It’s true lol. I had to install Docker for teaching on my old drive and that instantly maxed out my root partition even when I kept deleting intermediate builds and unused data. Now I have this fun paranoia for all apps :)

just fyi I moved Discord, GIMP, Obsidian, and OBS over to flatpak and my root partition jumped from 19GB to 23GB. I’m kinda sad about it tbh

I was there on TeamSpeak over a decade ago. It was good for its time

It looks like you’re right. Uninstalling the deb and then installing the flatpak consumed an additional 2GB on my root, but I have a handful of other electron-based apps that are mildly obnoxious snaps, and migrating them might help amortize that cost

Oh that’s great! I’ll switch to the flatpak tonight then

I haven’t tried flatpacks yet. How do they compare to, say, snaps in terms of storage/redundancy?

username checks out

It’s a way to force an 18 year old into a life of indentured servitude under the guise of “financial assistance” by simply clicking accept on a couple online forms, only for 40% of them to end up working jobs that don’t require a college degree in the first place.

Definitely RE for me. I couldn’t sleep after the first time I saw a crimson head. The sharks were terrifying too

wttr gang

I believe your “They use attention mechanisms to figure out which parts of the text are important” is just a restatement of my “break it into contextual chunks”, no?

Large language models literally do subspace projections on text to break it into contextual chunks, and then memorize the chunks. That’s how they’re defined.

Source: the paper that defined the transformer architecture and formulas for large language models, which has been cited in academic sources 85,000 times alone https://arxiv.org/abs/1706.03762