.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "auto_examples/resume.py"
.. LINE NUMBERS ARE GIVEN BELOW.

.. only:: html

    .. note::
        :class: sphx-glr-download-link-note

        :ref:`Go to the end <sphx_glr_download_auto_examples_resume.py>`
        to download the full example code.

.. rst-class:: sphx-glr-example-title

.. _sphx_glr_auto_examples_resume.py:


Coming back after model training
================================

.. GENERATED FROM PYTHON SOURCE LINES 7-17

After training a model, you might want to test its performance, make
predictions or do whatever you want with it.

.. note::
    This example assumes:
        * Training was performed with :doc:`AIdsorb CLI <../cli>` or :ref:`AIdsorb +
          PyTorch Lightning <aidsorb_with_pytorch_and_lightning>`.
        * `PyTorch Lightning checkpoints
          <https://lightning.ai/docs/pytorch/stable/common/checkpointing_basic.html#lightningmodule-from-checkpoint>`_
          are enabled during training.

.. GENERATED FROM PYTHON SOURCE LINES 17-37

.. code-block:: Python


    import lightning as L
    import torch

    from aidsorb.datamodules import PCDDataModule
    from aidsorb.litmodules import PCDLit

    # Restore lightning modules from checkpoint.
    ckpt_path = 'path/to/checkpoint.ckpt'
    litmodel = PCDLit.load_from_checkpoint(ckpt_path)
    dm = PCDDataModule.load_from_checkpoint(ckpt_path)

    # Set the model for inference (disable grads & enable eval mode).
    litmodel.freeze()
    print(f'Model in evaluation mode: {not litmodel.training}')

    # Your code goes here.
    ...



.. GENERATED FROM PYTHON SOURCE LINES 38-40

Measure performance
-------------------

.. GENERATED FROM PYTHON SOURCE LINES 40-47

.. code-block:: Python


    # Instantiate a trainer object.
    trainer = L.Trainer(...)

    # Measure performance on test set.
    trainer.test(litmodel, datamodule=dm)


.. GENERATED FROM PYTHON SOURCE LINES 48-50

Make predictions
----------------

.. GENERATED FROM PYTHON SOURCE LINES 50-59

.. code-block:: Python


    # Setup the datamodule.
    dm.setup()

    # Predict on the test set.
    y_pred = torch.cat(trainer.predict(litmodel, dm.test_dataloader()))

    # Predict on the train set.
    y_pred = torch.cat(trainer.predict(litmodel, dm.train_dataloader()))


.. _sphx_glr_download_auto_examples_resume.py:

.. only:: html

  .. container:: sphx-glr-footer sphx-glr-footer-example

    .. container:: sphx-glr-download sphx-glr-download-jupyter

      :download:`Download Jupyter notebook: resume.ipynb <resume.ipynb>`

    .. container:: sphx-glr-download sphx-glr-download-python

      :download:`Download Python source code: resume.py <resume.py>`

    .. container:: sphx-glr-download sphx-glr-download-zip

      :download:`Download zipped: resume.zip <resume.zip>`


.. only:: html

 .. rst-class:: sphx-glr-signature

    `Gallery generated by Sphinx-Gallery <https://sphinx-gallery.github.io>`_