tf.keras.metrics.Metric

View source on GitHub

Encapsulates metric logic and state.

Inherits From: Layer

View aliases

Main aliases

tf.metrics.Metric

Compat aliases for migration

See Migration guide for more details.

tf.compat.v1.keras.metrics.Metric

tf.keras.metrics.Metric(
    name=None, dtype=None, **kwargs
)

Args
name	(Optional) string name of the metric instance.
dtype	(Optional) data type of the metric result.
**kwargs	Additional layer keywords arguments.

Standalone usage:

m = SomeMetric(...)
for input in ...:
  m.update_state(input)
print('Final result: ', m.result().numpy())

Usage with compile() API:

model = tf.keras.Sequential()
model.add(tf.keras.layers.Dense(64, activation='relu'))
model.add(tf.keras.layers.Dense(64, activation='relu'))
model.add(tf.keras.layers.Dense(10, activation='softmax'))

model.compile(optimizer=tf.keras.optimizers.RMSprop(0.01),
              loss=tf.keras.losses.CategoricalCrossentropy(),
              metrics=[tf.keras.metrics.CategoricalAccuracy()])

data = np.random.random((1000, 32))
labels = np.random.random((1000, 10))

dataset = tf.data.Dataset.from_tensor_slices((data, labels))
dataset = dataset.batch(32)

model.fit(dataset, epochs=10)

To be implemented by subclasses:

• __init__(): All state variables should be created in this method by calling self.add_weight() like: self.var = self.add_weight(...)
• update_state(): Has all updates to the state variables like: self.var.assign_add(...).
• result(): Computes and returns a value for the metric from the state variables.

Example subclass implementation:

class BinaryTruePositives(tf.keras.metrics.Metric):

  def __init__(self, name='binary_true_positives', **kwargs):
    super(BinaryTruePositives, self).__init__(name=name, **kwargs)
    self.true_positives = self.add_weight(name='tp', initializer='zeros')

  def update_state(self, y_true, y_pred, sample_weight=None):
    y_true = tf.cast(y_true, tf.bool)
    y_pred = tf.cast(y_pred, tf.bool)

    values = tf.logical_and(tf.equal(y_true, True), tf.equal(y_pred, True))
    values = tf.cast(values, self.dtype)
    if sample_weight is not None:
      sample_weight = tf.cast(sample_weight, self.dtype)
      sample_weight = tf.broadcast_to(sample_weight, values.shape)
      values = tf.multiply(values, sample_weight)
    self.true_positives.assign_add(tf.reduce_sum(values))

  def result(self):
    return self.true_positives

Methods

add_weight

View source

add_weight(
    name, shape=(), aggregation=tf.compat.v1.VariableAggregation.SUM,
    synchronization=tf.VariableSynchronization.ON_READ, initializer=None, dtype=None
)

Adds state variable. Only for use by subclasses.

reset_states

View source

reset_states()

Resets all of the metric state variables.

This function is called between epochs/steps, when a metric is evaluated during training.

result

View source

@abc.abstractmethod
result()

Computes and returns the metric value tensor.

Result computation is an idempotent operation that simply calculates the metric value using the state variables.

update_state

View source

@abc.abstractmethod
update_state(
    *args, **kwargs
)

Accumulates statistics for the metric.

Note: This function is executed as a graph function in graph mode. This means: a) Operations on the same resource are executed in textual order. This should make it easier to do things like add the updated value of a variable to another, for example. b) You don't need to worry about collecting the update ops to execute. All update ops added to the graph by this function will be executed. As a result, code should generally work the same way with graph or eager execution.

Args
*args
**kwargs	A mini-batch of inputs to the Metric.