Keras Inference Monitoring and Profiling

See the Quick Start guide on how to install and configure Graphsignal.

Use a callback for evaluation and prediction phases:

from graphsignal.callbacks.keras import GraphsignalCallback

model.predict(..., callbacks=[GraphsignalCallback(endpoint='predict')])
# or model.evaluate(..., callbacks=[GraphsignalCallback(endpoint='predict')])


The Keras MNIST example illustrates how to add the callback.

Model serving

Graphsignal provides a built-in support for server applications. See Model Serving guide for more information.