[GaggiX]

At 4-bit quantization it should already fit quite nicely.

[Aurornis]

Unfortunately not with a reasonable context length.

[regularfry]

I've got 139k context with the UD-Q4_K_XL on a 4090, q8_0 ctk/v. Could probably squeeze a little more but that's enough for me for the moment.

[corysama]

Hey, buddy! Can I bum a command line arg list off ya?

[GaggiX]

The model uses Gated DeltaNet and Gated Attention so the memory usage of the KV cache is very low, even at BF16 precision.

[kkzz99]

It really depends on what you think a reasonable context length is, but I can get 50k-60k on a 4090.