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
- [BUG]DeepSpeed ZeRO-2 Offload: backward does not populate IPG buckets even though autograd graph reaches parameters (PyTorch 2.0 / Python 3.8 / DeepSpeed 0.17.6)
- [REQUEST] How to achieve DeepNVMe (ZeRO 3 offloading) via GPU Direct Storage?
- [BUG] DeepCompile: Training hang on random-sized inputs
- [REQUEST] Can deepspeed initialize multiple datasets/dataloaders?
- Git tag and version.txt mismatch
- [REQUEST] Add optional "subsequence overlap" to UlyssesSPDataLoaderAdapter
- [REQUEST] Fast access to whole model gradient during training
- Question about Deepspeed-Domino
- [BUG]how to use zero_to_fp32 when my model list is like below
- [REQUEST]Support for modules_to_save with LoRA under DeepSpeed ZeRO-3
- Docs
- Python not yet supported