Module: tf.contrib.eager.metrics

View source on GitHub

Metrics namespace.

Classes

class Accuracy: Calculates how often predictions matches labels.

class BinaryAccuracy: Calculates how often predictions matches labels.

class CategoricalAccuracy: Calculates how often predictions matches labels.

class Mean: Computes the (weighted) mean of the given values.

class Metric: A metric holds state for aggregating statistics over an evaluation run.

class SparseAccuracy: Calculates how often predictions matches labels.