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
- FSDP 2 + cpu_ram_efficient_loading=True fails for models with buffers (e.g. DS 3.1)
- FSDP2: Activation-checkpoint modules directly matching auto-wrap policy, not those whose parent module matches policy
- Docs: clarify MS-AMP limitations and alternatives
- Docs: index distributed inference examples
- `datasets` Iterable Dataset sharding support
- Deadlock when running multi tasks
- Feature Request: Add API to Materialize Meta Tensors for Device Movement
- Feature request: FSDP2 QLoRA
- Fail to `load_state` immediately after `save_state` with FSDP plugins.
- activation_checkpointing=True uses *a lot* more memory than gradient_checkpointing=True *with FSDP 2.0* (FSDP 1.0 is fine)
- Docs
- Python not yet supported