Skip to content

Commit

Permalink
[AIR/Docs] Small improvements to Train user guide (#26577)
Browse files Browse the repository at this point in the history
Co-authored-by: matthewdeng <[email protected]>
  • Loading branch information
Yard1 and matthewdeng authored Jul 16, 2022
1 parent 6217138 commit fb6f3cf
Show file tree
Hide file tree
Showing 5 changed files with 323 additions and 264 deletions.
2 changes: 2 additions & 0 deletions doc/source/ray-air/checkpoints.rst
Original file line number Diff line number Diff line change
@@ -1,3 1,5 @@
.. _air-checkpoints:

Checkpoints
===========

Expand Down
3 changes: 2 additions & 1 deletion doc/source/ray-air/key-concepts.rst
Original file line number Diff line number Diff line change
Expand Up @@ -39,13 39,14 @@ See the documentation on :ref:`Trainers <air-trainer-ref>`.



Trainer objects will produce a :ref:`Results <air-results-ref>` object after calling ``.fit()``. These objects will contain training metrics as long as checkpoints to retrieve the best model.
Trainer objects will produce a :ref:`Result <air-results-ref>` object after calling ``.fit()``. These objects will contain training metrics as long as checkpoints to retrieve the best model.

.. literalinclude:: doc_code/air_key_concepts.py
:language: python
:start-after: __air_trainer_output_start__
:end-before: __air_trainer_output_end__

.. _air-session-key-concepts:

Session
-------
Expand Down
4 changes: 2 additions & 2 deletions doc/source/ray-air/package-ref.rst
Original file line number Diff line number Diff line change
Expand Up @@ -45,6 45,8 @@ Abstract Classes
:members:
:show-inheritance:

.. _air-results-ref:

Training Result
###############

Expand Down Expand Up @@ -113,8 115,6 @@ Serving

.. autoclass:: ray.serve.air_integrations.PredictorWrapper

.. _air-results-ref:

Trainer and Predictor Integrations
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Expand Down
2 changes: 1 addition & 1 deletion doc/source/ray-air/preprocessors.rst
Original file line number Diff line number Diff line change
Expand Up @@ -92,7 92,7 @@ the ``Preprocessor`` and the fitting and transformation logic will occur once pe
Checkpoint
~~~~~~~~~~

``Trainer.fit()`` returns a ``Results`` object which contains a ``Checkpoint``.
``Trainer.fit()`` returns a ``Result`` object which contains a ``Checkpoint``.
If a ``Preprocessor`` was passed into the ``Trainer``, then it will be saved in the ``Checkpoint`` along with any fitted state.

As a sanity check, let's confirm the ``Preprocessor`` is available in the ``Checkpoint``. In practice you should not need to do this.
Expand Down
Loading

0 comments on commit fb6f3cf

Please sign in to comment.