For anyone looking to experiment with these models who doesn't have 210GB of VRAM on tap-we're working as quickly as we can to get cheap access to 4x80GB A100 instances running at thundercompute.com (aiming for sub-$5/hr). For quantized versions, we have cheaper 1-2 GPU nodes available today. If you're interested, join our Discord for updates: https://discord.com/invite/nwuETS9jJK