Confidential Containers for GPU Compute: Incorporating LLMs in a Lift-and-Shift Strategy for AI