ashpy.metrics¶

Collection of Metrics.

Metric

metric.Metric Metric is the abstract class that every AshPy Metric must implement.

Classifier

classifier.ClassifierLoss A handy way to measure the classification loss.
classifier.ClassifierMetric Wrap a metric using argmax to extract predictions out of a classifier’s output.

GAN

gan.DiscriminatorLoss The Discriminator loss value.
gan.GeneratorLoss Generator loss value.
gan.EncoderLoss Encoder Loss value.
gan.InceptionScore Inception Score Metric.
gan.EncodingAccuracy Generator and Encoder accuracy performance.
sliced_wasserstein_metric.SlicedWassersteinDistance Sliced Wasserstein Distance.
ssim_multiscale.SSIM_Multiscale Multiscale Structural Similarity.

Modules

classifier The classification metrics.
gan GAN metrics.
metric Metric is the abstract class that every ash metric must implement.
sliced_wasserstein_metric Sliced Wasserstein Distance metric.
ssim_multiscale Multiscale Structural Similarity metric.