Simon Willison’s Weblog


3 items tagged “justine-tunney”


gemma-2-27b-it-llamafile (via) Justine Tunney shipped llamafile packages of Google's new openly licensed (though definitely not open source) Gemma 2 27b model this morning.

I downloaded the gemma-2-27b-it.Q5_1.llamafile version (20.5GB) to my Mac, ran chmod 755 gemma-2-27b-it.Q5_1.llamafile and then ./gemma-2-27b-it.Q5_1.llamafile and now I'm trying it out through the llama.cpp default web UI in my browser. It works great.

It's a very capable model - currently sitting at position 12 on the LMSYS Arena making it the highest ranked open weights model - one position ahead of Llama-3-70b-Instruct and within striking distance of the GPT-4 class models.

# 2nd July 2024, 10:38 pm / google, ai, generative-ai, homebrew-llms, llms, llamafile, justine-tunney


llamafile is the new best way to run a LLM on your own computer

Visit llamafile is the new best way to run a LLM on your own computer

Mozilla’s innovation group and Justine Tunney just released llamafile, and I think it’s now the single best way to get started running Large Language Models (think your own local copy of ChatGPT) on your own computer.

[... 650 words]


Cosmopolitan: Compiling Python. Cosmopolitan is Justine Tunney’s “build-once run-anywhere C library”—part of the αcτµαlly pδrταblε εxεcµταblε effort, which produces wildly clever binary executable files that work on multiple different platforms, and is the secret sauce behind redbean. I hadn’t realized this was happening but there’s an active project to get Python to work as this format, producing a new way of running Python applications as standalone executables, only these ones have the potential to run unmodified on Windows, Linux and macOS.

# 26th July 2022, 8:43 pm / python, redbean, cosmopolitan, justine-tunney