{
  "best_metric": 0.288575142621994,
  "best_model_checkpoint": "./fine-tuned/checkpoint-2000",
  "epoch": 3.99667497921862,
  "eval_steps": 100,
  "global_step": 2404,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0831255195344971,
      "grad_norm": 81058.0859375,
      "learning_rate": 4.896006655574044e-05,
      "loss": 1.2803,
      "step": 50
    },
    {
      "epoch": 0.1662510390689942,
      "grad_norm": 80198.9375,
      "learning_rate": 4.792013311148087e-05,
      "loss": 0.6108,
      "step": 100
    },
    {
      "epoch": 0.1662510390689942,
      "eval_loss": 0.4418589174747467,
      "eval_runtime": 38.3971,
      "eval_samples_per_second": 12.866,
      "eval_steps_per_second": 1.615,
      "step": 100
    },
    {
      "epoch": 0.24937655860349128,
      "grad_norm": 70821.9921875,
      "learning_rate": 4.68801996672213e-05,
      "loss": 0.5174,
      "step": 150
    },
    {
      "epoch": 0.3325020781379884,
      "grad_norm": 75381.640625,
      "learning_rate": 4.5840266222961734e-05,
      "loss": 0.4743,
      "step": 200
    },
    {
      "epoch": 0.3325020781379884,
      "eval_loss": 0.3926239609718323,
      "eval_runtime": 38.3601,
      "eval_samples_per_second": 12.878,
      "eval_steps_per_second": 1.616,
      "step": 200
    },
    {
      "epoch": 0.41562759767248547,
      "grad_norm": 87610.4453125,
      "learning_rate": 4.480033277870216e-05,
      "loss": 0.4777,
      "step": 250
    },
    {
      "epoch": 0.49875311720698257,
      "grad_norm": 74392.109375,
      "learning_rate": 4.3760399334442597e-05,
      "loss": 0.4536,
      "step": 300
    },
    {
      "epoch": 0.49875311720698257,
      "eval_loss": 0.36712247133255005,
      "eval_runtime": 38.3686,
      "eval_samples_per_second": 12.875,
      "eval_steps_per_second": 1.616,
      "step": 300
    },
    {
      "epoch": 0.5818786367414797,
      "grad_norm": 152670.1875,
      "learning_rate": 4.272046589018303e-05,
      "loss": 0.4386,
      "step": 350
    },
    {
      "epoch": 0.6650041562759768,
      "grad_norm": 59777.48046875,
      "learning_rate": 4.1680532445923466e-05,
      "loss": 0.449,
      "step": 400
    },
    {
      "epoch": 0.6650041562759768,
      "eval_loss": 0.3487951159477234,
      "eval_runtime": 38.1805,
      "eval_samples_per_second": 12.939,
      "eval_steps_per_second": 1.624,
      "step": 400
    },
    {
      "epoch": 0.7481296758104738,
      "grad_norm": 60153.67578125,
      "learning_rate": 4.06405990016639e-05,
      "loss": 0.4116,
      "step": 450
    },
    {
      "epoch": 0.8312551953449709,
      "grad_norm": 60111.30078125,
      "learning_rate": 3.960066555740433e-05,
      "loss": 0.4145,
      "step": 500
    },
    {
      "epoch": 0.8312551953449709,
      "eval_loss": 0.3375319242477417,
      "eval_runtime": 38.2231,
      "eval_samples_per_second": 12.924,
      "eval_steps_per_second": 1.622,
      "step": 500
    },
    {
      "epoch": 0.914380714879468,
      "grad_norm": 60626.75,
      "learning_rate": 3.856073211314476e-05,
      "loss": 0.405,
      "step": 550
    },
    {
      "epoch": 0.9975062344139651,
      "grad_norm": 51910.04296875,
      "learning_rate": 3.752079866888519e-05,
      "loss": 0.4186,
      "step": 600
    },
    {
      "epoch": 0.9975062344139651,
      "eval_loss": 0.3285529315471649,
      "eval_runtime": 38.3187,
