llama.cpp

3979f2bb - docs: add hunyuan-ocr gguf, also add test [no ci] (#21490)