llama.cpp

gguf-py : fix SpecialVocab parsing when post_processor is null

#14330

Merged