Skip to main content

Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
TagsGame Engines

Version 1.03 should fix the double reply issue. As for laggy, assuming you are talking about how the game actually runs, yeah it seems unoptimized, though you can try staying near the edge of the apartment and looking away from the center to help, it might be best to wait for a fix.

As for Chat GPT usage limit, not much the developer can do about that, though it looks like there is a on host machine GPT model that is in the works that, with enough effort, could be hooked up to the game, though I'm not sure how hardware extensive running it is.

(1 edit)

GPT4All runs on a macbook air CPU. So easy a moldy potato can do it. Quality of responses would suffer somewhat. I'd be much more interested in GPT-J-6b or other open models that are used in KoboldAI, as they are much more open-ended and unbound.

Ah cool, wasn't sure how hard it was to run, I imagine it can't be too hard to program the game to take input from those instead of the Chat GPT API, but I could be wrong.

Though maybe an option for either would be good, since those on host models are trained on a smaller data set (At least I know it is for GPT4all).

It's really amazing how fast the tech is advancing, and I look forward to it being refined and used in games to handle most NPC interactions, and hopefully with enough programming to back it up so that the NPCs don't just say things they can't do.

i really hope they put it so it wont be in a limit anymore