trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 200,
  "global_step": 1968,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.41,
      "eval_loss": 5.472805500030518,
      "eval_runtime": 162.595,
      "eval_samples_per_second": 4.078,
      "eval_steps_per_second": 0.51,
      "eval_wer": 1.075673807878369,
      "step": 200
    },
    {
      "epoch": 0.81,
      "eval_loss": 5.127437114715576,
      "eval_runtime": 163.1607,
      "eval_samples_per_second": 4.063,
      "eval_steps_per_second": 0.509,
      "eval_wer": 1.003800967519005,
      "step": 400
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0007890792291220557,
      "loss": 6.5037,
      "step": 500
    },
    {
      "epoch": 1.22,
      "eval_loss": 0.6166694760322571,
      "eval_runtime": 161.3235,
      "eval_samples_per_second": 4.11,
      "eval_steps_per_second": 0.514,
      "eval_wer": 0.8130615065653075,
      "step": 600
    },
    {
      "epoch": 1.63,
      "eval_loss": 0.328411728143692,
      "eval_runtime": 162.053,
      "eval_samples_per_second": 4.091,
      "eval_steps_per_second": 0.512,
      "eval_wer": 0.582930200414651,
      "step": 800
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0005214132762312634,
      "loss": 1.0482,
      "step": 1000
    },
    {
      "epoch": 2.03,
      "eval_loss": 0.3169207274913788,
      "eval_runtime": 165.1028,
      "eval_samples_per_second": 4.016,
      "eval_steps_per_second": 0.503,
      "eval_wer": 0.5666897028334485,
      "step": 1000
    },
    {
      "epoch": 2.44,
      "eval_loss": 0.28758111596107483,
      "eval_runtime": 160.9496,
      "eval_samples_per_second": 4.119,
      "eval_steps_per_second": 0.516,
      "eval_wer": 0.5425017277125086,
      "step": 1200
    },
    {
      "epoch": 2.85,
      "eval_loss": 0.2846720516681671,
      "eval_runtime": 161.8788,
      "eval_samples_per_second": 4.096,
      "eval_steps_per_second": 0.513,
      "eval_wer": 0.5521769177608846,
      "step": 1400
    },
    {
      "epoch": 3.05,
      "learning_rate": 0.00025374732334047106,
      "loss": 0.4314,
      "step": 1500
    },
    {
      "epoch": 3.25,
      "eval_loss": 0.27460750937461853,
      "eval_runtime": 160.7333,
      "eval_samples_per_second": 4.125,
      "eval_steps_per_second": 0.516,
      "eval_wer": 0.5393918451969593,
      "step": 1600
    },
    {
      "epoch": 3.66,
      "eval_loss": 0.26981213688850403,
      "eval_runtime": 160.1257,
      "eval_samples_per_second": 4.14,
      "eval_steps_per_second": 0.518,
      "eval_wer": 0.5345542501727713,
      "step": 1800
    },
    {
      "epoch": 4.0,
      "step": 1968,
      "total_flos": 2.984133759613272e+18,
      "train_loss": 2.1205503649827913,
      "train_runtime": 5112.0325,
      "train_samples_per_second": 1.54,
      "train_steps_per_second": 0.385
    }
  ],
  "logging_steps": 500,
  "max_steps": 1968,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 200,
  "total_flos": 2.984133759613272e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}