Can I ask what you use to generate stuff? I haven’t dipped my toes in any LLM stuff yet and I prooobably should just for fun.
Running it at home with https://ollama.com/ and I think I used mistral7b model that released a few days ago.
I'm running on a 1080 ti. It's only a few seconds with the most recent small models. I did fall off on image generation though.
Oh shit that awesome! Thank you, I’ll give it a shot! I’m spoiled by watching Kitboga’s Western Amber running on a 4090.