Describe your dataset goal. We'll filter the annotation schema, fetch media, and create a fully annotated multimodal JSONL dataset.
Supports: Internet Archive, direct links, podcasts, Vimeo, and most platforms.