accelerate
https://github.com/huggingface/accelerate
Python
🚀 A simple way to train and use PyTorch models with multi-GPU, TPU, mixed-precision
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
Python not yet supported1 Subscribers
Add a CodeTriage badge to accelerate
Help out
- Issues
- Fix metric padding removal after skip_first_batches
- Fix boolean tensor memory accounting
- Fix dtype_byte_size reporting zero bytes for sub-byte and packed dtypes
- Add an FSDP2-native Ulysses sequence parallelism backend
- Skip automatic DataLoader sharding when the loader is already rank-sharded
- `use_stateful_dataloader` + `num_workers>0` draws epoch-0 permutation during `prepare()` before cross-rank RNG sync — corrupts DDP data partition
- Ask transformers whether a model supports context parallelism
- `use_stateful_dataloader` under multi-process training restores the cursor, not the shuffle order — sampler permutation is never serialized
- Allow tensor parallelism next to replicated data parallelism
- Fix stateful dataloader checkpointing across processes
- Docs
- Python not yet supported