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
- Honor CPU launch mode and update minimum-Torch CI
- Check trained results and requested precision
- test2
- gather_for_metrics drops or duplicates examples with SubsetRandomSampler
- [FSDP2] Full QLoRA support: PEFT adapter exclusion, LoRA-param upcast, detect helpers
- Fix pad across processes keep dtype
- Warn when DeepSpeed ZeRO silently changes custom optimizer semantics
- DataLoaderDispatcher.total_batch_size raises AttributeError with ordinary datasets
- Fix total batch size metadata for dispatched dataloaders
- Fix split-batch shard length for uneven tails
- Docs
- Python not yet supported