My LLM CLI tool for running prompts against large language models from the terminal (and logging the results to SQLite) can now be installed from Homebrew!
brew install llm
More details here: https://llm.datasette.io/
@simon Does this work on the linux version of homebrew too?
@mauve theoretically it should, but I haven't tried it myself
@simon It works! The ggml-replit model is having trouble with this fib function and got confused generatting error handling code. :P Are there any gpt4all things that are good with code completion? Or do they also need special treatment for how to prompt them?
@mauve I've honestly not spent enough time with the gpt4all models to have a great feel for how to get the best results out of them yet
@simon That's fair. For what it's worth I've been having decent luck with orca mini. I'll need to do some postprocessing to fetch just the code blocks out of responses though.
BTW! Sent in a PR to fix a bug when trying to run a gpt4all model offline. https://github.com/simonw/llm-gpt4all/pull/9
@mauve shipped, thank you!