trainer_state.json

{
  "best_metric": 12.457650398315174,
  "best_model_checkpoint": "./checkpoint-300",
  "epoch": 1.1027522935779817,
  "global_step": 300,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 0.5068,
      "step": 25
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.2745,
      "step": 50
    },
    {
      "epoch": 0.28,
      "learning_rate": 7.3e-06,
      "loss": 0.2341,
      "step": 75
    },
    {
      "epoch": 0.37,
      "learning_rate": 9.800000000000001e-06,
      "loss": 0.2097,
      "step": 100
    },
    {
      "epoch": 0.37,
      "eval_loss": 0.26155176758766174,
      "eval_runtime": 6495.4175,
      "eval_samples_per_second": 0.446,
      "eval_steps_per_second": 0.056,
      "eval_wer": 17.6700851570369,
      "step": 100
    },
    {
      "epoch": 0.46,
      "learning_rate": 8.85e-06,
      "loss": 0.2161,
      "step": 125
    },
    {
      "epoch": 0.55,
      "learning_rate": 7.600000000000001e-06,
      "loss": 0.2043,
      "step": 150
    },
    {
      "epoch": 0.64,
      "learning_rate": 6.35e-06,
      "loss": 0.1803,
      "step": 175
    },
    {
      "epoch": 0.73,
      "learning_rate": 5.150000000000001e-06,
      "loss": 0.1578,
      "step": 200
    },
    {
      "epoch": 0.73,
      "eval_loss": 0.21081551909446716,
      "eval_runtime": 6471.6616,
      "eval_samples_per_second": 0.447,
      "eval_steps_per_second": 0.056,
      "eval_wer": 14.098983609559564,
      "step": 200
    },
    {
      "epoch": 0.83,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.1696,
      "step": 225
    },
    {
      "epoch": 0.92,
      "learning_rate": 2.6500000000000005e-06,
      "loss": 0.144,
      "step": 250
    },
    {
      "epoch": 1.01,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 0.1274,
      "step": 275
    },
    {
      "epoch": 1.1,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 0.0806,
      "step": 300
    },
    {
      "epoch": 1.1,
      "eval_loss": 0.18695282936096191,
      "eval_runtime": 6462.2032,
      "eval_samples_per_second": 0.448,
      "eval_steps_per_second": 0.056,
      "eval_wer": 12.457650398315174,
      "step": 300
    },
    {
      "epoch": 1.1,
      "step": 300,
      "total_flos": 1.02039599259648e+19,
      "train_loss": 0.20877531051635742,
      "train_runtime": 20970.8306,
      "train_samples_per_second": 0.229,
      "train_steps_per_second": 0.014
    }
  ],
  "max_steps": 300,
  "num_train_epochs": 2,
  "total_flos": 1.02039599259648e+19,
  "trial_name": null,
  "trial_params": null
}