Final user interface to work with recommendation model
Generate videos from text prompts with optional images