.. _l-template-ml:

Templates
=========

.. contents::
    :local:

Scikit-learn template
+++++++++++++++++++++

Classes :class:`MLStorage <lightmlrestapi.mlapp.mlstorage.MLStorage>` and
:class:`MLStoragePost <lightmlrestapi.mlapp.mlstorage_rest.MLStoragePost>`
assume that a model is actionable by implementing
the following template:

.. runpython::

    from lightmlrestapi.testing.template_ml import __file__

    with open(__file__, "r", encoding="utf-8") as f:
        print(f.read())

Model with image
++++++++++++++++

The second template shows how to deal with images
with a dummy example which computes the distance
between two images.

.. runpython::

    from lightmlrestapi.testing.template_dl_light import __file__

    with open(__file__, "r", encoding="utf-8") as f:
        print(f.read())

Model with keras
++++++++++++++++

Thrid template with :epkg:`keras`
and a model trained on :epkg:`ImageNet`.

.. runpython::

    from lightmlrestapi.testing.template_dl_keras import __file__

    with open(__file__, "r", encoding="utf-8") as f:
        print(f.read())