Skip to content

A performant, memory-efficient checkpointing library for PyTorch applications, designed with large, complex distributed workloads in mind.

License

Notifications You must be signed in to change notification settings

galrotem/torchsnapshot

 
 

Repository files navigation

TorchSnapshot (Beta Release)

build status pypi version conda version pypi nightly version codecov bsd license

A performant, memory-efficient checkpointing library for PyTorch applications, designed with large, complex distributed workloads in mind.

Install

Requires Python >= 3.8 and PyTorch >= 2.0.0

From pip:

# Stable
pip install torchsnapshot
# Or, using conda
conda install -c conda-forge torchsnapshot

# Nightly
pip install --pre torchsnapshot-nightly

From source:

git clone https://github.com/pytorch/torchsnapshot
cd torchsnapshot
pip install -r requirements.txt
python setup.py install

Why TorchSnapshot

Performance

  • TorchSnapshot provides a fast checkpointing implementation employing various optimizations, including zero-copy serialization for most tensor types, overlapped device-to-host copy and storage I/O, parallelized storage I/O.
  • TorchSnapshot greatly speeds up checkpointing for DistributedDataParallel workloads by distributing the write load across all ranks (benchmark).
  • When host memory is abundant, TorchSnapshot allows training to resume before all storage I/O completes, reducing the time blocked by checkpoint saving.

Memory Usage

  • TorchSnapshot's memory usage adapts to the host's available resources, greatly reducing the chance of out-of-memory issues when saving and loading checkpoints.
  • TorchSnapshot supports efficient random access to individual objects within a snapshot, even when the snapshot is stored in a cloud object storage.

Usability

  • Simple APIs that are consistent between distributed and non-distributed workloads.
  • Out of the box integration with commonly used cloud object storage systems.
  • Automatic resharding (elasticity) on world size change for supported workloads (more details).

Security

  • Secure tensor serialization without pickle dependency [WIP].

Getting Started

from torchsnapshot import Snapshot

# Taking a snapshot
app_state = {"model": model, "optimizer": optimizer}
snapshot = Snapshot.take(path="/path/to/snapshot", app_state=app_state)

# Restoring from a snapshot
snapshot.restore(app_state=app_state)

See the documentation for more details.

License

torchsnapshot is BSD licensed, as found in the LICENSE file.

About

A performant, memory-efficient checkpointing library for PyTorch applications, designed with large, complex distributed workloads in mind.

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 99.7%
  • Shell 0.3%