Q3_K_M vs UD-Q3_K_XL

#5
by DrRos - opened

Thanks for quants, little question - I thought UD quants always bigger than their _K_M counterparts, why GLM's Q3_K_M is 360 GB and UD-Q3_K_XL is 332 GB - which one to choose if I can fit either on my hardware? I know the rule of thumb to get biggest, but this time I doubt which to choose.

Unsloth AI org

UD is actually 50% sometimes smaller and 50% sometimes bigger. Usually bigger is always better

@shimmyshimmer thanks!

DrRos changed discussion status to closed

Sign up or log in to comment