": 2.77,
      "eval_loss": 1.9870028495788574,
      "eval_runtime": 206.1419,
      "eval_samples_per_second": 30.435,
      "eval_steps_per_second": 3.808,
      "step": 3800
    },
    {
      "epoch": 2.84,
      "learning_rate": 0.0002,
      "loss": 1.9354,
      "step": 3900
    },
    {
      "epoch": 2.84,
      "eval_loss": 1.9740982055664062,
      "eval_runtime": 206.2491,
      "eval_samples_per_second": 30.42,
      "eval_steps_per_second": 3.806,
      "step": 3900
    }
  ],
  "logging_steps": 100,
  "max_steps": 4116,
  "num_train_epochs": 3,
  "save_steps": 100,
  "total_flos": 1.1452801265934336e+17,
  "trial_name": null,
  "trial_params": null
}