Skip to content

Comment on Mesh LLM: distributed AI computing on iroh

Comments

Does this support Qwen 3.6 (e.g. 27B) and the myriad of llama.cpp options (batch sizes, quantization, etc.)?

I'd love to see some performance data.

Yes! I’ve worked on the settings interface between our runtime and llamacpp, these are documented and available via our config.toml file

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.