      "eval_samples_per_second": 12.892,
      "eval_steps_per_second": 1.618,
      "step": 600
    },
    {
      "epoch": 1.0806317539484622,
      "grad_norm": 48182.8046875,
      "learning_rate": 3.6480865224625625e-05,
      "loss": 0.3925,
      "step": 650
    },
    {
      "epoch": 1.1637572734829593,
      "grad_norm": 51930.34765625,
      "learning_rate": 3.544093178036606e-05,
      "loss": 0.3705,
      "step": 700
    },
    {
      "epoch": 1.1637572734829593,
      "eval_loss": 0.3227428197860718,
      "eval_runtime": 38.1185,
      "eval_samples_per_second": 12.96,
      "eval_steps_per_second": 1.627,
      "step": 700
    },
    {
      "epoch": 1.2468827930174564,
      "grad_norm": 60283.91015625,
      "learning_rate": 3.4400998336106495e-05,
      "loss": 0.3823,
      "step": 750
    },
    {
      "epoch": 1.3300083125519535,
      "grad_norm": 55843.62109375,
      "learning_rate": 3.336106489184692e-05,
      "loss": 0.3763,
      "step": 800
    },
    {
      "epoch": 1.3300083125519535,
      "eval_loss": 0.3174193501472473,
      "eval_runtime": 38.4517,
      "eval_samples_per_second": 12.847,
      "eval_steps_per_second": 1.612,
      "step": 800
    },
    {
      "epoch": 1.4131338320864506,
      "grad_norm": 40623.88671875,
      "learning_rate": 3.232113144758736e-05,
      "loss": 0.3509,
      "step": 850
    },
    {
      "epoch": 1.4962593516209477,
      "grad_norm": 57212.0546875,
      "learning_rate": 3.128119800332779e-05,
      "loss": 0.3624,
      "step": 900
    },
    {
      "epoch": 1.4962593516209477,
      "eval_loss": 0.31218209862709045,
      "eval_runtime": 38.1558,
      "eval_samples_per_second": 12.947,
      "eval_steps_per_second": 1.625,
      "step": 900
    },
    {
      "epoch": 1.5793848711554448,
      "grad_norm": 49043.25390625,
      "learning_rate": 3.0241264559068223e-05,
      "loss": 0.3656,
      "step": 950
    },
    {
      "epoch": 1.6625103906899419,
      "grad_norm": 50702.92578125,
      "learning_rate": 2.9201331114808654e-05,
      "loss": 0.3741,
      "step": 1000
    },
    {
      "epoch": 1.6625103906899419,
      "eval_loss": 0.3063213527202606,
      "eval_runtime": 38.2912,
      "eval_samples_per_second": 12.901,
      "eval_steps_per_second": 1.619,
      "step": 1000
    },
    {
      "epoch": 1.745635910224439,
      "grad_norm": 57559.86328125,
      "learning_rate": 2.816139767054909e-05,
      "loss": 0.3649,
      "step": 1050
    },
    {
      "epoch": 1.828761429758936,
      "grad_norm": 54892.80078125,
      "learning_rate": 2.7121464226289517e-05,
      "loss": 0.3518,
      "step": 1100
    },
    {
      "epoch": 1.828761429758936,
      "eval_loss": 0.30387237668037415,
      "eval_runtime": 38.3202,
      "eval_samples_per_second": 12.891,
      "eval_steps_per_second": 1.618,
      "step": 1100
    },
    {
      "epoch": 1.9118869492934332,
      "grad_norm": 56018.88671875,
      "learning_rate": 2.608153078202995e-05,
      "loss": 0.3515,
      "step": 1150
    },
    {
      "epoch": 1.9950124688279303,
      "grad_norm": 46875.38671875,
      "learning_rate": 2.5041597337770382e-05,
      "loss": 0.3459,
      "step": 1200
    },
    {
      "epoch": 1.9950124688279303,
      "eval_loss": 0.29989051818847656,
      "eval_runtime": 38.3688,
      "eval_samples_per_second": 12.875,
      "eval_steps_per_second": 1.616,
      "step": 1200
