{
  "best_metric": 0.4116077423095703,
  "best_model_checkpoint": "mikhail-panzo/zlm_b64_le4_s4000/checkpoint-1000",
  "epoch": 0.837696335078534,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.041884816753926704,
      "grad_norm": 3.0341947078704834,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 1.0529,
      "step": 50
    },
    {
      "epoch": 0.08376963350785341,
      "grad_norm": 3.274244785308838,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.8525,
      "step": 100
    },
    {
      "epoch": 0.1256544502617801,
      "grad_norm": 7.409469127655029,
      "learning_rate": 7.4e-06,
      "loss": 0.7528,
      "step": 150
    },
    {
      "epoch": 0.16753926701570682,
      "grad_norm": 3.476987600326538,
      "learning_rate": 9.900000000000002e-06,
      "loss": 0.689,
      "step": 200
    },
    {
      "epoch": 0.2094240837696335,
      "grad_norm": 2.8871872425079346,
      "learning_rate": 1.24e-05,
      "loss": 0.6403,
      "step": 250
    },
    {
      "epoch": 0.2513089005235602,
      "grad_norm": 2.007760763168335,
      "learning_rate": 1.49e-05,
      "loss": 0.6108,
      "step": 300
    },
    {
      "epoch": 0.2931937172774869,
      "grad_norm": 1.8842569589614868,
      "learning_rate": 1.74e-05,
      "loss": 0.5696,
      "step": 350
    },
    {
      "epoch": 0.33507853403141363,
      "grad_norm": 3.3749866485595703,
      "learning_rate": 1.9900000000000003e-05,
      "loss": 0.5544,
      "step": 400
    },
    {
      "epoch": 0.3769633507853403,
      "grad_norm": 3.5075504779815674,
      "learning_rate": 2.2400000000000002e-05,
      "loss": 0.5441,
      "step": 450
    },
    {
      "epoch": 0.418848167539267,
      "grad_norm": 3.789675712585449,
      "learning_rate": 2.4900000000000002e-05,
      "loss": 0.5277,
      "step": 500
    },
    {
      "epoch": 0.418848167539267,
      "eval_loss": 0.4806458353996277,
      "eval_runtime": 268.4116,
      "eval_samples_per_second": 31.627,
      "eval_steps_per_second": 3.957,
      "step": 500
    },
    {
      "epoch": 0.4607329842931937,
      "grad_norm": 7.912121772766113,
      "learning_rate": 2.7400000000000002e-05,
      "loss": 0.527,
      "step": 550
    },
    {
      "epoch": 0.5026178010471204,
      "grad_norm": 1.7653332948684692,
      "learning_rate": 2.9900000000000002e-05,
      "loss": 0.5152,
      "step": 600
    },
    {
      "epoch": 0.5445026178010471,
      "grad_norm": 2.491381883621216,
      "learning_rate": 3.24e-05,
      "loss": 0.5163,
      "step": 650
    },
    {
      "epoch": 0.5863874345549738,
      "grad_norm": 3.3132810592651367,
      "learning_rate": 3.49e-05,
      "loss": 0.4987,
      "step": 700
    },
    {
      "epoch": 0.6282722513089005,
      "grad_norm": 3.01517915725708,
      "learning_rate": 3.74e-05,
      "loss": 0.4909,
      "step": 750
    },
    {
      "epoch": 0.6701570680628273,
      "grad_norm": 3.681798219680786,
      "learning_rate": 3.99e-05,
      "loss": 0.4856,
      "step": 800
    },
    {
      "epoch": 0.7120418848167539,
      "grad_norm": 2.726991653442383,
      "learning_rate": 4.235e-05,
      "loss": 0.4901,
      "step": 850
    },
    {
      "epoch": 0.7539267015706806,
      "grad_norm": 1.8803215026855469,
      "learning_rate": 4.4850000000000006e-05,
      "loss": 0.4686,
      "step": 900
    },
    {
      "epoch": 0.7958115183246073,
      "grad_norm": 2.8432154655456543,
      "learning_rate": 4.735e-05,
      "loss": 0.4782,
      "step": 950
    },
    {
      "epoch": 0.837696335078534,
      "grad_norm": 2.698971748352051,
      "learning_rate": 4.9850000000000006e-05,
      "loss": 0.4582,
      "step": 1000
    },
    {
      "epoch": 0.837696335078534,
      "eval_loss": 0.4116077423095703,
      "eval_runtime": 270.0841,
      "eval_samples_per_second": 31.431,
      "eval_steps_per_second": 3.932,
      "step": 1000
    }
  ],
  "logging_steps": 50,
  "max_steps": 4000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "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": 8963407491426432.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}