quantize.* missing

#10
by phymbert - opened

Hello,

Thanks for the quantum models.

I am interested to know why quantize.* imatrix metadata are not included:

https://github.com/ggml-org/llama.cpp/blob/master/examples%2Fquantize%2Fquantize.cpp#L61

phymbert changed discussion title from quantuze.* missing to quantize.* missing
Unsloth AI org

Only Q1 and Q2 are imatrix, the rest aren't

Still the metadata are not there in Q1, Q2 but present in the original quants from @bartowski . Strange.

phymbert changed discussion status to closed

Sign up or log in to comment