vllm-project/vllm-ascend

[Contribution] 任务 #1：vllm主社区多模态能力加强，视频场景evs能力压缩

Open

#10,648 opened on Jun 18, 2026

View on GitHub
 (3 comments) (0 reactions) (0 assignees)C++ (1,318 forks)github user discovery
help wanted

Repository metrics

Stars
 (2,180 stars)
PR merge metrics
 (Avg merge 4d 5h) (559 merged PRs in 30d)

Description

在提交新问题之前，请确保您已经在社区中搜索过相关问题，并使用了社区中提供的资源/工具后，仍未找到满意的解决方式。

📚 任务问题描述

任务描述

该任务拆分自 #9079，任务 ID：#1。

vllm主社区目前已经支持视频压缩的evs能力，介绍见 https://docs.vllm.ai/en/stable/api/vllm/multimodal/evs/ ，接入vllm-ascend插件后，让该特性和对应接口可以在NPU上也进行，拿到对应的收益。

验收标准

  1. 使用qwen3.5模型，打通evs框架上能力，收益符合预期。
  2. 配套使用说明及对应用例上库，社区提交RFC及对应的特性使用指导。

对接人

@jyoung6652

认领状态

待认领

建议的替代/修复方案

请根据任务描述提交实现、测试或文档 PR，并在 PR 中关联本 issue 与 #9079；如实现过程中发现任务边界、使用指导或现有行为存在不合理之处，请同步补充说明和最小复现。

欢迎加入社区，感谢您对社区的贡献 🎉!

Contributor guide

Research direction
Study the existing EVS implementation in vLLM main repo (docs/api/vllm/multimodal/evs/) and the vllm ascend plugin structure. Identify the interfaces that need to be adapted for NPU. Implement the EVS integration in vllm ascend, test with Qwen3.5 model, and document usage. Submit a PR linking this issue and #9079.
Tech stack
python
Domain
aibackend
Issue type
Feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
3-5 days
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
PythonvLLMAscend NPUEVS
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
40