The short version
Community Jinja chat templates tuned for running Qwen models with vLLM.
From the HackerLinks archive
Qwen Fixed Chat Templates
Community Jinja chat templates tuned for running Qwen models with vLLM.
At a glance:
First seen:2026-09-02
Last seen:2026-09-02
Times seen:1
Website:huggingface.co
Why it caught our attention
A production user reported good agentic results and emphatically recommended the templates.
Where it surfaced on Hacker News
2026-09-02
“Don't sleep on the froggeric templates.”
hadlock · recommendation · production use
Direct HN comment
Show HN: Running 104GB Qwen3.8-Flash-Next on 48GB Mac with at ~12 tok/s