π“Ÿ host1.bitchan.one

𓀐 @LknCME?7

⏱ 1w seenβ›“ 0πŸ‘ 0β§– 1

filter [-]

languages:onother

This might be old new, but I recently found out about Strata and decided to give it a try https://github.com/Niko1221/Strata my setup is 64gb ram with a 3090 Results are impressive, a 2-3x boost compared to llama.cpp, even with llama's expert cache enabled⏱ 1w↩