Show HN: Open-source AMDGCN kernels for optimizing LLM inference

This project offers open-source AMDGCN kernels that accelerate large language model inference, reducing execution time on AMD GPUs. To adopt, first verify compa