I've got one. I'm using it to power an LLM for a chatbot me and a few friends use. The memory and disk speeds are really slow, but there's plenty of memory to

I've got one. I'm using it to power an LLM for a chatbot me and a few friends use. The memory and disk speeds are really slow, but there's plenty of memory to keep whatever you need loaded. I can mmap massive models and get decent inferencing performance without spending a dime. One of these days I feel like we'll draw too much attention to the program, and they'll have to shut it down. Until then...