trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.3086838534599727,
  "eval_steps": 500,
  "global_step": 16,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09,
      "learning_rate": 1.9974521146102535e-05,
      "loss": 1.3794,
      "step": 1
    },
    {
      "epoch": 0.35,
      "eval_loss": 1.3587424755096436,
      "eval_runtime": 21.5907,
      "eval_samples_per_second": 13.617,
      "eval_steps_per_second": 1.714,
      "step": 4
    },
    {
      "epoch": 1.07,
      "learning_rate": 1.936949724999762e-05,
      "loss": 1.3642,
      "step": 5
    },
    {
      "epoch": 1.33,
      "eval_loss": 1.3150031566619873,
      "eval_runtime": 21.6398,
      "eval_samples_per_second": 13.586,
      "eval_steps_per_second": 1.71,
      "step": 8
    },
    {
      "epoch": 2.15,
      "learning_rate": 1.7557495743542586e-05,
      "loss": 1.3179,
      "step": 10
    },
    {
      "epoch": 2.32,
      "eval_loss": 1.2843962907791138,
      "eval_runtime": 21.6342,
      "eval_samples_per_second": 13.59,
      "eval_steps_per_second": 1.71,
      "step": 12
    },
    {
      "epoch": 3.22,
      "learning_rate": 1.479248986720057e-05,
      "loss": 1.2789,
      "step": 15
    },
    {
      "epoch": 3.31,
      "eval_loss": 1.2639315128326416,
      "eval_runtime": 21.5826,
      "eval_samples_per_second": 13.622,
      "eval_steps_per_second": 1.714,
      "step": 16
    },
    {
      "epoch": 3.31,
      "step": 16,
      "total_flos": 7.473328135392461e+17,
      "train_loss": 1.3643869757652283,
      "train_runtime": 5803.8585,
      "train_samples_per_second": 4.063,
      "train_steps_per_second": 0.008
    }
  ],
  "logging_steps": 5,
  "max_steps": 44,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 7.473328135392461e+17,
  "trial_name": null,
  "trial_params": null
}