trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9673590504451037,
  "eval_steps": 200,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.29673590504451036,
      "grad_norm": 4.922755241394043,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.6132,
      "step": 100
    },
    {
      "epoch": 0.5934718100890207,
      "grad_norm": 4.277930736541748,
      "learning_rate": 8.933333333333333e-06,
      "loss": 0.303,
      "step": 200
    },
    {
      "epoch": 0.5934718100890207,
      "eval_loss": 0.24335336685180664,
      "eval_runtime": 493.5362,
      "eval_samples_per_second": 4.652,
      "eval_steps_per_second": 0.146,
      "eval_wer": 0.4226068014038485,
      "step": 200
    },
    {
      "epoch": 0.8902077151335311,
      "grad_norm": 4.590033054351807,
      "learning_rate": 7.822222222222224e-06,
      "loss": 0.2564,
      "step": 300
    },
    {
      "epoch": 1.1869436201780414,
      "grad_norm": 3.3569984436035156,
      "learning_rate": 6.711111111111111e-06,
      "loss": 0.2,
      "step": 400
    },
    {
      "epoch": 1.1869436201780414,
      "eval_loss": 0.20352379977703094,
      "eval_runtime": 482.7378,
      "eval_samples_per_second": 4.756,
      "eval_steps_per_second": 0.149,
      "eval_wer": 0.39138327483964663,
      "step": 400
    },
    {
      "epoch": 1.4836795252225519,
      "grad_norm": 3.2953569889068604,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.1637,
      "step": 500
    },
    {
      "epoch": 1.7804154302670623,
      "grad_norm": 3.363754987716675,
      "learning_rate": 4.488888888888889e-06,
      "loss": 0.1633,
      "step": 600
    },
    {
      "epoch": 1.7804154302670623,
      "eval_loss": 0.18764939904212952,
      "eval_runtime": 484.1747,
      "eval_samples_per_second": 4.742,
      "eval_steps_per_second": 0.149,
      "eval_wer": 0.3469079026987777,
      "step": 600
    },
    {
      "epoch": 2.077151335311573,
      "grad_norm": 2.2284867763519287,
      "learning_rate": 3.377777777777778e-06,
      "loss": 0.1428,
      "step": 700
    },
    {
      "epoch": 2.373887240356083,
      "grad_norm": 2.0610995292663574,
      "learning_rate": 2.266666666666667e-06,
      "loss": 0.106,
      "step": 800
    },
    {
      "epoch": 2.373887240356083,
      "eval_loss": 0.18498285114765167,
      "eval_runtime": 480.7097,
      "eval_samples_per_second": 4.776,
      "eval_steps_per_second": 0.15,
      "eval_wer": 0.3487837347210456,
      "step": 800
    },
    {
      "epoch": 2.6706231454005933,
      "grad_norm": 2.554095983505249,
      "learning_rate": 1.1555555555555556e-06,
      "loss": 0.1029,
      "step": 900
    },
    {
      "epoch": 2.9673590504451037,
      "grad_norm": 2.727163314819336,
      "learning_rate": 4.444444444444445e-08,
      "loss": 0.1005,
      "step": 1000
    },
    {
      "epoch": 2.9673590504451037,
      "eval_loss": 0.18132926523685455,
      "eval_runtime": 479.4261,
      "eval_samples_per_second": 4.789,
      "eval_steps_per_second": 0.15,
      "eval_wer": 0.3367421033522934,
      "step": 1000
    },
    {
      "epoch": 2.9673590504451037,
      "step": 1000,
      "total_flos": 9.234732878187725e+18,
      "train_loss": 0.21518024158477783,
      "train_runtime": 7211.1645,
      "train_samples_per_second": 4.438,
      "train_steps_per_second": 0.139
    }
  ],
  "logging_steps": 100,
  "max_steps": 1000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.234732878187725e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}