{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.4496088357109986,
  "eval_steps": 500,
  "global_step": 6300,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05752416014726185,
      "grad_norm": 0.5827537775039673,
      "learning_rate": 0.00016513000460193283,
      "loss": 5.0544,
      "step": 250
    },
    {
      "epoch": 0.1150483202945237,
      "grad_norm": 0.5847251415252686,
      "learning_rate": 0.00016024045098941557,
      "loss": 4.2362,
      "step": 500
    },
    {
      "epoch": 0.1150483202945237,
      "eval_loss": 3.4322922229766846,
      "eval_runtime": 35.9031,
      "eval_samples_per_second": 109.74,
      "eval_steps_per_second": 3.231,
      "step": 500
    },
    {
      "epoch": 0.17257248044178555,
      "grad_norm": 0.6557937264442444,
      "learning_rate": 0.0001553508973768983,
      "loss": 4.0566,
      "step": 750
    },
    {
      "epoch": 0.2300966405890474,
      "grad_norm": 0.6734243631362915,
      "learning_rate": 0.00015046134376438104,
      "loss": 3.9349,
      "step": 1000
    },
    {
      "epoch": 0.2300966405890474,
      "eval_loss": 3.1976470947265625,
      "eval_runtime": 35.3562,
      "eval_samples_per_second": 111.437,
      "eval_steps_per_second": 3.281,
      "step": 1000
    },
    {
      "epoch": 0.28762080073630925,
      "grad_norm": 0.6826881766319275,
      "learning_rate": 0.0001455717901518638,
      "loss": 3.856,
      "step": 1250
    },
    {
      "epoch": 0.3451449608835711,
      "grad_norm": 0.6653180122375488,
      "learning_rate": 0.00014068223653934653,
      "loss": 3.7903,
      "step": 1500
    },
    {
      "epoch": 0.3451449608835711,
      "eval_loss": 3.0786449909210205,
      "eval_runtime": 36.0086,
      "eval_samples_per_second": 109.418,
      "eval_steps_per_second": 3.221,
      "step": 1500
    },
    {
      "epoch": 0.40266912103083297,
      "grad_norm": 0.7236284613609314,
      "learning_rate": 0.0001357926829268293,
      "loss": 3.7295,
      "step": 1750
    },
    {
      "epoch": 0.4601932811780948,
      "grad_norm": 0.7614400386810303,
      "learning_rate": 0.000130903129314312,
      "loss": 3.692,
      "step": 2000
    },
    {
      "epoch": 0.4601932811780948,
      "eval_loss": 3.0163052082061768,
      "eval_runtime": 35.3697,
      "eval_samples_per_second": 111.395,
      "eval_steps_per_second": 3.28,
      "step": 2000
    },
    {
      "epoch": 0.5177174413253567,
      "grad_norm": 0.7231118083000183,
      "learning_rate": 0.00012601357570179476,
      "loss": 3.6578,
      "step": 2250
    },
    {
      "epoch": 0.5752416014726185,
      "grad_norm": 0.7403096556663513,
      "learning_rate": 0.00012112402208927751,
      "loss": 3.6348,
      "step": 2500
    },
    {
      "epoch": 0.5752416014726185,
      "eval_loss": 2.9728314876556396,
      "eval_runtime": 35.3987,
      "eval_samples_per_second": 111.303,
      "eval_steps_per_second": 3.277,
      "step": 2500
    },
    {
      "epoch": 0.6327657616198803,
      "grad_norm": 0.8307600021362305,
      "learning_rate": 0.00011623446847676025,
      "loss": 3.5985,
      "step": 2750
    },
    {
      "epoch": 0.6902899217671422,
      "grad_norm": 0.692379891872406,
      "learning_rate": 0.000111344914864243,
      "loss": 3.5853,
      "step": 3000
    },
    {
      "epoch": 0.6902899217671422,
      "eval_loss": 2.949444055557251,
      "eval_runtime": 35.5037,
      "eval_samples_per_second": 110.974,
      "eval_steps_per_second": 3.267,
      "step": 3000
    },
    {
      "epoch": 0.747814081914404,
      "grad_norm": 0.6413500905036926,
      "learning_rate": 0.00010645536125172573,
