.. Copyright © 2022 Idiap Research Institute <contact@idiap.ch>
..
.. SPDX-License-Identifier: GPL-3.0-or-later

.. _mednet.config:

Preset Configurations
---------------------

This module contains preset configurations for baseline CNN architectures and
DataModules.


.. _mednet.config.models:

Pre-configured Models
=====================

Pre-configured models you can readily use.

.. autosummary::
   :toctree: api/config.models
   :template: config.rst

   mednet.config.models.alexnet
   mednet.config.models.alexnet_pretrained
   mednet.config.models.densenet
   mednet.config.models.densenet_pretrained
   mednet.config.models.densenet_rs
   mednet.config.models.logistic_regression
   mednet.config.models.mlp
   mednet.config.models.pasa


.. _mednet.config.datamodules:

DataModule support
==================

Base DataModules and raw data loaders for the various databases currently
supported in this package, for your reference.  Each pre-configured DataModule
can receive the name of one or more splits as argument to build a fully
functional DataModule that can be used in training, prediction or testing.

.. autosummary::
   :toctree: api/config.datamodules

   mednet.config.data.hivtb.datamodule
   mednet.config.data.indian.datamodule
   mednet.config.data.montgomery.datamodule
   mednet.config.data.montgomery_shenzhen.datamodule
   mednet.config.data.montgomery_shenzhen_indian.datamodule
   mednet.config.data.montgomery_shenzhen_indian_padchest.datamodule
   mednet.config.data.montgomery_shenzhen_indian_tbx11k.datamodule
   mednet.config.data.nih_cxr14.datamodule
   mednet.config.data.nih_cxr14_padchest.datamodule
   mednet.config.data.padchest.datamodule
   mednet.config.data.shenzhen.datamodule
   mednet.config.data.tbpoc.datamodule
   mednet.config.data.tbx11k.datamodule


.. _mednet.config.datamodule-instances:

Pre-configured DataModules
==========================

DataModules provide access to preset pytorch dataloaders for training,
validating, testing and running prediction tasks.  Each of the pre-configured
DataModule is based on one (or more) of the :ref:`supported base DataModules
<mednet.config.datamodules>`.

.. autosummary::
   :toctree: api/config.datamodule-instances
   :template: config.rst

   mednet.config.data.indian.default
   mednet.config.data.montgomery.default
   mednet.config.data.montgomery_shenzhen.default
   mednet.config.data.montgomery_shenzhen_indian.default
   mednet.config.data.montgomery_shenzhen_indian_padchest.default
   mednet.config.data.montgomery_shenzhen_indian_tbx11k.v1_healthy_vs_atb
   mednet.config.data.montgomery_shenzhen_indian_tbx11k.v2_others_vs_atb
   mednet.config.data.nih_cxr14.default
   mednet.config.data.nih_cxr14_padchest.idiap
   mednet.config.data.padchest.idiap
   mednet.config.data.shenzhen.default
   mednet.config.data.tbx11k.v1_healthy_vs_atb
   mednet.config.data.tbx11k.v2_others_vs_atb
   mednet.config.data.tbx11k.v2_others_vs_atb


.. _mednet.config.datamodule-instances.folds:

Cross-validation DataModules
============================

We support cross-validation with precise preset folds.  In this section, you
will find the configuration for the first fold (fold-0) for all supported
DataModules.  Nine other folds are available for every configuration (from 1 to
9), making up 10 folds per supported DataModule.


.. autosummary::
   :toctree: api/config.datamodule-folds
   :template: config.rst

   mednet.config.data.hivtb.fold_0
   mednet.config.data.indian.fold_0
   mednet.config.data.montgomery.fold_0
   mednet.config.data.montgomery_shenzhen.fold_0
   mednet.config.data.montgomery_shenzhen_indian.fold_0
   mednet.config.data.montgomery_shenzhen_indian_tbx11k.v1_fold_0
   mednet.config.data.montgomery_shenzhen_indian_tbx11k.v2_fold_0
   mednet.config.data.shenzhen.fold_0
   mednet.config.data.tbpoc.fold_0
   mednet.config.data.tbx11k.v1_fold_0
   mednet.config.data.tbx11k.v2_fold_0


.. include:: links.rst