trainer_state.json

{
  "best_metric": 0.26802200078964233,
  "best_model_checkpoint": "./xls-r-1b-bem-fsv/checkpoint-1000",
  "epoch": 4.990403071017274,
  "eval_steps": 500,
  "global_step": 1300,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.7677543186180422,
      "grad_norm": 7.090035915374756,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 4.9139,
      "step": 200
    },
    {
      "epoch": 1.5355086372360844,
      "grad_norm": 2.271395683288574,
      "learning_rate": 3.960000000000001e-05,
      "loss": 2.6967,
      "step": 400
    },
    {
      "epoch": 1.9193857965451055,
      "eval_loss": 0.4789997339248657,
      "eval_runtime": 47.1718,
      "eval_samples_per_second": 10.578,
      "eval_steps_per_second": 1.336,
      "eval_wer": 0.9859719438877755,
      "step": 500
    },
    {
      "epoch": 2.3032629558541267,
      "grad_norm": 1.4331015348434448,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3851,
      "step": 600
    },
    {
      "epoch": 3.071017274472169,
      "grad_norm": 0.58216392993927,
      "learning_rate": 3.15e-05,
      "loss": 0.257,
      "step": 800
    },
    {
      "epoch": 3.838771593090211,
      "grad_norm": 0.7515180110931396,
      "learning_rate": 1.9e-05,
      "loss": 0.1815,
      "step": 1000
    },
    {
      "epoch": 3.838771593090211,
      "eval_loss": 0.26802200078964233,
      "eval_runtime": 46.97,
      "eval_samples_per_second": 10.624,
      "eval_steps_per_second": 1.341,
      "eval_wer": 0.8797595190380761,
      "step": 1000
    },
    {
      "epoch": 4.606525911708253,
      "grad_norm": 0.4551451504230499,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.1455,
      "step": 1200
    },
    {
      "epoch": 4.990403071017274,
      "step": 1300,
      "total_flos": 1.1936063667019008e+19,
      "train_loss": 1.3302699947357177,
      "train_runtime": 3428.0062,
      "train_samples_per_second": 6.072,
      "train_steps_per_second": 0.379
    }
  ],
  "logging_steps": 200,
  "max_steps": 1300,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 1.1936063667019008e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}