LLM at home

I have been playing a little bit with LLMs at home, following this guide from the Register. That was simple and running slow. I do have some other scenarios where a suitable processor can add to the experience such as with Immich which can do face recognition and image classification.

Linkdump

For lack of a better title, for now, Linkdump Things I read, things that make me think, things that make an impression. Things I would like to come back to at some point Models All The Way Down A Knowing Machines project/article about the open training set LAION-5B, how it is constructed etc

Linkdump

For lack of a better title, for now, Linkdump Things I read, things that make me think, things that make an impression. Things I would like to come back to at some point To See Black Holes in Detail, She Uses ‘Echoes’ Like a Bat | Quanta Magazine Gary explains: What is x86-64-v3?

Linkdump

For lack of a better title, for now, Linkdump Things I read, things that make me think, things that make an impression. Things I would like to come back to at some point Bruce Schneier - AI related AI and Lossy Bottlenecks - Schneier on Security AI and Trust | Belfer Center for Science and International Affairs The part about trust is very interesting

RSS - websites - rabbit hole

I follow several blogs via RSS and today I read A splash of color by Matt Birchler over at Birchtree. He points at a few websites and Pedro Duarte personal website made me go wow. Checking other websites he referred to lead me directly or inderectly to Chris Nicholas, Benji Taylor, Rauno Freiberg, Lee Robinson all looking at design and content creation from various angles.