rocm 6.2 Build for Qwen7B Inference – AMD RX 7900, Python 3.12, Safetensors, Multi-LLM Setup - View it on GitHub