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
- [Feature request] Support already-sharded DataLoaders in Accelerator.prepare
- fix(modeling): reject sharded-checkpoint weight_map entries that escape the checkpoint folder (#4067)
- fix: preserve dataloader iteration state when resuming from checkpoint
- [Security] Path traversal + malicious-model DoS via sharded-checkpoint `weight_map` in ….
- Question: Manual Dataset Sharding per GPU Rank with Accelerate + DistributedSampler (Avoid Double DataLoader Length Split)
- Add register_tracker_class to register custom trackers by name
- [Windows] RTX 5070 Ti (Blackwell sm_120) - setup and deployment notes
- `datasets` Iterable Dataset sharding support
- Feature Request: Add API to Materialize Meta Tensors for Device Movement
- Feature request: FSDP2 QLoRA
- Docs
- Python not yet supported