{
  "best_metric": 7.983619689941406,
  "best_model_checkpoint": "./models/magic/checkpoint-2440",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 2440,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.20491803278688525,
      "grad_norm": 66683.3203125,
      "learning_rate": 1.918032786885246e-05,
      "loss": 8.3446,
      "step": 100
    },
    {
      "epoch": 0.4098360655737705,
      "grad_norm": 88343.609375,
      "learning_rate": 1.836065573770492e-05,
      "loss": 8.344,
      "step": 200
    },
    {
      "epoch": 0.6147540983606558,
      "grad_norm": 112013.890625,
      "learning_rate": 1.7540983606557377e-05,
      "loss": 8.3305,
      "step": 300
    },
    {
      "epoch": 0.819672131147541,
      "grad_norm": 107013.5390625,
      "learning_rate": 1.6721311475409837e-05,
      "loss": 8.3148,
      "step": 400
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.0054761904761904765,
      "eval_loss": 8.26321792602539,
      "eval_runtime": 37.0437,
      "eval_samples_per_second": 113.38,
      "eval_steps_per_second": 3.563,
      "step": 488
    },
    {
      "epoch": 1.0245901639344261,
      "grad_norm": 109507.2734375,
      "learning_rate": 1.5901639344262295e-05,
      "loss": 8.2911,
      "step": 500
    },
    {
      "epoch": 1.2295081967213115,
      "grad_norm": 107816.046875,
      "learning_rate": 1.5081967213114754e-05,
      "loss": 8.2472,
      "step": 600
    },
    {
      "epoch": 1.4344262295081966,
      "grad_norm": 108237.4453125,
      "learning_rate": 1.4262295081967214e-05,
      "loss": 8.2257,
      "step": 700
    },
    {
      "epoch": 1.639344262295082,
      "grad_norm": 111140.1640625,
      "learning_rate": 1.3442622950819673e-05,
      "loss": 8.2074,
      "step": 800
    },
    {
      "epoch": 1.8442622950819674,
      "grad_norm": 107701.734375,
      "learning_rate": 1.2622950819672132e-05,
      "loss": 8.1958,
      "step": 900
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.05976190476190476,
      "eval_loss": 8.151880264282227,
      "eval_runtime": 37.1118,
      "eval_samples_per_second": 113.172,
      "eval_steps_per_second": 3.557,
      "step": 976
    },
    {
      "epoch": 2.0491803278688523,
      "grad_norm": 111871.875,
      "learning_rate": 1.1803278688524591e-05,
      "loss": 8.1707,
      "step": 1000
    },
    {
      "epoch": 2.2540983606557377,
      "grad_norm": 117045.8203125,
      "learning_rate": 1.0983606557377052e-05,
      "loss": 8.1275,
      "step": 1100
    },
    {
      "epoch": 2.459016393442623,
      "grad_norm": 110316.90625,
      "learning_rate": 1.0163934426229509e-05,
      "loss": 8.1193,
      "step": 1200
    },
    {
      "epoch": 2.663934426229508,
      "grad_norm": 114694.4609375,
      "learning_rate": 9.344262295081968e-06,
      "loss": 8.1023,
      "step": 1300
    },
    {
      "epoch": 2.8688524590163933,
      "grad_norm": 113438.8828125,
      "learning_rate": 8.524590163934427e-06,
      "loss": 8.089,
      "step": 1400
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.15666666666666668,
      "eval_loss": 8.05963134765625,
      "eval_runtime": 36.9672,
      "eval_samples_per_second": 113.614,
      "eval_steps_per_second": 3.571,
      "step": 1464
    },
    {
      "epoch": 3.0737704918032787,
      "grad_norm": 130567.7890625,
      "learning_rate": 7.704918032786886e-06,
      "loss": 8.0675,
      "step": 1500
    },
    {
      "epoch": 3.278688524590164,
      "grad_norm": 111654.6484375,
      "learning_rate": 6.885245901639345e-06,
      "loss": 8.049,
      "step": 1600
    },
    {
      "epoch": 3.4836065573770494,
      "grad_norm": 110857.171875,
      "learning_rate": 6.065573770491804e-06,
      "loss": 8.0304,
      "step": 1700
    },
    {
      "epoch": 3.6885245901639343,
      "grad_norm": 120024.4765625,
      "learning_rate": 5.245901639344263e-06,
      "loss": 8.0182,
      "step": 1800
    },
    {
      "epoch": 3.8934426229508197,
      "grad_norm": 113248.8359375,
      "learning_rate": 4.426229508196722e-06,
      "loss": 8.0208,
      "step": 1900
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.2276190476190476,
      "eval_loss": 8.003293991088867,
      "eval_runtime": 36.914,
      "eval_samples_per_second": 113.778,
      "eval_steps_per_second": 3.576,
      "step": 1952
    },
    {
      "epoch": 4.098360655737705,
      "grad_norm": 114861.390625,
      "learning_rate": 3.6065573770491806e-06,
      "loss": 8.0023,
      "step": 2000
    },
    {
      "epoch": 4.30327868852459,
      "grad_norm": 114188.84375,
      "learning_rate": 2.786885245901639e-06,
      "loss": 7.9841,
      "step": 2100
    },
    {
      "epoch": 4.508196721311475,
      "grad_norm": 114966.8828125,
      "learning_rate": 1.9672131147540985e-06,
      "loss": 7.9843,
      "step": 2200
    },
    {
      "epoch": 4.713114754098361,
      "grad_norm": 124759.4375,
      "learning_rate": 1.1475409836065575e-06,
      "loss": 7.9772,
      "step": 2300
    },
    {
      "epoch": 4.918032786885246,
      "grad_norm": 110823.4609375,
      "learning_rate": 3.278688524590164e-07,
      "loss": 7.983,
      "step": 2400
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.25452380952380954,
      "eval_loss": 7.983619689941406,
      "eval_runtime": 37.5516,
      "eval_samples_per_second": 111.846,
      "eval_steps_per_second": 3.515,
      "step": 2440
    },
    {
      "epoch": 5.0,
      "step": 2440,
      "total_flos": 6.271796898791424e+18,
      "train_loss": 8.131903288794346,
      "train_runtime": 1755.8039,
      "train_samples_per_second": 44.424,
      "train_steps_per_second": 1.39
    }
  ],
  "logging_steps": 100,
  "max_steps": 2440,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 6.271796898791424e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}