DIfferent results here in the chat and locally
#10
by
KarBik
- opened
Hello,
I was testing some prompts here and they worked very well. But when I was testing locally the results for the same prompts were absolutely different and very bad. Is there something specific you do during the inference?
Thanks in advance
are you using the chat formatting? what generation settings are you using?
I have the same issue. I also tried chat formatting.