NVIDIA-NeMo / NVIDIA-NeMo/Automodel

Support configurable HSDP in both expert and non-expert groups for MoEs

Open
#2,267 0 comments 1 reaction 0 assignees View on GitHub

Nobody has claimed this yet.

enhancement
Dominant language
Python
Stars
963
Forks
318
Avg merge
3d 20h
Merged PRs (30d)
143

Description

From Cursor 2.5 blog:

HSDP forms multiple FSDP replicas and all-reduces gradients across corresponding shards. We use separate HSDP layouts for non-expert and expert weights: non-expert weights are comparatively small, so their FSDP groups can stay narrow, often within a node or rack, while expert weights hold most of the parameters and most of the Muon compute, so they use a wider expert sharding mesh.

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

The issue names no files, tests, or entry points. Start by locating the existing HSDP and FSDP layout handling for mixture-of-experts models, then determine how expert and non-expert groups are configured. Done means both groups support configurable HSDP layouts with coverage for the relevant training paths.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
distributed-systems, machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
30/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.