Yeah. I've been considering buying a Ryzen 395 Max+ so I can use those big models locally. I only have 12GB of vram right now. (3080 Ti).
Yeah. I've been considering buying a Ryzen 395 Max+ so I can use those big models locally. I only have 12GB of vram right now. (3080 Ti).
I would recommend trying Hugging Faces Spaces to trial the model you're interested in, see if it justifies the cost for your use case. I've found that anything above 70B has diminishing returns when compared to fine tuning and specialization. Leave the big general models to corps if unnecessary.
What I hate so much is the big model censorship. Claude Sonnet is absolutely incredible at roleplay and narrative generation in dynamic RPG games. It rarely outright refuses to perform, but it is hard to risk an account ban if they somehow think your story about a knight murdering someone is real.
Yep! The censorship is my biggest issue. I wasn't planning on running anything larger than 70B. Mostly the intention with the Ryzen AI 395 is to run a couple of models at once. Been thinking about creating a mixture of experts set up for Home Assistant, but I haven't had time to play with it yet.