What was the command/method used to make these quants?
#1
by Downtown-Case - opened
Just one, as an example.
I'm trying to quantize models myself in sd.cpp, but the filesize ends up different. To you specify certain tensors to be F16 or something?
Hi,
Sorry, only to reply now, but I use my script to get city96 scripts to create this - https://github.com/Santodan/GGUF-Converter-GUI