from keras import metrics
model.compile(loss='mean_squared_error',
              optimizer='sgd',
              metrics=[metrics.mae, metrics.categorical_accuracy])

A metric function is similar to a , except that the results from evaluating a metric are not used when training the model. You may use any of the loss functions as a metric function.

Arguments

• y_true: True labels. Theano/TensorFlow tensor.
• y_pred: Predictions. Theano/TensorFlow tensor of the same shape as y_true.

Returns

Single tensor value representing the mean of the output array across all datapoints.

keras.metrics.accuracy(y_true, y_pred)

binary_accuracy

categorical_accuracy

top_k_categorical_accuracy

keras.metrics.top_k_categorical_accuracy(y_true, y_pred, k=5)

sparse_top_k_categorical_accuracy

keras.metrics.sparse_top_k_categorical_accuracy(y_true, y_pred, k=5)

clone_metric

keras.metrics.clone_metric(metric)

clone_metrics

keras.metrics.clone_metrics(metrics)

In addition to the metrics above, you may use any of the loss functions described in the loss function page as metrics.

def mean_pred(y_true, y_pred):
    return K.mean(y_pred)
model.compile(optimizer='rmsprop',
              loss='binary_crossentropy',