      "loss": 3.553,
      "step": 3250
    },
    {
      "epoch": 0.8053382420616659,
      "grad_norm": 0.747322678565979,
      "learning_rate": 0.00010156580763920848,
      "loss": 3.5418,
      "step": 3500
    },
    {
      "epoch": 0.8053382420616659,
      "eval_loss": 2.9215950965881348,
      "eval_runtime": 35.7029,
      "eval_samples_per_second": 110.355,
      "eval_steps_per_second": 3.249,
      "step": 3500
    },
    {
      "epoch": 0.8628624022089277,
      "grad_norm": 0.7153074741363525,
      "learning_rate": 9.667625402669122e-05,
      "loss": 3.5174,
      "step": 3750
    },
    {
      "epoch": 0.9203865623561897,
      "grad_norm": 0.7129898071289062,
      "learning_rate": 9.178670041417396e-05,
      "loss": 3.5023,
      "step": 4000
    },
    {
      "epoch": 0.9203865623561897,
      "eval_loss": 2.8931143283843994,
      "eval_runtime": 35.6334,
      "eval_samples_per_second": 110.57,
      "eval_steps_per_second": 3.255,
      "step": 4000
    },
    {
      "epoch": 0.9779107225034515,
      "grad_norm": 0.6831061244010925,
      "learning_rate": 8.68971468016567e-05,
      "loss": 3.4839,
      "step": 4250
    },
    {
      "epoch": 1.0354348826507134,
      "grad_norm": 0.6215568780899048,
      "learning_rate": 8.200759318913945e-05,
      "loss": 3.4785,
      "step": 4500
    },
    {
      "epoch": 1.0354348826507134,
      "eval_loss": 2.888288974761963,
      "eval_runtime": 35.5637,
      "eval_samples_per_second": 110.787,
      "eval_steps_per_second": 3.262,
      "step": 4500
    },
    {
      "epoch": 1.0929590427979752,
      "grad_norm": 0.7072056531906128,
      "learning_rate": 7.711803957662219e-05,
      "loss": 3.4706,
      "step": 4750
    },
    {
      "epoch": 1.150483202945237,
      "grad_norm": 0.7484083771705627,
      "learning_rate": 7.222848596410494e-05,
      "loss": 3.4573,
      "step": 5000
    },
    {
      "epoch": 1.150483202945237,
      "eval_loss": 2.8630568981170654,
      "eval_runtime": 34.7986,
      "eval_samples_per_second": 113.223,
      "eval_steps_per_second": 3.333,
      "step": 5000
    },
    {
      "epoch": 1.2080073630924988,
      "grad_norm": 0.8469358682632446,
      "learning_rate": 6.733893235158767e-05,
      "loss": 3.4449,
      "step": 5250
    },
    {
      "epoch": 1.2655315232397606,
      "grad_norm": 0.7269773483276367,
      "learning_rate": 6.244937873907042e-05,
      "loss": 3.43,
      "step": 5500
    },
    {
      "epoch": 1.2655315232397606,
      "eval_loss": 2.853081226348877,
      "eval_runtime": 34.7036,
      "eval_samples_per_second": 113.533,
      "eval_steps_per_second": 3.343,
      "step": 5500
    },
    {
      "epoch": 1.3230556833870226,
      "grad_norm": 0.653972864151001,
      "learning_rate": 5.755982512655316e-05,
      "loss": 3.4265,
      "step": 5750
    },
    {
      "epoch": 1.3805798435342844,
      "grad_norm": 0.6091383695602417,
      "learning_rate": 5.267027151403589e-05,
      "loss": 3.4206,
      "step": 6000
    },
    {
      "epoch": 1.3805798435342844,
      "eval_loss": 2.84901762008667,
      "eval_runtime": 34.7272,
      "eval_samples_per_second": 113.456,
      "eval_steps_per_second": 3.34,
      "step": 6000
    },
    {
      "epoch": 1.4381040036815462,
      "grad_norm": 0.8110650181770325,
      "learning_rate": 4.778071790151864e-05,
      "loss": 3.4127,
      "step": 6250
    }
  ],
  "logging_steps": 250,
  "max_steps": 8692,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 350,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4833295532556288.0,
  "train_batch_size": 34,
  "trial_name": null,
  "trial_params": null
}