checkpoint-3078/trainer_state.json

{
  "best_metric": 0.41836002868754485,
  "best_model_checkpoint": "w2v-bert-2.0-CV_Fleurs-lg-10hrs-v4/checkpoint-3078",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 3078,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 30.792112350463867,
      "learning_rate": 9.998031496062993e-05,
      "loss": 1.4106,
      "step": 513
    },
    {
      "epoch": 1.0,
      "eval_cer": 0.1264185884550246,
      "eval_loss": 0.5293365716934204,
      "eval_runtime": 31.0205,
      "eval_samples_per_second": 26.628,
      "eval_steps_per_second": 6.673,
      "eval_wer": 0.5665790102797036,
      "step": 513
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.16055679321289,
      "learning_rate": 9.897047244094488e-05,
      "loss": 0.4775,
      "step": 1026
    },
    {
      "epoch": 2.0,
      "eval_cer": 0.11182513176846737,
      "eval_loss": 0.4882170259952545,
      "eval_runtime": 30.4398,
      "eval_samples_per_second": 27.135,
      "eval_steps_per_second": 6.8,
      "eval_wer": 0.5251016017212526,
      "step": 1026
    },
    {
      "epoch": 3.0,
      "grad_norm": 24.60751724243164,
      "learning_rate": 9.796062992125985e-05,
      "loss": 0.3696,
      "step": 1539
    },
    {
      "epoch": 3.0,
      "eval_cer": 0.09981164627486981,
      "eval_loss": 0.397329181432724,
      "eval_runtime": 30.5019,
      "eval_samples_per_second": 27.08,
      "eval_steps_per_second": 6.786,
      "eval_wer": 0.4564905570164953,
      "step": 1539
    },
    {
      "epoch": 4.0,
      "grad_norm": 8.973501205444336,
      "learning_rate": 9.695078740157481e-05,
      "loss": 0.3007,
      "step": 2052
    },
    {
      "epoch": 4.0,
      "eval_cer": 0.09322717991737761,
      "eval_loss": 0.3895888030529022,
      "eval_runtime": 29.8209,
      "eval_samples_per_second": 27.699,
      "eval_steps_per_second": 6.941,
      "eval_wer": 0.4311498924217069,
      "step": 2052
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.461097240447998,
      "learning_rate": 9.594094488188978e-05,
      "loss": 0.2564,
      "step": 2565
    },
    {
      "epoch": 5.0,
      "eval_cer": 0.09659855331676664,
      "eval_loss": 0.3743041455745697,
      "eval_runtime": 29.9644,
      "eval_samples_per_second": 27.566,
      "eval_steps_per_second": 6.908,
      "eval_wer": 0.4372459956968683,
      "step": 2565
    },
    {
      "epoch": 6.0,
      "grad_norm": 9.60593032836914,
      "learning_rate": 9.493110236220472e-05,
      "loss": 0.2273,
      "step": 3078
    },
    {
      "epoch": 6.0,
      "eval_cer": 0.0879880973108153,
      "eval_loss": 0.3549309968948364,
      "eval_runtime": 30.2119,
      "eval_samples_per_second": 27.34,
      "eval_steps_per_second": 6.852,
      "eval_wer": 0.41836002868754485,
      "step": 3078
    }
  ],
  "logging_steps": 500,
  "max_steps": 51300,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 100,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 20,
        "early_stopping_threshold": 0.001
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6.544282515665142e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}