skills.rest
vLLM-HUST
vLLM-HUSTOFFICIAL
Explore all skills by this author →
WebsiteGitHub

vllm-ascend-model-adapter

Adapt and validate models for vLLM on Ascend hardware with a signed commit.

3|17|Updated Mar 21, 2026
Search Tags:#deployment#validation#vllm#ascend#model-adaptation#commit-signoff
One-click install
npx skills add https://github.com/vLLM-HUST/vllm-ascend-hust --skill vllm-ascend-model-adapter-vllm-hust
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: vllm-ascend-model-adapter
Source: https://github.com/vLLM-HUST/vllm-ascend-hust/tree/main/.agents/skills/vllm-ascend-model-adapter
Command: npx skills add https://github.com/vLLM-HUST/vllm-ascend-hust --skill vllm-ascend-model-adapter-vllm-hust

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Adapt and debug existing or new models for vLLM on Ascend NPU. Implement in /vllm-workspace/vllm and /vllm-workspace/vllm-ascend, validate via direct vllm serve from /workspace, and deliver one signed commit in the current repo.

Core Features & Use Cases

  • Adapt Hugging Face or local models to run on vLLM Ascend with minimal changes and deterministic validation.
  • Validate model integration via vllm serve from /workspace and sign off with a single commit in the repo.

Quick Start

Run the two-stage serve workflow from /workspace using the provided model path and default port to validate startup and a smoke request.

Frequently Asked Questions about vllm-ascend-model-adapter

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I adapt a Hugging Face model to run on vLLM Ascend?▼

Model adaptation for vLLM Ascend is automated by implementing minimal code changes in /vllm-workspace/vllm and /vllm-workspace/vllm-ascend, validated through deterministic testing via vllm serve on port 8000.

What is the two-stage validation workflow for vLLM model adaptation?▼

The two-stage validation workflow for vLLM model adaptation first tests startup and smoke requests using dummy weights, then validates with real weights before delivering a single signed commit.

How do I validate model integration on Ascend hardware using vllm serve?▼

Validate model integration on Ascend hardware by starting vllm serve directly from /workspace on port 8000, which executes deterministic testing to confirm the adapted model runs correctly on Ascend NPU.

Does vLLM model adaptation for Ascend require manual commit signing?▼

No, vLLM model adaptation for Ascend automates the delivery process by requiring minimal code changes and producing an auditable final delivery with a single signed commit in the current repository.

Can I debug a new model for vLLM on Ascend NPU with deterministic testing?▼

Yes, you can debug new models for vLLM on Ascend NPU using deterministic testing, which validates the adaptation through direct vllm serve from /workspace before signing off the commit.

Related Skills

Explore More Software Engineering Skills→
Alex-stack-hubAlex-stack-hub

vllm-ascend-model-adapter

Adapt Hugging Face or local models for vLLM on Ascend NPU.

Community
Advanced
handsomelong922handsomelong922

serving-llms-vllm

Deploy scalable LLM serving with vLLM for OpenAI-compatible endpoints.

Community
Advanced
nv-actionnv-action
4

vllm-ascend-model-adapter

Adapt and validate models for vLLM Ascend NPU with a signed commit.

Official
Advanced
iani-kuliiani-kuli

huggingface-community-evals

Execute local Hugging Face model evaluations with inspect-ai or lighteval frameworks.

Community
Intermediate