felsiq

Can you please elaborate on that second one, or drop a name so I can look into it? Sounds very counterintuitive and like something I wanna know

Wow you’re more patient than I am, if you type on your phone here a lot lmao. Thanks for the answers!

You make a good point about systemd being monolithic, and I hate to add to your replies fully ignoring it to only talk about the thorn… but I gotta admit I’m really curious how you type it.

I’m guessing you’re not using text replacement and that you’re typing it instead, but do you have it bound to a key combo, replacing a little-used character, etc? Do you use the same method on mobile, if you also use the thorn there? If you type like this everywhere, are you concerned about your distinct typing patterns making you easy to dox?

Sorry to hit you with a bunch of questions unrelated to your actual comment, I don’t have strong opinions on systemd so don’t have much to contribute there lol

I’ve given up on running ASA using a 5900x and a 10gb 3080, cuz I can barely get a stable 40fps at 4K with clouds turned off and the settings as low as I can tolerate them (including upscaling turned on). If you’re playing at a lower resolution you’ll obviously have a much better time with it than I do, but fair warning this game is just unoptimized as fuck.

My performance is entirely gpu-bound, but that’s also what you’d expect at 4K so if you run 1080p for example a 5000-series cpu would probably give you a big jump in performance. A 5950x, 5900xt, 5900x, 5800xt, etc would be a really nice upgrade from your current cpu and with a bios update your board should support them all. If you’re able to find one, a 5800x3d or 5700x3d would be ideal if this is a purely gaming pc for you, though they’re not as powerful for productivity stuff.

The only other thing I’d check out is just to see if you have vsync enabled / a framerate cap, your gpu usage being low is probably just from the cpu bottleneck but those two things can also cause it

I never thought I’d say this, but in this one specific scenario I’m actually perfectly happy with him being alive too

To be clear, I wasn’t saying you’re wrong. I just like homomorphic encryption a lot and love a chance to tell people about it lmao

There are no LLMs that process encrypted tokens.

Check out homomorphic encryption! AFAIK it’s not used in any LLMs just yet but the plans are in place and it’s tantalizingly close

Hold up, tech noir… 2??? Thank you so much for introducing this to my life

So if I’m understanding you right, your crush is primarily on the person herself and/or who she wants to be and the character aspect of it is secondary? That’s kinda the opposite of what my ignorant guess would be so it’s really interesting to read. Thanks for answering!

Did you use a heavily quantized version? Those models are much smaller than the state of the art ones to begin with, and if you chop their weights from float16 to float2 or something it reduces their capabilities a lot more

Yep, the OpenAI api and/or the ollama one work for this no problem in most projects. You just give it the address and port you want to connect to, and that port can be localhost, lan, another server on another network, whatever.

If you don’t mind me asking, how does that work? I don’t know a lot about vtubers - is your crush on the character? On the vtuber themself? Do you know anything about the person behind the character or are they fully in character at all times?

No pressure to answer this if it’s too prying, tho!

Oof rip sorry lol, thought I’d seen a post about boost and piefed

I’ve seen a couple of other clients add support for piefed recently as well, although I don’t follow them closely cuz I’m happy with voyager. Maybe boost or mlem supports this with piefed already?

It’s come a long way recently, I’m currently using piefed on voyager rn and it works great

Worth clarifying from your proton link, it’s still behind an experimental feature flag but the support for Wayland in proton is coming along really well. I’ve just blanket enabled Wayland for all my games and with the latest proton-ge version they all work great.

It works in the sense that the operations are performed on binary numbers, so text handling works the same way it normally does assuming the handler function is encrypted to match. Once you have multiplication and addition, you can make logic gates and general computing follows from there - although with the noise being amplified thru each logic gate, the more complex the functions the more bootstrapping is required and the less I see this being doable in the short term.

For a working example, check out apple’s homomorphic encryption page, they use it for landmark identification and afaik will be using it for siri whenever they get to that update. It’s slow but it’s already usable - I’m not personally convinced it’ll be used everywhere, but the technology is super cool and I hope it shows up more

I don’t know if anyone reading this will ever have this problem (if you got this far without installing an adblocker, this is your wake up call - go get one now), but ctrl+W is the shortcut to kill a tab and that should work regardless site focus or popups

I need soft circular bears in my life

This really feels like it didn’t need an announcement