/lmg/ - Local Models General - AI development discussion

  • 🐕 I am attempting to get the site runnning as fast as possible. If you are experiencing slow page load times, please report it.
Oh, you guys are starting up again here. Would be funny if some desperate coomer tries to start up /aicg/ here of all places but I'm sure they can maintain post-nut clarity for long enough not to do so.
 
  • Thunk-Provoking
Reactions: Lieutenant Rasczak
Not really any faster that i can see, genning a 30s clip right now. However i think vramlets will leap on it due to the very low requirements. It's taken 30 mins to do 10s so far. No doubt if there's some gains hidden in there they will be extracted for use elsewhere.
Turbo sloppa inc etc.
 
A thread for this subject already exists in the shape of the ChatGPT one, it evolved over time into a more general LLM discussion board. There was talk of making a new LLM general thread, but no one bothered until now.
 
Is this related to the ads I get for milfs and models in my local area who are rather insistent on me personally fucking their frankly mid looking asses?
 
There was talk of making a new LLM general thread, but no one bothered until now.
I was writing some unstructured text spoonfeeding the general gist of LLMs and setting up frontends etc. but it's mostly about using APIs and SaaS, I've only ever used local once through kobold.
 
i just want novelai at home, something that gives me control of the story when the ai goes off rails and doesnt write the entire damn thing for me and read like complete shite
 
Anybody recommend any model in particular for generating mostly boilerplate code? Don't worry - it's not for production. Languages being C#, Typescript/Javascript and maybe Python or PHP. But more intended as a general question.

Needing something that runs locally.
 
Back