llama.cpp tokenization fix?
#1
by
int-19h
- opened
Is this quantized using a version of llama.cpp that contains https://github.com/ggerganov/llama.cpp/commit/ffe666572f98a686b17a2cd1dbf4c0a982e5ac0a?
Yes, I just uploaded the latest version
JaaackXD
changed discussion status to
closed