deepspeedai / deepspeedai/DeepSpeed

[REQUEST] Support for CUDA Graphs

Open
#1,746 2 comments 2 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

enhancement
Dominant language
Python
Stars
43.1k
Forks
5k
Avg merge
4d 15h
Merged PRs (30d)
112

Description

Does DeepSpeed support Pytorch code with CUDA Graphs? If not, do think it may be helpful to DeepSpeed users for further speedups?

Context: I am trying to optimize my training code that can benefit from launching multiple kernels together.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reviewing the linked PyTorch CUDA Graphs documentation and the DeepSpeed integration points used by the reported training code. Determine whether CUDA Graphs are currently supported and define the DeepSpeed changes and validation needed for kernel-capture speedups; the issue provides no specific files or tests.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
machine-learning, performance
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.