deepspeed
https://github.com/microsoft/deepspeed
Python
DeepSpeed is a deep learning optimization library that makes distributed training easy, efficient, and effective.
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 supported21 Subscribers
Add a CodeTriage badge to deepspeed
Help out
- Issues
- Universal checkpoint conversion only knows Adam's state keys, so a Muon checkpoint cannot be converted
- Report the unscaled gradient norm from FP16_UnfusedOptimizer
- Decay OneCycle momentum instead of growing it past 1.0
- Fix: Replace assert statements with ValueError in runtime config validation
- init_inference rejects dtype=torch.int8 on every accelerator
- Fix: handle None buffers in AutoTP checkpoint load
- Document Windows source-checkout symlink workaround
- 2 tests fail: assert isinstance(model.llm[0], DistributedAttention)
- Keep runtime configuration validation active under python -O
- [BUG] Runtime configuration validation is bypassed under python -O
- Docs
- Python not yet supported