llama.cpp

60e9cf7a - batch: migrate the rest of examples to llama_batch_ext (#29601)