trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 2000.0,
  "global_step": 8340,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.3,
      "learning_rate": 2.5e-05,
      "loss": 1.5483,
      "step": 500
    },
    {
      "epoch": 0.6,
      "learning_rate": 5e-05,
      "loss": 1.1393,
      "step": 1000
    },
    {
      "epoch": 0.9,
      "learning_rate": 7.500000000000001e-05,
      "loss": 1.0633,
      "step": 1500
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0001,
      "loss": 1.0197,
      "step": 2000
    },
    {
      "epoch": 1.5,
      "learning_rate": 9.211356466876973e-05,
      "loss": 0.9953,
      "step": 2500
    },
    {
      "epoch": 1.8,
      "learning_rate": 8.422712933753944e-05,
      "loss": 0.9726,
      "step": 3000
    },
    {
      "epoch": 2.1,
      "learning_rate": 7.634069400630916e-05,
      "loss": 0.9514,
      "step": 3500
    },
    {
      "epoch": 2.4,
      "learning_rate": 6.845425867507887e-05,
      "loss": 0.9302,
      "step": 4000
    },
    {
      "epoch": 2.7,
      "learning_rate": 6.056782334384858e-05,
      "loss": 0.9206,
      "step": 4500
    },
    {
      "epoch": 3.0,
      "learning_rate": 5.2681388012618305e-05,
      "loss": 0.9098,
      "step": 5000
    },
    {
      "epoch": 3.3,
      "learning_rate": 4.4794952681388013e-05,
      "loss": 0.8869,
      "step": 5500
    },
    {
      "epoch": 3.6,
      "learning_rate": 3.690851735015773e-05,
      "loss": 0.8798,
      "step": 6000
    },
    {
      "epoch": 3.9,
      "learning_rate": 2.9022082018927448e-05,
      "loss": 0.8706,
      "step": 6500
    },
    {
      "epoch": 4.2,
      "learning_rate": 2.113564668769716e-05,
      "loss": 0.8603,
      "step": 7000
    },
    {
      "epoch": 4.5,
      "learning_rate": 1.3249211356466879e-05,
      "loss": 0.8558,
      "step": 7500
    },
    {
      "epoch": 4.8,
      "learning_rate": 5.3627760252365935e-06,
      "loss": 0.8524,
      "step": 8000
    },
    {
      "epoch": 5.0,
      "step": 8340,
      "total_flos": 4.06800921526272e+16,
      "train_loss": 0.9733701600731134,
      "train_runtime": 6864.6479,
      "train_samples_per_second": 19.438,
      "train_steps_per_second": 1.215
    }
  ],
  "logging_steps": 500,
  "max_steps": 8340,
  "num_train_epochs": 5,
  "save_steps": 2000,
  "total_flos": 4.06800921526272e+16,
  "trial_name": null,
  "trial_params": null
}