{
  "best_metric": 3.346250057220459,
  "best_model_checkpoint": "/tmp/berties-pretrained/checkpoint-6500",
  "epoch": 3.0,
  "global_step": 6993,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.21,
      "learning_rate": 4.642499642499642e-05,
      "loss": 5.7047,
      "step": 500
    },
    {
      "epoch": 0.21,
      "eval_loss": 5.037093639373779,
      "eval_runtime": 8.5329,
      "eval_samples_per_second": 114.967,
      "eval_steps_per_second": 14.415,
      "step": 500
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.2849992849992856e-05,
      "loss": 4.4116,
      "step": 1000
    },
    {
      "epoch": 0.43,
      "eval_loss": 4.514286041259766,
      "eval_runtime": 8.497,
      "eval_samples_per_second": 115.452,
      "eval_steps_per_second": 14.476,
      "step": 1000
    },
    {
      "epoch": 0.64,
      "learning_rate": 3.9274989274989276e-05,
      "loss": 4.0711,
      "step": 1500
    },
    {
      "epoch": 0.64,
      "eval_loss": 4.327243804931641,
      "eval_runtime": 8.5359,
      "eval_samples_per_second": 114.926,
      "eval_steps_per_second": 14.41,
      "step": 1500
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.56999856999857e-05,
      "loss": 3.8657,
      "step": 2000
    },
    {
      "epoch": 0.86,
      "eval_loss": 4.17035436630249,
      "eval_runtime": 8.541,
      "eval_samples_per_second": 114.858,
      "eval_steps_per_second": 14.401,
      "step": 2000
    },
    {
      "epoch": 1.07,
      "learning_rate": 3.212498212498213e-05,
      "loss": 3.6971,
      "step": 2500
    },
    {
      "epoch": 1.07,
      "eval_loss": 3.9950718879699707,
      "eval_runtime": 8.4686,
      "eval_samples_per_second": 115.84,
      "eval_steps_per_second": 14.524,
      "step": 2500
    },
    {
      "epoch": 1.29,
      "learning_rate": 2.854997854997855e-05,
      "loss": 3.5093,
      "step": 3000
    },
    {
      "epoch": 1.29,
      "eval_loss": 3.768172264099121,
      "eval_runtime": 8.4968,
      "eval_samples_per_second": 115.455,
      "eval_steps_per_second": 14.476,
      "step": 3000
    },
    {
      "epoch": 1.5,
      "learning_rate": 2.4974974974974976e-05,
      "loss": 3.3598,
      "step": 3500
    },
    {
      "epoch": 1.5,
      "eval_loss": 3.655388832092285,
      "eval_runtime": 8.4823,
      "eval_samples_per_second": 115.652,
      "eval_steps_per_second": 14.501,
      "step": 3500
    },
    {
      "epoch": 1.72,
      "learning_rate": 2.13999713999714e-05,
      "loss": 3.2389,
      "step": 4000
    },
    {
      "epoch": 1.72,
      "eval_loss": 3.6167075634002686,
      "eval_runtime": 8.4937,
      "eval_samples_per_second": 115.497,
      "eval_steps_per_second": 14.481,
      "step": 4000
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.7824967824967826e-05,
      "loss": 3.1984,
      "step": 4500
    },
    {
      "epoch": 1.93,
      "eval_loss": 3.4444828033447266,
      "eval_runtime": 8.5242,
      "eval_samples_per_second": 115.084,
      "eval_steps_per_second": 14.429,
      "step": 4500
    },
    {
      "epoch": 2.15,
      "learning_rate": 1.4249964249964251e-05,
      "loss": 3.1063,
      "step": 5000
    },
    {
      "epoch": 2.15,
      "eval_loss": 3.4205729961395264,
      "eval_runtime": 8.5145,
      "eval_samples_per_second": 115.215,
      "eval_steps_per_second": 14.446,
      "step": 5000
    },
    {
      "epoch": 2.36,
      "learning_rate": 1.0674960674960675e-05,
      "loss": 3.0063,
      "step": 5500
    },
    {
      "epoch": 2.36,
      "eval_loss": 3.416414976119995,
      "eval_runtime": 8.5224,
      "eval_samples_per_second": 115.108,
      "eval_steps_per_second": 14.432,
      "step": 5500
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.0999570999571e-06,
      "loss": 3.0069,
      "step": 6000
    },
    {
      "epoch": 2.57,
      "eval_loss": 3.4115302562713623,
      "eval_runtime": 8.4853,
      "eval_samples_per_second": 115.611,
      "eval_steps_per_second": 14.496,
      "step": 6000
    },
    {
      "epoch": 2.79,
      "learning_rate": 3.5249535249535254e-06,
      "loss": 2.9954,
      "step": 6500
    },
    {
      "epoch": 2.79,
      "eval_loss": 3.346250057220459,
      "eval_runtime": 8.4716,
      "eval_samples_per_second": 115.799,
      "eval_steps_per_second": 14.519,
      "step": 6500
    },
    {
      "epoch": 3.0,
      "step": 6993,
      "total_flos": 1415406837275046.0,
      "train_loss": 3.5777804869601746,
      "train_runtime": 2142.1553,
      "train_samples_per_second": 26.114,
      "train_steps_per_second": 3.264
    }
  ],
  "max_steps": 6993,
  "num_train_epochs": 3,
  "total_flos": 1415406837275046.0,
  "trial_name": null,
  "trial_params": null
}