sgl-project/sglang

[Feature] Support `return_hidden_states` in async engine

Closed

#6,528 opened on May 22, 2025

View on GitHubContributor guide
 (1 comment) (0 reactions) (0 assignees)Python (6,216 forks)auto 404
good first issue

Repository metrics

Stars
 (28,442 stars)
PR merge metrics
 (Avg merge 2d 1h) (1,000 merged PRs in 30d)

Description

Checklist

Motivation

Currently, it is possible to get hidden states using the offline interface, i.e. Engine.generate, but Engine.async_generate doesn't support this function.

result = await model.async_generate(
        ["hello", "how are you doing?"],
        sampling_params={"temperature": 0, "max_new_tokens": 1},
        return_hidden_states=True)

TypeError: Engine.async_generate() got an unexpected keyword argument 'return_hidden_states'

This can be useful when developing web services where some additional hidden states postprocess logics is done.

Related resources

No response

Contributor guide

Research direction
Examine how `return hidden states` is handled in `Engine.generate` and add the same support to `async generate`.
Tech stack
python
Domain
backend
Issue type
Feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
2
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-3 hours
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
Python
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
70