trainer_state.json

{
  "best_metric": 3.7265218830814386,
  "best_model_checkpoint": "./checkpoint-100",
  "epoch": 1.0,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.25,
      "learning_rate": 5e-08,
      "loss": 0.1602,
      "step": 25
    },
    {
      "epoch": 0.5,
      "learning_rate": 1e-07,
      "loss": 0.1588,
      "step": 50
    },
    {
      "epoch": 0.5,
      "eval_loss": 0.14945247769355774,
      "eval_runtime": 6007.4304,
      "eval_samples_per_second": 1.55,
      "eval_steps_per_second": 0.097,
      "eval_wer": 3.7649995973262462,
      "step": 50
    },
    {
      "epoch": 0.75,
      "learning_rate": 5e-08,
      "loss": 0.1321,
      "step": 75
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0,
      "loss": 0.1267,
      "step": 100
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.12812651693820953,
      "eval_runtime": 5985.3223,
      "eval_samples_per_second": 1.556,
      "eval_steps_per_second": 0.097,
      "eval_wer": 3.7265218830814386,
      "step": 100
    },
    {
      "epoch": 1.0,
      "step": 100,
      "total_flos": 1.846946562048e+18,
      "train_loss": 0.144483003616333,
      "train_runtime": 13002.902,
      "train_samples_per_second": 0.492,
      "train_steps_per_second": 0.008
    }
  ],
  "logging_steps": 25,
  "max_steps": 100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 50,
  "total_flos": 1.846946562048e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}