llama.cpp

save number of parameters and the size in llama_model, fixes #10285

#10286

Merged