[p_l]

Generally, unquantized - double the number and that's the amount of VRAM in GB you need + some extra, because most models use fp16 weights so it's 2 bytes per parameter -> 32B parameters = 64GB

typical quantization to 4bit will cut 32B model into 16GB of weights plus some of the runtime data, which makes it possibly usable (if slow) on 16GB GPU. You can sometimes viably use smaller quantizations, which will reduce memory use even more.

[regularfry]

You always want a bit of headroom for context. It's a problem I keep bumping into with 32B models on a 24GB card: the decent quants fit, but the context you have available on the card isn't quite as much as I'd like.