Why is UD-IQ3-XSS slower than UD-Q4K and even UD-Q5K? (same prompt/chat, same offloading)
#8 opened 6 days ago
by
tnuvkeg
Hot Damn This Model Cooks!
π
6
8
#5 opened 21 days ago
by
aaron-newsome
Does it make sense to have UD-IQ4_XS?
2
#4 opened 22 days ago
by
tarruda
Report: getting 20 t/s with UD-Q4_K_XL and 72 VRAM
π₯
1
10
#2 opened 22 days ago
by
SlavikF