{
  "best_metric": 0.00016613505431450903,
  "best_model_checkpoint": "./results/checkpoint-500",
  "epoch": 4.62962962962963,
  "eval_steps": 20,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.18518518518518517,
      "grad_norm": 15.828068733215332,
      "learning_rate": 1.925925925925926e-05,
      "loss": 0.3468,
      "step": 20
    },
    {
      "epoch": 0.18518518518518517,
      "eval_accuracy": 0.9019607843137255,
      "eval_loss": 0.20517832040786743,
      "eval_runtime": 0.628,
      "eval_samples_per_second": 81.206,
      "eval_steps_per_second": 20.699,
      "step": 20
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 9.892717361450195,
      "learning_rate": 1.851851851851852e-05,
      "loss": 0.2089,
      "step": 40
    },
    {
      "epoch": 0.37037037037037035,
      "eval_accuracy": 0.9215686274509803,
      "eval_loss": 0.12760986387729645,
      "eval_runtime": 0.6405,
      "eval_samples_per_second": 79.628,
      "eval_steps_per_second": 20.297,
      "step": 40
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.21783480048179626,
      "learning_rate": 1.7777777777777777e-05,
      "loss": 0.214,
      "step": 60
    },
    {
      "epoch": 0.5555555555555556,
      "eval_accuracy": 0.9803921568627451,
      "eval_loss": 0.02514837123453617,
      "eval_runtime": 0.632,
      "eval_samples_per_second": 80.694,
      "eval_steps_per_second": 20.569,
      "step": 60
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 0.05077819526195526,
      "learning_rate": 1.7037037037037038e-05,
      "loss": 0.0807,
      "step": 80
    },
    {
      "epoch": 0.7407407407407407,
      "eval_accuracy": 0.9803921568627451,
      "eval_loss": 0.016518110409379005,
      "eval_runtime": 0.629,
      "eval_samples_per_second": 81.079,
      "eval_steps_per_second": 20.667,
      "step": 80
    },
    {
      "epoch": 0.9259259259259259,
      "grad_norm": 0.04394271969795227,
      "learning_rate": 1.6296296296296297e-05,
      "loss": 0.1064,
      "step": 100
    },
    {
      "epoch": 0.9259259259259259,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0030651080887764692,
      "eval_runtime": 0.6365,
      "eval_samples_per_second": 80.124,
      "eval_steps_per_second": 20.424,
      "step": 100
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.08775356411933899,
      "learning_rate": 1.555555555555556e-05,
      "loss": 0.0134,
      "step": 120
    },
    {
      "epoch": 1.1111111111111112,
      "eval_accuracy": 1.0,
      "eval_loss": 0.002504199743270874,
      "eval_runtime": 0.6271,
      "eval_samples_per_second": 81.321,
      "eval_steps_per_second": 20.729,
      "step": 120
    },
    {
      "epoch": 1.2962962962962963,
      "grad_norm": 0.028400583192706108,
      "learning_rate": 1.4814814814814815e-05,
      "loss": 0.0349,
      "step": 140
    },
    {
      "epoch": 1.2962962962962963,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0006583676440641284,
      "eval_runtime": 0.6233,
      "eval_samples_per_second": 81.817,
      "eval_steps_per_second": 20.855,
      "step": 140
    },
    {
      "epoch": 1.4814814814814814,
      "grad_norm": 0.00876949355006218,
      "learning_rate": 1.4074074074074075e-05,
      "loss": 0.0819,
      "step": 160
    },
    {
      "epoch": 1.4814814814814814,
      "eval_accuracy": 0.9607843137254902,
      "eval_loss": 0.07063630223274231,
      "eval_runtime": 0.6218,
      "eval_samples_per_second": 82.021,
      "eval_steps_per_second": 20.907,
      "step": 160
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.014950844459235668,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.0586,
      "step": 180
    },
    {
      "epoch": 1.6666666666666665,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0005163149326108396,
      "eval_runtime": 0.6311,
      "eval_samples_per_second": 80.809,
      "eval_steps_per_second": 20.598,
      "step": 180
    },
    {
      "epoch": 1.8518518518518519,
      "grad_norm": 0.01813225820660591,
      "learning_rate": 1.2592592592592593e-05,
      "loss": 0.0368,
      "step": 200
    },
    {
      "epoch": 1.8518518518518519,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0004809926904272288,
      "eval_runtime": 0.6213,
      "eval_samples_per_second": 82.09,
      "eval_steps_per_second": 20.925,
      "step": 200
    },
    {
      "epoch": 2.037037037037037,
      "grad_norm": 0.018220530822873116,
      "learning_rate": 1.1851851851851852e-05,
      "loss": 0.0485,
      "step": 220
    },
    {
      "epoch": 2.037037037037037,
      "eval_accuracy": 1.0,
      "eval_loss": 0.003055064007639885,
      "eval_runtime": 0.6259,
      "eval_samples_per_second": 81.483,
      "eval_steps_per_second": 20.77,
      "step": 220
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.012761042453348637,
      "learning_rate": 1.1111111111111113e-05,
      "loss": 0.017,
      "step": 240
    },
    {
      "epoch": 2.2222222222222223,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0003475048579275608,
      "eval_runtime": 0.6263,
      "eval_samples_per_second": 81.436,
      "eval_steps_per_second": 20.758,
      "step": 240
    },
    {
      "epoch": 2.4074074074074074,
      "grad_norm": 0.012553339824080467,
      "learning_rate": 1.037037037037037e-05,
      "loss": 0.0017,
      "step": 260
    },
    {
