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
- Make PipelineModule.set_checkpoint_interval actually change the interval
- Add REFUTE scientific critique + calibration benchmark
- [BUG] autotuning get_val_by_key stops at the first nested subdict, missing keys in sibling subdicts
- Fix autotuning get_val_by_key to search all nested subdicts
- Stabilize DeepCompile ZeRO-3 memory scheduling
- Stabilize fork-sensitive nightly tests and AutoSP coverage
- [RFC] Support sharding LM heads and adopting Online Softmax
- Enable support for Tutel when k != 1 for shared moe
- Guard LRRangeTest and OneCycle schedulers against zero step sizes
- Fix WarmupLR collapsing multi-group base LRs to group 0's
- Docs
- Python not yet supported