Welcome to the TorchSnapshot documentation!
===========================================

TorchSnapshot is a PyTorch library for adding fault tolerance to large-scale PyTorch distributed training workloads.

`Installation instructions <https://github.com/pytorch/torchsnapshot#readme>`_


TorchSnapshot API
-----------------

.. toctree::
   :maxdepth: 2
   :caption: Contents:

   getting_started.rst
   api_reference.rst

Examples
--------

* `Simple example <https://github.com/pytorch/torchsnapshot/blob/main/examples/simple_example.py>`_
* `Using TorchSnapshot with DistributedDataParallel (DDP) <https://github.com/pytorch/torchsnapshot/blob/main/examples/ddp_example.py>`_
* `Using TorchSnapshot with TorchRec <https://github.com/pytorch/torchsnapshot/blob/main/examples/torchrec_example.py>`_