{
  "best_metric": 0.9801412180052956,
  "best_model_checkpoint": "audio/train/checkpoint/whisper-tiny/superb_ks_42/checkpoint-15970",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 15970,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.6261740763932373,
      "grad_norm": 17.53887176513672,
      "learning_rate": 3.1308703819661866e-05,
      "loss": 0.8486,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9385113268608414,
      "eval_loss": 0.23804208636283875,
      "eval_runtime": 29.9994,
      "eval_samples_per_second": 226.605,
      "eval_steps_per_second": 0.9,
      "step": 1597
    },
    {
      "epoch": 1.2523481527864746,
      "grad_norm": 0.1602136492729187,
      "learning_rate": 4.859806581785292e-05,
      "loss": 0.1349,
      "step": 2000
    },
    {
      "epoch": 1.878522229179712,
      "grad_norm": 0.37793028354644775,
      "learning_rate": 4.51193209490016e-05,
      "loss": 0.0986,
      "step": 3000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9598411297440423,
      "eval_loss": 0.1777295619249344,
      "eval_runtime": 27.7034,
      "eval_samples_per_second": 245.385,
      "eval_steps_per_second": 0.975,
      "step": 3194
    },
    {
      "epoch": 2.504696305572949,
      "grad_norm": 15.802234649658203,
      "learning_rate": 4.1640576080150286e-05,
      "loss": 0.0773,
      "step": 4000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9738158281847602,
      "eval_loss": 0.12492171674966812,
      "eval_runtime": 28.5966,
      "eval_samples_per_second": 237.721,
      "eval_steps_per_second": 0.944,
      "step": 4791
    },
    {
      "epoch": 3.1308703819661865,
      "grad_norm": 10.594197273254395,
      "learning_rate": 3.8161831211298965e-05,
      "loss": 0.0669,
      "step": 5000
    },
    {
      "epoch": 3.757044458359424,
      "grad_norm": 0.004856509622186422,
      "learning_rate": 3.468308634244764e-05,
      "loss": 0.0532,
      "step": 6000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9782288908502501,
      "eval_loss": 0.1077924370765686,
      "eval_runtime": 30.2849,
      "eval_samples_per_second": 224.468,
      "eval_steps_per_second": 0.892,
      "step": 6388
    },
    {
      "epoch": 4.383218534752661,
      "grad_norm": 0.8512524962425232,
      "learning_rate": 3.120434147359633e-05,
      "loss": 0.0472,
      "step": 7000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9766107678729038,
      "eval_loss": 0.12581859529018402,
      "eval_runtime": 25.8895,
      "eval_samples_per_second": 262.577,
      "eval_steps_per_second": 1.043,
      "step": 7985
    },
    {
      "epoch": 5.009392611145898,
      "grad_norm": 0.2138238549232483,
      "learning_rate": 2.7725596604745007e-05,
      "loss": 0.0411,
      "step": 8000
    },
    {
      "epoch": 5.635566687539136,
      "grad_norm": 0.004156901966780424,
      "learning_rate": 2.4246851735893692e-05,
      "loss": 0.0322,
      "step": 9000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9771991762283024,
      "eval_loss": 0.13651490211486816,
      "eval_runtime": 27.6161,
      "eval_samples_per_second": 246.161,
      "eval_steps_per_second": 0.978,
      "step": 9582
    },
    {
      "epoch": 6.261740763932373,
      "grad_norm": 1.517451286315918,
      "learning_rate": 2.076810686704237e-05,
      "loss": 0.0357,
      "step": 10000
    },
    {
      "epoch": 6.88791484032561,
      "grad_norm": 0.0043836492113769054,
      "learning_rate": 1.7289361998191053e-05,
      "loss": 0.0258,
      "step": 11000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9798470138275963,
      "eval_loss": 0.13380160927772522,
      "eval_runtime": 28.81,
      "eval_samples_per_second": 235.96,
      "eval_steps_per_second": 0.937,
      "step": 11179
    },
    {
      "epoch": 7.514088916718848,
      "grad_norm": 0.0014675065176561475,
      "learning_rate": 1.3810617129339735e-05,
      "loss": 0.0231,
      "step": 12000
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.979552809649897,
      "eval_loss": 0.14468975365161896,
      "eval_runtime": 26.2171,
      "eval_samples_per_second": 259.296,
      "eval_steps_per_second": 1.03,
      "step": 12776
    },
    {
      "epoch": 8.140262993112085,
      "grad_norm": 3.176596164703369,
      "learning_rate": 1.0331872260488417e-05,
      "loss": 0.021,
      "step": 13000
    },
    {
      "epoch": 8.766437069505322,
      "grad_norm": 0.0021598339080810547,
      "learning_rate": 6.853127391637097e-06,
      "loss": 0.0171,
      "step": 14000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9796999117387467,
      "eval_loss": 0.14350397884845734,
      "eval_runtime": 25.9468,
      "eval_samples_per_second": 261.997,
      "eval_steps_per_second": 1.041,
      "step": 14373
    },
    {
      "epoch": 9.39261114589856,
      "grad_norm": 0.0006765660946257412,
      "learning_rate": 3.3743825227857787e-06,
      "loss": 0.0142,
      "step": 15000
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9801412180052956,
      "eval_loss": 0.14339740574359894,
      "eval_runtime": 26.2356,
      "eval_samples_per_second": 259.114,
      "eval_steps_per_second": 1.029,
      "step": 15970
    },
    {
      "epoch": 10.0,
      "step": 15970,
      "total_flos": 5.6903389435008e+18,
      "train_loss": 0.09692646174708529,
      "train_runtime": 3764.7682,
      "train_samples_per_second": 135.716,
      "train_steps_per_second": 4.242
    }
  ],
  "logging_steps": 1000,
  "max_steps": 15970,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 5.6903389435008e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}