TensorFlow 2.0 Beta is available Learn more

tfma.evaluators.MetricsAndPlotsEvaluator

tfma.evaluators.MetricsAndPlotsEvaluator(
    eval_shared_model,
    desired_batch_size=None,
    metrics_key=constants.METRICS_KEY,
    plots_key=constants.PLOTS_KEY,
    run_after=slice_key_extractor.SLICE_KEY_EXTRACTOR_STAGE_NAME,
    num_bootstrap_samples=1
)

Defined in evaluators/metrics_and_plots_evaluator.py.

Creates an Evaluator for evaluating metrics and plots.

Args:

  • eval_shared_model: Shared model parameters for EvalSavedModel.
  • desired_batch_size: Optional batch size for batching in Aggregate.
  • metrics_key: Name to use for metrics key in Evaluation output.
  • plots_key: Name to use for plots key in Evaluation output.
  • run_after: Extractor to run after (None means before any extractors).
  • num_bootstrap_samples: Number of bootstrap samples to draw. If more than 1, confidence intervals will be computed for metrics. Suggested value is at least 20.

Returns:

Evaluator for evaluating metrics and plots. The output will be stored under 'metrics' and 'plots' keys.