I can't answer for the OP but it works fine under llama.cpp's container.