Efficiency nerd. Compress your content please!
degeneracy knows no limits
“B” refer to the size of parameters of the model in the billion but people has started referring them to “bits”.
This is not entirely correct. B does stand for “Billion” parameters, but bits are a different thing. You can have, for example, a 4-bit 13B model or an 8-bit 3B model. They don’t correlate at all.
The player could describe what they saw, but nobody except the GM knows if it’s true.
It’s equally plausible that the paladin failed the check and saw a monster when there was only a dog, or that they passed the check and saw a monster because there was in fact a monster. Their argument to the party would be the same in either case: “that’s no dog, it’s a space station monster”.
The party then must question who saw the correct thing. Did the paladin actually see something everyone else missed? Or are they just seeing things? My point was that the players should not immediately be able to discern the truth. I find that this kind of uncertainty breeds intrigue!
I’ve never actually played, so this might be standard practice anyway, but I think this would be a great time to have the DM roll privately for each player and not tell them if they passed or failed. If the players only know what their character saw (and not if they pass or fail the check, or even get an idea based on the roll) then metagaming is impossible. This could produce a situation where it’s just a dog but the paladin thinks they saw a monster because they failed the roll, or it could be the other way around.
Doing it with DM-only rolls ensures the players have to actually figure out what they saw rather than knowing based on what they rolled or if they passed.
As I said, this could be standard practice, I have no idea. But I hope it is.
I don’t have any on me right now, but I could probably dig up some old ones later if I can remember to come back here.
If you can’t afford the hardware, perhaps try the AI Horde (it’s basically free access to crowdsourced LLMs). Your chats are likely to be ignored but there is nothing stopping whoever you’re connected to from logging your conversations. The only thing you can do to prevent that is to run one locally, which is actually getting easier and cheaper to do all the time. You can even run a (relatively small) model on a smartphone now!
The future is bright. Keep an eye on the progress and you may be able to join in sooner than you think.
You can already run LLMs on consumer hardware. They’re slow, but they definitely work. Mine is midrange from about 5 years ago and I can run a small one at a pretty reasonable speed or a medium (but “smarter”) one quite slowly at roughly 2 seconds per token.
I believe CloudFlare doesn’t upcharge any domains at all.
I was only joking :)
And I never really thought about running it like that, it’s interesting to think about.
I think you need to run the interface man :)
Giving “Feddit” a whole new meaning
If she’s unable to move after casting explosion, can she give consent? 🤔
Based foxgirl enjoyer
But what would the taste be? 🤔
I suppose that would explain why it’s hiding!
We kinda do already! !pregnanthentai@burggit.moe isn’t specifically for lolis but it definitely allows them, so I don’t see the point in a separate new community when this one already exists.
There’s a few other similar images by this artist. It might be worth adding them to this post!
I think most loli art involves some element of shock/confusion/correction/lust/ahegao etc but much more rarely cuter stuff like shyness, romance, and love. That’s what I feel makes the difference here, I wish there was more like this.
…what does that mean?
An absolutely moronic decision.