{
  "best_metric": 0.40509888529777527,
  "best_model_checkpoint": "mikhail_panzo/ceb_b128_le3_s4000/checkpoint-500",
  "epoch": 118.81188118811882,
  "eval_steps": 500,
  "global_step": 1500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 3.9603960396039604,
      "grad_norm": 1.6000345945358276,
      "learning_rate": 2.5e-05,
      "loss": 0.7102,
      "step": 50
    },
    {
      "epoch": 7.920792079207921,
      "grad_norm": 1.240767478942871,
      "learning_rate": 5e-05,
      "loss": 0.5129,
      "step": 100
    },
    {
      "epoch": 11.881188118811881,
      "grad_norm": 2.545292377471924,
      "learning_rate": 7.5e-05,
      "loss": 0.4701,
      "step": 150
    },
    {
      "epoch": 15.841584158415841,
      "grad_norm": 3.054067611694336,
      "learning_rate": 0.0001,
      "loss": 0.4502,
      "step": 200
    },
    {
      "epoch": 19.801980198019802,
      "grad_norm": 3.884960889816284,
      "learning_rate": 0.000125,
      "loss": 0.4433,
      "step": 250
    },
    {
      "epoch": 23.762376237623762,
      "grad_norm": 2.102992296218872,
      "learning_rate": 0.00015,
      "loss": 0.4361,
      "step": 300
    },
    {
      "epoch": 27.722772277227723,
      "grad_norm": 3.1947975158691406,
      "learning_rate": 0.000175,
      "loss": 0.4335,
      "step": 350
    },
    {
      "epoch": 31.683168316831683,
      "grad_norm": 4.315099716186523,
      "learning_rate": 0.0002,
      "loss": 0.4322,
      "step": 400
    },
    {
      "epoch": 35.64356435643565,
      "grad_norm": 6.909496784210205,
      "learning_rate": 0.00022500000000000002,
      "loss": 0.4346,
      "step": 450
    },
    {
      "epoch": 39.603960396039604,
      "grad_norm": 2.0800621509552,
      "learning_rate": 0.00025,
      "loss": 0.42,
      "step": 500
    },
    {
      "epoch": 39.603960396039604,
      "eval_loss": 0.40509888529777527,
      "eval_runtime": 8.3422,
      "eval_samples_per_second": 21.577,
      "eval_steps_per_second": 2.757,
      "step": 500
    },
    {
      "epoch": 43.56435643564357,
      "grad_norm": 5.034981727600098,
      "learning_rate": 0.000275,
      "loss": 0.4336,
      "step": 550
    },
    {
      "epoch": 47.524752475247524,
      "grad_norm": 6.329326152801514,
      "learning_rate": 0.0003,
      "loss": 0.4265,
      "step": 600
    },
    {
      "epoch": 51.48514851485149,
      "grad_norm": 2.624539852142334,
      "learning_rate": 0.00032500000000000004,
      "loss": 0.4249,
      "step": 650
    },
    {
      "epoch": 55.445544554455445,
      "grad_norm": 1.945346474647522,
      "learning_rate": 0.00035,
      "loss": 0.4155,
      "step": 700
    },
    {
      "epoch": 59.40594059405941,
      "grad_norm": 3.682614326477051,
      "learning_rate": 0.000375,
      "loss": 0.4117,
      "step": 750
    },
    {
      "epoch": 63.366336633663366,
      "grad_norm": 1.726725459098816,
      "learning_rate": 0.0004,
      "loss": 0.4139,
      "step": 800
    },
    {
      "epoch": 67.32673267326733,
      "grad_norm": 1.665935754776001,
      "learning_rate": 0.000425,
      "loss": 0.4138,
      "step": 850
    },
    {
      "epoch": 71.2871287128713,
      "grad_norm": 2.6523916721343994,
      "learning_rate": 0.00045000000000000004,
      "loss": 0.4098,
      "step": 900
    },
    {
      "epoch": 75.24752475247524,
      "grad_norm": 4.059382915496826,
      "learning_rate": 0.000475,
      "loss": 0.4075,
      "step": 950
    },
    {
      "epoch": 79.20792079207921,
      "grad_norm": 4.141351222991943,
      "learning_rate": 0.0005,
      "loss": 0.4187,
      "step": 1000
    },
    {
      "epoch": 79.20792079207921,
      "eval_loss": 0.44087010622024536,
      "eval_runtime": 7.9097,
      "eval_samples_per_second": 22.757,
      "eval_steps_per_second": 2.908,
      "step": 1000
    },
    {
      "epoch": 83.16831683168317,
      "grad_norm": 4.268443584442139,
      "learning_rate": 0.0005250000000000001,
      "loss": 0.4233,
      "step": 1050
    },
    {
      "epoch": 87.12871287128714,
      "grad_norm": 2.5200088024139404,
      "learning_rate": 0.00055,
      "loss": 0.4165,
      "step": 1100
    },
    {
      "epoch": 91.08910891089108,
      "grad_norm": 1.1965175867080688,
      "learning_rate": 0.000575,
      "loss": 0.4131,
      "step": 1150
    },
    {
      "epoch": 95.04950495049505,
      "grad_norm": 5.397756099700928,
      "learning_rate": 0.0006,
      "loss": 0.4209,
      "step": 1200
    },
    {
      "epoch": 99.00990099009901,
      "grad_norm": 4.361856460571289,
      "learning_rate": 0.000625,
      "loss": 0.4289,
      "step": 1250
    },
    {
      "epoch": 102.97029702970298,
      "grad_norm": 8.962231636047363,
      "learning_rate": 0.0006500000000000001,
      "loss": 0.4418,
      "step": 1300
    },
    {
      "epoch": 106.93069306930693,
      "grad_norm": 5.27103328704834,
      "learning_rate": 0.000675,
      "loss": 0.4297,
      "step": 1350
    },
    {
      "epoch": 110.89108910891089,
      "grad_norm": 5.550845623016357,
      "learning_rate": 0.0007,
      "loss": 0.4394,
      "step": 1400
    },
    {
      "epoch": 114.85148514851485,
      "grad_norm": 1.9186675548553467,
      "learning_rate": 0.000725,
      "loss": 0.436,
      "step": 1450
    },
    {
      "epoch": 118.81188118811882,
      "grad_norm": 2.9912400245666504,
      "learning_rate": 0.00075,
      "loss": 0.4401,
      "step": 1500
    },
    {
      "epoch": 118.81188118811882,
      "eval_loss": 0.4779699444770813,
      "eval_runtime": 7.8042,
      "eval_samples_per_second": 23.065,
      "eval_steps_per_second": 2.947,
      "step": 1500
    }
  ],
  "logging_steps": 50,
  "max_steps": 4000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 334,
  "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": 3.2419037301352416e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}