Files
ollama-for-amd/llama/ggml-cuda/template-instances/mmq-instance-q4_1.cu