sgl-project/sglang

[Feature] Enable EPLB in Draft Models

Open

#7.893 aberto em 9 de jul. de 2025

Ver no GitHub
 (6 comments) (0 reactions) (1 assignee)Python (6.216 forks)auto 404
good first issue

Métricas do repositório

Stars
 (28.442 stars)
Métricas de merge de PR
 (Mesclagem média 2d 1h) (1.000 fundiu PRs em 30d)

Description

Checklist

Motivation

Currently, EPLB is not supported in draft models, which constrains the parallelism size. For example, with EPLB, we can set the EP size as 72/144 for the DeepSeek model as it has 288 experts in total. However, these parallelism settings cannot be adopted when MTP is enabled as the draft model as has 256 experts.

Related resources

No response

Guia do colaborador