llama.cpp

requirements: relax torch~=2.6.0 to torch>=2.6.0 for convert_hf_to_gguf

#23503

Merged