    },
    {
      "epoch": 2.0781379883624274,
      "grad_norm": 46399.90234375,
      "learning_rate": 2.4001663893510817e-05,
      "loss": 0.3293,
      "step": 1250
    },
    {
      "epoch": 2.1612635078969245,
      "grad_norm": 57348.53515625,
      "learning_rate": 2.296173044925125e-05,
      "loss": 0.3387,
      "step": 1300
    },
    {
      "epoch": 2.1612635078969245,
      "eval_loss": 0.29858091473579407,
      "eval_runtime": 38.3821,
      "eval_samples_per_second": 12.871,
      "eval_steps_per_second": 1.615,
      "step": 1300
    },
    {
      "epoch": 2.2443890274314215,
      "grad_norm": 32873.6796875,
      "learning_rate": 2.1921797004991683e-05,
      "loss": 0.3366,
      "step": 1350
    },
    {
      "epoch": 2.3275145469659186,
      "grad_norm": 47216.03125,
      "learning_rate": 2.0881863560732114e-05,
      "loss": 0.3528,
      "step": 1400
    },
    {
      "epoch": 2.3275145469659186,
      "eval_loss": 0.29636165499687195,
      "eval_runtime": 38.2663,
      "eval_samples_per_second": 12.91,
      "eval_steps_per_second": 1.62,
      "step": 1400
    },
    {
      "epoch": 2.4106400665004157,
      "grad_norm": 52153.40234375,
      "learning_rate": 1.9841930116472545e-05,
      "loss": 0.3379,
      "step": 1450
    },
    {
      "epoch": 2.493765586034913,
      "grad_norm": 57533.625,
      "learning_rate": 1.880199667221298e-05,
      "loss": 0.3175,
      "step": 1500
    },
    {
      "epoch": 2.493765586034913,
      "eval_loss": 0.29390034079551697,
      "eval_runtime": 38.2441,
      "eval_samples_per_second": 12.917,
      "eval_steps_per_second": 1.621,
      "step": 1500
    },
    {
      "epoch": 2.57689110556941,
      "grad_norm": 42211.703125,
      "learning_rate": 1.776206322795341e-05,
      "loss": 0.3199,
      "step": 1550
    },
    {
      "epoch": 2.660016625103907,
      "grad_norm": 41404.1953125,
      "learning_rate": 1.6722129783693842e-05,
      "loss": 0.3321,
      "step": 1600
    },
    {
      "epoch": 2.660016625103907,
      "eval_loss": 0.2914765477180481,
      "eval_runtime": 38.3949,
      "eval_samples_per_second": 12.866,
      "eval_steps_per_second": 1.615,
      "step": 1600
    },
    {
      "epoch": 2.743142144638404,
      "grad_norm": 65816.5546875,
      "learning_rate": 1.5682196339434277e-05,
      "loss": 0.3314,
      "step": 1650
    },
    {
      "epoch": 2.826267664172901,
      "grad_norm": 80084.921875,
      "learning_rate": 1.464226289517471e-05,
      "loss": 0.3163,
      "step": 1700
    },
    {
      "epoch": 2.826267664172901,
      "eval_loss": 0.29080235958099365,
      "eval_runtime": 38.4641,
      "eval_samples_per_second": 12.843,
      "eval_steps_per_second": 1.612,
      "step": 1700
    },
    {
      "epoch": 2.9093931837073983,
      "grad_norm": 41843.58203125,
      "learning_rate": 1.3602329450915141e-05,
      "loss": 0.3334,
      "step": 1750
    },
    {
      "epoch": 2.9925187032418954,
      "grad_norm": 46427.6796875,
      "learning_rate": 1.2562396006655574e-05,
      "loss": 0.3196,
      "step": 1800
    },
    {
      "epoch": 2.9925187032418954,
      "eval_loss": 0.2890726923942566,
      "eval_runtime": 38.2139,
      "eval_samples_per_second": 12.927,
      "eval_steps_per_second": 1.622,
      "step": 1800
    },
    {
      "epoch": 3.0756442227763925,
      "grad_norm": 45502.953125,
