Home Server Experiments
Can a mining card help run a language model at home?
My home server experiments include repurposing a mining card as an extra GPU and running an LLM on my own hardware. Anyone with a Gmail account can try talking to it; its personality is still a work in progress.
What is this project?
I call it servOS: a home-built, local-first system for private conversation, speech, making, and observability. sahil.ing is its public window, with a machine overview and coarse service health while operations stay private.
One experiment is repurposing a mining card as an extra GPU. The LLM runs on my home server, and anyone with a Gmail account can try talking to it through Talk.
The model’s personality is still on my to-do list. For now, this is an experiment you can try while I work out what I want it to become.