If you can't make it live, register anyways to have access to the VOD after the event: go.chainguard.dev/ll-raspberry...

If you're curious, I'll be presenting this demo next week at @chainguard.dev 's Learning Labs. Here's the registration link!

go.chainguard.dev/ll-raspberry...

This is running on a containerized environment based on Wolfi, as a result the image is smaller than the official llama.cpp image and has zero CVEs. Everything OSS based, free, running locally!

For example, I started digging into open source LLMs and llama.cpp for help with ALT descriptions for images. I got my own personal assistant chat running on a Raspberry Pi 5 in my local network, and it is working pretty well for describing images.

PROST!

Glad people are taking this issue more seriously and the whole industry is now paying attention to software supply chain risks. When I joined Chainguard 3+ years ago, even the term was new to me - I never really thought about the software supply chain, always taking open source for granted.

Plugged directly into your build and deployment workflows, dependency management tools have privileged access to your production codebase. You're installing third party software based on trust, but you're not overseeing every single link in the chain to validate the origin of these artifacts.

Dependency management tools were a crucial improvement in the software industry for all languages, since it allowed us to have more control over the code we are reusing from third-parties and to update these dependencies more easily. The problem is, we started to depend and trust **too much**.

I don’t think Ai will take your job

But the budget spent on it, the cap ex, that could make you redundant or get you paid less

As will the data already out there on what you use it for, what they can provide instead of paying you. Think on that.