      "learning_rate": 1.1522462562396007e-05,
      "loss": 0.3226,
      "step": 1850
    },
    {
      "epoch": 3.1587697423108896,
      "grad_norm": 57836.49609375,
      "learning_rate": 1.048252911813644e-05,
      "loss": 0.3216,
      "step": 1900
    },
    {
      "epoch": 3.1587697423108896,
      "eval_loss": 0.2888147532939911,
      "eval_runtime": 38.2133,
      "eval_samples_per_second": 12.927,
      "eval_steps_per_second": 1.622,
      "step": 1900
    },
    {
      "epoch": 3.2418952618453867,
      "grad_norm": 51613.7734375,
      "learning_rate": 9.442595673876873e-06,
      "loss": 0.3169,
      "step": 1950
    },
    {
      "epoch": 3.3250207813798838,
      "grad_norm": 53367.69921875,
      "learning_rate": 8.402662229617304e-06,
      "loss": 0.3183,
      "step": 2000
    },
    {
      "epoch": 3.3250207813798838,
      "eval_loss": 0.288575142621994,
      "eval_runtime": 38.3741,
      "eval_samples_per_second": 12.873,
      "eval_steps_per_second": 1.616,
      "step": 2000
    },
    {
      "epoch": 3.408146300914381,
      "grad_norm": 48414.92578125,
      "learning_rate": 7.362728785357738e-06,
      "loss": 0.3053,
      "step": 2050
    },
    {
      "epoch": 3.491271820448878,
      "grad_norm": 56748.72265625,
      "learning_rate": 6.32279534109817e-06,
      "loss": 0.3253,
      "step": 2100
    },
    {
      "epoch": 3.491271820448878,
      "eval_loss": 0.28712254762649536,
      "eval_runtime": 38.3593,
      "eval_samples_per_second": 12.878,
      "eval_steps_per_second": 1.616,
      "step": 2100
    },
    {
      "epoch": 3.574397339983375,
      "grad_norm": 52447.25390625,
      "learning_rate": 5.282861896838603e-06,
      "loss": 0.3188,
      "step": 2150
    },
    {
      "epoch": 3.657522859517872,
      "grad_norm": 47730.953125,
      "learning_rate": 4.242928452579035e-06,
      "loss": 0.3041,
      "step": 2200
    },
    {
      "epoch": 3.657522859517872,
      "eval_loss": 0.2862567901611328,
      "eval_runtime": 38.2735,
      "eval_samples_per_second": 12.907,
      "eval_steps_per_second": 1.62,
      "step": 2200
    },
    {
      "epoch": 3.7406483790523692,
      "grad_norm": 48164.21484375,
      "learning_rate": 3.2029950083194676e-06,
      "loss": 0.3013,
      "step": 2250
    },
    {
      "epoch": 3.8237738985868663,
      "grad_norm": 44376.41015625,
      "learning_rate": 2.1630615640599005e-06,
      "loss": 0.3166,
      "step": 2300
    },
    {
      "epoch": 3.8237738985868663,
      "eval_loss": 0.2858003079891205,
      "eval_runtime": 38.3193,
      "eval_samples_per_second": 12.892,
      "eval_steps_per_second": 1.618,
      "step": 2300
    },
    {
      "epoch": 3.9068994181213634,
      "grad_norm": 53910.06640625,
      "learning_rate": 1.1231281198003328e-06,
      "loss": 0.3034,
      "step": 2350
    },
    {
      "epoch": 3.9900249376558605,
      "grad_norm": 36512.5546875,
      "learning_rate": 8.319467554076539e-08,
      "loss": 0.2983,
      "step": 2400
    },
    {
      "epoch": 3.9900249376558605,
      "eval_loss": 0.2859017550945282,
      "eval_runtime": 38.2723,
      "eval_samples_per_second": 12.908,
      "eval_steps_per_second": 1.62,
      "step": 2400
    }
  ],
  "logging_steps": 50,
  "max_steps": 2404,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.342112942882816e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}