Do you think we are near the theoretical maximums of how much model we cram into 20B?
I don't consider myself qualified to make that call, though I believe there's plenty of room for improvement. But I don't mean inference compute, which is fairly accessible for well-off enthusiasts even now. Training is a lot more expensive though.
alt Hacker News
I don't consider myself qualified to make that call, though I believe there's plenty of room for improvement. But I don't mean inference compute, which is fairly accessible for well-off enthusiasts even now. Training is a lot more expensive though.