I’ve seen people use Jan for this kind of stuff but as far as I know it can’t do image generation or voiced responses. I’m kind of looking for something along the lines of moemate that I can run on my computer and not have to pay a subscription for.
Tell me more about your use case.
Porn, the answers always porn.
I’m not sure what else you’re looking for. I mentioned dirty talking in the post. Would just be nice to see some titties as well and not have to pay a subscription for it.
If I could import characters or situations that would be even better.
All the pieces are there, uncensored gen AI models exist for both text and images. But it’s not clear to me how you want these things to connect or how you want to interact with the framework. I’m guessing there is some existing platform that you’re trying to replicate locally? But without any context it’s difficult to help
If you are expecting you to just click go and have 10 pages of erotic stories then no it does not exist as far as I can see. Ollama is a tool to just download and run a llm on a computer and most models are designed to tell you to go away if you ask for anything like that. Some have been uncensored and do not have guardrails but that is not your problem.
A large language model is really advanced auto correct. We have taken all text we can, done some fancy statistics to work out what words are used neer each other and found what the average most common next word would be. Keep rolling the dice and work out what the next most average word is and repeat. A llm does not know what a word is and is based on converting combinations of words to a number and some fancy statistics to work out how common “th” and "e " are so it can spit out "the " randomly when it thinks it is statistically needed.
Llms cannot keep track of context and the more you use them thr more likely they will forget character names etc