      "epoch": 2.4074074074074074,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0020730593241751194,
      "eval_runtime": 0.623,
      "eval_samples_per_second": 81.864,
      "eval_steps_per_second": 20.867,
      "step": 260
    },
    {
      "epoch": 2.5925925925925926,
      "grad_norm": 0.009718580171465874,
      "learning_rate": 9.62962962962963e-06,
      "loss": 0.0004,
      "step": 280
    },
    {
      "epoch": 2.5925925925925926,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0004216369998175651,
      "eval_runtime": 0.6294,
      "eval_samples_per_second": 81.028,
      "eval_steps_per_second": 20.654,
      "step": 280
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 0.009635565802454948,
      "learning_rate": 8.888888888888888e-06,
      "loss": 0.0444,
      "step": 300
    },
    {
      "epoch": 2.7777777777777777,
      "eval_accuracy": 1.0,
      "eval_loss": 0.00025334919337183237,
      "eval_runtime": 0.6299,
      "eval_samples_per_second": 80.96,
      "eval_steps_per_second": 20.637,
      "step": 300
    },
    {
      "epoch": 2.962962962962963,
      "grad_norm": 0.005175964906811714,
      "learning_rate": 8.148148148148148e-06,
      "loss": 0.0005,
      "step": 320
    },
    {
      "epoch": 2.962962962962963,
      "eval_accuracy": 1.0,
      "eval_loss": 0.00028179946821182966,
      "eval_runtime": 0.6242,
      "eval_samples_per_second": 81.709,
      "eval_steps_per_second": 20.828,
      "step": 320
    },
    {
      "epoch": 3.148148148148148,
      "grad_norm": 0.008158649317920208,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.0003,
      "step": 340
    },
    {
      "epoch": 3.148148148148148,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0002456614456605166,
      "eval_runtime": 0.6258,
      "eval_samples_per_second": 81.502,
      "eval_steps_per_second": 20.775,
      "step": 340
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.004232426173985004,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.0238,
      "step": 360
    },
    {
      "epoch": 3.3333333333333335,
      "eval_accuracy": 1.0,
      "eval_loss": 0.00026214588433504105,
      "eval_runtime": 0.6234,
      "eval_samples_per_second": 81.814,
      "eval_steps_per_second": 20.855,
      "step": 360
    },
    {
      "epoch": 3.5185185185185186,
      "grad_norm": 0.004335971549153328,
      "learning_rate": 5.925925925925926e-06,
      "loss": 0.0003,
      "step": 380
    },
    {
      "epoch": 3.5185185185185186,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0002457990194670856,
      "eval_runtime": 0.6226,
      "eval_samples_per_second": 81.91,
      "eval_steps_per_second": 20.879,
      "step": 380
    },
    {
      "epoch": 3.7037037037037037,
      "grad_norm": 0.007228133734315634,
      "learning_rate": 5.185185185185185e-06,
      "loss": 0.0003,
      "step": 400
    },
    {
      "epoch": 3.7037037037037037,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0002311759744770825,
      "eval_runtime": 0.6205,
      "eval_samples_per_second": 82.185,
      "eval_steps_per_second": 20.949,
      "step": 400
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 0.004542021080851555,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.0003,
      "step": 420
    },
    {
      "epoch": 3.888888888888889,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0002200859016738832,
      "eval_runtime": 0.6236,
      "eval_samples_per_second": 81.782,
      "eval_steps_per_second": 20.846,
      "step": 420
    },
    {
      "epoch": 4.074074074074074,
      "grad_norm": 0.004315598402172327,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.0007,
      "step": 440
    },
    {
      "epoch": 4.074074074074074,
      "eval_accuracy": 1.0,
      "eval_loss": 0.0001800174795789644,
      "eval_runtime": 0.6201,
      "eval_samples_per_second": 82.247,
      "eval_steps_per_second": 20.965,
      "step": 440
    },
    {
      "epoch": 4.2592592592592595,
      "grad_norm": 0.014702328480780125,
      "learning_rate": 2.962962962962963e-06,
      "loss": 0.0002,
      "step": 460
    },
    {
      "epoch": 4.2592592592592595,
      "eval_accuracy": 1.0,
      "eval_loss": 0.00017176676192320883,
      "eval_runtime": 0.6211,
      "eval_samples_per_second": 82.115,
      "eval_steps_per_second": 20.931,
      "step": 460
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 0.005943182855844498,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.0003,
      "step": 480
    },
    {
      "epoch": 4.444444444444445,
      "eval_accuracy": 1.0,
      "eval_loss": 0.00016821030294522643,
      "eval_runtime": 0.6229,
      "eval_samples_per_second": 81.877,
      "eval_steps_per_second": 20.871,
      "step": 480
    },
    {
      "epoch": 4.62962962962963,
      "grad_norm": 0.036252692341804504,
      "learning_rate": 1.4814814814814815e-06,
      "loss": 0.0002,
      "step": 500
    },
    {
      "epoch": 4.62962962962963,
      "eval_accuracy": 1.0,
      "eval_loss": 0.00016613505431450903,
      "eval_runtime": 0.627,
      "eval_samples_per_second": 81.342,
      "eval_steps_per_second": 20.734,
      "step": 500
    }
  ],
  "logging_steps": 20,
  "max_steps": 540,
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 411116162887800.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}