dsh-llm-qwen-local

English | 简体中文

DeepSeek Harness LLM adapter plugin for a locally deployed Qwen model (e.g. Qwen3.8-27B) served by vLLM behind its OpenAI-compatible /v1/chat/completions endpoint.

v0.3.1 · exact compatibility target: DSH 0.1.1-rc.2 · MIT · community-maintained and not a DeepSeek or Qwen product.

dsh plugin --profile web add dsh-llm-qwen-local

Two deployment-specific knobs are first-class:

• Per-model multimodal switch (multimodal: true/false) — declares whether the deployment serves the model with vision.
• Fully configurable reasoning efforts — every selectable level, its display name, its reasoning_effort wire spelling, the default level, and how off is expressed on the wire all come from configuration, matching whatever vocabulary your vLLM build accepts.

- id: llm-qwen-local
  name: dsh-llm-qwen-local
  config:
    baseURL: http://127.0.0.1:8000/v1
    models:
      - id: qwen3.8
        name: Qwen3.8 (local)
        multimodal: true
        reasoning:
          efforts:
            - { id: off, wire: none }
            - { id: low, wire: low }
            - { id: medium, wire: medium }
            - { id: xhigh, wire: xhigh }
          defaultEffort: xhigh

Documentation

	English	中文
Installation & usage	(this README)	(此 README)
Configuration reference — every field	docs/configuration.md	docs/configuration.zh.md
Design notes — wire dialect, model parameters, framework compatibility, error paths, limitations	docs/design.md	docs/design.zh.md

Requirements

• An installed dsh (the CLI) 0.1.1-rc.2 or newer, and a vLLM instance serving your Qwen model with the OpenAI-compatible API.
• Node.js with global fetch (18+).
• A profile whose composition mounts @deepseek-ai/dsh-attachment — the standard web and headless profiles do, via dsh-base.

Required vLLM serve flags (per the official vLLM recipe): --reasoning-parser qwen3 is effectively mandatory — without it the whole reasoning block lands in message.content — plus --enable-auto-tool-choice --tool-call-parser qwen3_coder for tool calling and --max-model-len 262144 (or higher).

Install

# install from npm (recommended — prebuilt, no build step on install):
dsh plugin --profile web add dsh-llm-qwen-local

# install from git (the prepare script builds lib/ on install):
dsh plugin --profile web add github:starefinger/dsh-llm-qwen-local

# or from a local checkout (same prepare build runs on install):
dsh plugin --profile web add ./path/to/qwen3.8-LLM-plugin

# or from a packed tarball (prebuilt — no build step on install):
dsh plugin --profile web add ./dsh-llm-qwen-local-0.3.1.tgz

# verify the contributed layer, then start:
dsh --profile web --dump-config
dsh --profile web

Version-pinned install (tag)

…