trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 8979,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.17,
      "learning_rate": 4.7215725581913354e-05,
      "loss": 1.6236,
      "step": 500
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.443145116382671e-05,
      "loss": 1.525,
      "step": 1000
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.164717674574006e-05,
      "loss": 1.4783,
      "step": 1500
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.8862902327653414e-05,
      "loss": 1.4528,
      "step": 2000
    },
    {
      "epoch": 0.84,
      "learning_rate": 3.607862790956677e-05,
      "loss": 1.4441,
      "step": 2500
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.329435349148012e-05,
      "loss": 1.4137,
      "step": 3000
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.0510079073393477e-05,
      "loss": 1.3863,
      "step": 3500
    },
    {
      "epoch": 1.34,
      "learning_rate": 2.7725804655306832e-05,
      "loss": 1.3796,
      "step": 4000
    },
    {
      "epoch": 1.5,
      "learning_rate": 2.4941530237220183e-05,
      "loss": 1.3692,
      "step": 4500
    },
    {
      "epoch": 1.67,
      "learning_rate": 2.2157255819133534e-05,
      "loss": 1.3598,
      "step": 5000
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.937298140104689e-05,
      "loss": 1.3458,
      "step": 5500
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.6588706982960243e-05,
      "loss": 1.3418,
      "step": 6000
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.3804432564873596e-05,
      "loss": 1.3283,
      "step": 6500
    },
    {
      "epoch": 2.34,
      "learning_rate": 1.1020158146786947e-05,
      "loss": 1.3118,
      "step": 7000
    },
    {
      "epoch": 2.51,
      "learning_rate": 8.235883728700302e-06,
      "loss": 1.3161,
      "step": 7500
    },
    {
      "epoch": 2.67,
      "learning_rate": 5.451609310613654e-06,
      "loss": 1.3111,
      "step": 8000
    },
    {
      "epoch": 2.84,
      "learning_rate": 2.6673348925270076e-06,
      "loss": 1.3042,
      "step": 8500
    },
    {
      "epoch": 3.0,
      "step": 8979,
      "total_flos": 7.5623004331776e+16,
      "train_loss": 1.3888828911883806,
      "train_runtime": 5861.5075,
      "train_samples_per_second": 49.006,
      "train_steps_per_second": 1.532
    }
  ],
  "logging_steps": 500,
  "max_steps": 8979,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 7.5623004331776e+16,
  "trial_name": null,
  "trial_params": null
}