I tried this model using q8 in llama.cpp and it's in reasoning mode forever just to reply to my chat "hi". the reasoning not yet end even after i finish writing this discussion post lol. is there any setting or fix i could try to prevent this happening?
