trainer_state.json

{
  "best_metric": 2.877540349960327,
  "best_model_checkpoint": "/scratch/skscla001/speech/results/wav2vec2-300m-librispeech-test/checkpoint-500",
  "epoch": 2.0,
  "eval_steps": 100,
  "global_step": 690,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2898550724637681,
      "grad_norm": 0.9390692114830017,
      "learning_rate": 0.00029699999999999996,
      "loss": 6.037,
      "step": 100
    },
    {
      "epoch": 0.2898550724637681,
      "eval_loss": 2.9196243286132812,
      "eval_runtime": 150.904,
      "eval_samples_per_second": 17.256,
      "eval_steps_per_second": 4.314,
      "eval_wer": 1.0,
      "step": 100
    },
    {
      "epoch": 0.5797101449275363,
      "grad_norm": 0.5911043882369995,
      "learning_rate": 0.0002496610169491525,
      "loss": 2.8559,
      "step": 200
    },
    {
      "epoch": 0.5797101449275363,
      "eval_loss": 2.888011932373047,
      "eval_runtime": 150.2657,
      "eval_samples_per_second": 17.329,
      "eval_steps_per_second": 4.332,
      "eval_wer": 1.0,
      "step": 200
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 0.5367380380630493,
      "learning_rate": 0.0001988135593220339,
      "loss": 2.8547,
      "step": 300
    },
    {
      "epoch": 0.8695652173913043,
      "eval_loss": 2.8900482654571533,
      "eval_runtime": 149.7042,
      "eval_samples_per_second": 17.394,
      "eval_steps_per_second": 4.349,
      "eval_wer": 1.0,
      "step": 300
    },
    {
      "epoch": 1.1594202898550725,
      "grad_norm": 0.5129782557487488,
      "learning_rate": 0.00014796610169491524,
      "loss": 2.8418,
      "step": 400
    },
    {
      "epoch": 1.1594202898550725,
      "eval_loss": 2.9081904888153076,
      "eval_runtime": 149.1662,
      "eval_samples_per_second": 17.457,
      "eval_steps_per_second": 4.364,
      "eval_wer": 1.0,
      "step": 400
    },
    {
      "epoch": 1.4492753623188406,
      "grad_norm": 0.4716997444629669,
      "learning_rate": 9.71186440677966e-05,
      "loss": 2.8436,
      "step": 500
    },
    {
      "epoch": 1.4492753623188406,
      "eval_loss": 2.877540349960327,
      "eval_runtime": 149.8814,
      "eval_samples_per_second": 17.374,
      "eval_steps_per_second": 4.343,
      "eval_wer": 1.0,
      "step": 500
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 0.19197961688041687,
      "learning_rate": 4.6271186440677964e-05,
      "loss": 2.8373,
      "step": 600
    },
    {
      "epoch": 1.7391304347826086,
      "eval_loss": 2.882120132446289,
      "eval_runtime": 149.6054,
      "eval_samples_per_second": 17.406,
      "eval_steps_per_second": 4.351,
      "eval_wer": 1.0,
      "step": 600
    },
    {
      "epoch": 2.0,
      "step": 690,
      "total_flos": 2.1389390179654113e+18,
      "train_loss": 3.307427580460258,
      "train_runtime": 2183.7213,
      "train_samples_per_second": 2.527,
      "train_steps_per_second": 0.316
    }
  ],
  "logging_steps": 100,
  "max_steps": 690,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 400,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.1389390179654113e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}