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 supported9 Subscribers
Add a CodeTriage badge to deepspeed
Help out
- Issues
- Error while saving T5-11B checkpoint
- [BUG] a huge memory leak when using `register_full_backward_hook`
- Questions about the AWM in the ZeRO-Infinity paper
- [BUG] memory overhead issue with optimizer leading to OOM
- [BUG] fp16 optimizers: printing log info on all ranks
- [BUG] ZeRO-1 hangs in optimizer step when used in Pipeline
- Multi Node Distributing - RuntimeError: Connection reset by peer
- [REQUEST] Model serving via deepspeed's inference module
- [QUESTION] About Pipeline Parallelism save_checkpoint
- [QUESTION] Skipping examples with NaN loss
- Docs
- Python not yet supported