Multimodal Generation

Hugging Face Transformers task for any-to-any multimodal generation; it supports combinations of text, image, audio, and video inputs and can generate outputs across those modalities.