Generate multimodal forecasts with visual attention maps
Generate music from text and optional melody