Skip to content
Achu Mukundan
Work Notes Resume Contact

Notes

Writing about local inference, coding tools, and audio software as I work on them.

Ditching llm-scaler: Qwen3.8-27B on my B70

Updated 2026-09-12

Leaving llm-scaler behind, patching vLLM with GPT-5.6 Sol, and getting MTP4 working. My setup as of August 20.

© 2026 Achu Mukundan. Toronto, Canada.
Email GitHub LinkedIn