Decodering: Get started with local LLMs using ollama! AND Navi intro
Details
The open-source Ollama server makes it straightforward to run a variety of models (Mistral, Gemma, Phi, Llama) on your own machine in a podman/docker container. The advantage of running them inside a container is that you can experiment with them and then discard when you are bored - and the queries stay on your own machine, of course.
This talk will show you how to get Mistral, a pretty decent LLM chatbot, up and running in short order without any fuss! To follow along, you will need:
- `podman` or `docker`
- `make` or `just`
- and a reasonably fast internet connection!
NAVI:
If you often forget commands, hate context switching, and want to automate more of what you do, this talk aims to help you do that!
It will be a quick overview of a tool I've found really neat over the last few months and has saved me loads of time - Navi (https://github.com/denisidoro/navi)
It’s an interactive command cheat sheet with dynamic customisable placeholders. Great for making more use of CLIs without learning them by hand, and working in the terminal faster