Performance Analysis
====================

.. :py:currentmodule:: spdl.pipeline

When optimizing ML model pipeline, it is important to determine if data loading is bottleneck,
and where the bottleneck is.

SPDL facilitates the pipeline performance analysis. :py:class:`Pipeline` records runtime
performance statistics and export them.
In this section, we look at how to log the performance statistics and
analyze the pipeline performance.

.. toctree::

   setup
   logging
   analysis