Fresh builds of llama.cpp with AMD ROCm™ 7 acceleration - View it on GitHub