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 BatchSamplerShard.__len__ overcounting when split_batches=True and even_batches=False
- BatchSamplerShard.__len__ overcounts for some ranks when split_batches=True and even_batches=False
- Add missing -> bool return type hints in utils/imports.py
- Fix write_basic_config clobbering MLU config to CPU
- Fix IterableDatasetShard.__len__ to account for split_batches
- Fix concatenate() for non-tensor batch fields across multiple batches
- docs: clarify multi-process non-tensor batch limitation
- Fix environment conversion docstring example
- Fix typos in comments, docs, and examples
- concatenate() still raises TypeError on non-tensor batch fields when num_processes > 1
- Docs
- Python not yet supported