[Style] Voice Assistant style, Podcast style, Audiobook style, and Online Learning style (four styles in total). [Duration] Approximately 2 hours per speaker, with about 0.5 hours for each style. [Content] Voice assistant data is recorded as single utterances, while the other styles are recorded in paragraph-level segments.