About

I like running LLMs on my own hardware. I don't like maintaining launch scripts, manually running benchmarks, and documenting results. That's why I started automating the process, and this website shows the results.

I hope the site also helps fellow enthusiasts to see what they can expect from some of the hardware available to us. And maybe how to get inference going by building on the published launch commands.

If you find an error or have feedback, feel free to drop me an email at click.