{
  "best_metric": 0.39529716968536377,
  "best_model_checkpoint": "mikhail-panzo/ceb_b128_le4_s4000/checkpoint-1000",
  "epoch": 79.20792079207921,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 3.9603960396039604,
      "grad_norm": 2.952680826187134,
      "learning_rate": 2.5e-06,
      "loss": 0.7707,
      "step": 50
    },
    {
      "epoch": 7.920792079207921,
      "grad_norm": 1.126000165939331,
      "learning_rate": 5e-06,
      "loss": 0.6934,
      "step": 100
    },
    {
      "epoch": 11.881188118811881,
      "grad_norm": 2.129329204559326,
      "learning_rate": 7.5e-06,
      "loss": 0.5982,
      "step": 150
    },
    {
      "epoch": 15.841584158415841,
      "grad_norm": 1.1155015230178833,
      "learning_rate": 1e-05,
      "loss": 0.5093,
      "step": 200
    },
    {
      "epoch": 19.801980198019802,
      "grad_norm": 1.3643360137939453,
      "learning_rate": 1.25e-05,
      "loss": 0.4912,
      "step": 250
    },
    {
      "epoch": 23.762376237623762,
      "grad_norm": 0.9774149060249329,
      "learning_rate": 1.5e-05,
      "loss": 0.4764,
      "step": 300
    },
    {
      "epoch": 27.722772277227723,
      "grad_norm": 0.8793525695800781,
      "learning_rate": 1.75e-05,
      "loss": 0.4648,
      "step": 350
    },
    {
      "epoch": 31.683168316831683,
      "grad_norm": 0.7882798910140991,
      "learning_rate": 2e-05,
      "loss": 0.459,
      "step": 400
    },
    {
      "epoch": 35.64356435643565,
      "grad_norm": 1.2033296823501587,
      "learning_rate": 2.25e-05,
      "loss": 0.4519,
      "step": 450
    },
    {
      "epoch": 39.603960396039604,
      "grad_norm": 0.8620955348014832,
      "learning_rate": 2.5e-05,
      "loss": 0.4433,
      "step": 500
    },
    {
      "epoch": 39.603960396039604,
      "eval_loss": 0.4056467115879059,
      "eval_runtime": 7.4199,
      "eval_samples_per_second": 24.259,
      "eval_steps_per_second": 3.1,
      "step": 500
    },
    {
      "epoch": 43.56435643564357,
      "grad_norm": 1.266213297843933,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.444,
      "step": 550
    },
    {
      "epoch": 47.524752475247524,
      "grad_norm": 1.174134373664856,
      "learning_rate": 3e-05,
      "loss": 0.437,
      "step": 600
    },
    {
      "epoch": 51.48514851485149,
      "grad_norm": 0.8188551664352417,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.4318,
      "step": 650
    },
    {
      "epoch": 55.445544554455445,
      "grad_norm": 1.3083854913711548,
      "learning_rate": 3.5e-05,
      "loss": 0.4297,
      "step": 700
    },
    {
      "epoch": 59.40594059405941,
      "grad_norm": 0.8094576001167297,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.4238,
      "step": 750
    },
    {
      "epoch": 63.366336633663366,
      "grad_norm": 1.2506986856460571,
      "learning_rate": 4e-05,
      "loss": 0.4233,
      "step": 800
    },
    {
      "epoch": 67.32673267326733,
      "grad_norm": 1.7902193069458008,
      "learning_rate": 4.25e-05,
      "loss": 0.4208,
      "step": 850
    },
    {
      "epoch": 71.2871287128713,
      "grad_norm": 2.209643602371216,
      "learning_rate": 4.5e-05,
      "loss": 0.4195,
      "step": 900
    },
    {
      "epoch": 75.24752475247524,
      "grad_norm": 2.256385087966919,
      "learning_rate": 4.75e-05,
      "loss": 0.4202,
      "step": 950
    },
    {
      "epoch": 79.20792079207921,
      "grad_norm": 0.9167351126670837,
      "learning_rate": 5e-05,
      "loss": 0.4142,
      "step": 1000
    },
    {
      "epoch": 79.20792079207921,
      "eval_loss": 0.39529716968536377,
      "eval_runtime": 6.6198,
      "eval_samples_per_second": 27.191,
      "eval_steps_per_second": 3.474,
      "step": 1000
    }
  ],
  "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": 2.161155031107257e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}