MODUS: Decoder-Only Any-to-Any Modeling of Diverse Modalities

Part of

Deep Learning Model Evaluation Benchmarks