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 supported20 Subscribers
Add a CodeTriage badge to deepspeed
Help out
- Issues
- Fix DeepCompile ZeRO-3 gathered parameter ownership
- Scope DeepCompile compiler state to graph and engine lifecycles
- Fix ZeRO-3 all_reduce param fetch stride for padded parameters
- compute_elastic_config raises bare ZeroDivisionError for return_microbatch=True on non-0.2 elasticity versions without explicit world_size
- Clamp capacity to num_tokens in MoE gating drop branches (complete #5353)
- [BUG] AutoEP + ZeRO-1/2 universal checkpoint conversion silently leaves expert optimizer momentum stale/EP-unaware
- Add logging support for better observability
- fix: add zero-guards for divisor/scale edge cases (#7838)
- Add governance process page to deepspeed.ai
- Feature/zero multi loss separate backward
- Docs
- Python not yet supported