deepspeedai / deepspeedai/DeepSpeed

[REQUEST] Mixture of Experts (MoE) Segmentation Task

Open
#3,701 0 comments 0 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

Feature relates to MoE End-to-End inference
i would to know if MoE used in DeepSpeed can implement in the Segmentation task

Describe the solution
i was working on the MoE problem I read the paper on DeepSpeed-MoE i read the documentation and I found only work on the MLP Linear Gate Network,

Feature
make MoE in DeepSpeed support Sgemnatattion

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 reading the DeepSpeed-MoE documentation and the existing MLP Linear Gate Network support mentioned in the issue. Determine the requirements and integration points for using MoE in a segmentation task and for end-to-end inference. Done should mean that segmentation is supported and documented with a working inference path.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
18/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.