[Feature]: MaskGCT long-form audio

[fakerybakery]

Hi,
Thanks for releasing MaskGCT! Are there any plans to support long-form speech synthesis besides using chunking?
Thanks!

[HeCheng0625]

Hi, thank you for your attention. In the future, we will expand the training data to the minute level and use a codec with a higher compression rate to generate longer audio.

[JonathanFly]

Hi, thank you for your attention. In the future, we will expand the training data to the minute level and use a codec with a higher compression rate to generate longer audio.

In the current version, what was the duration of the training data "chunk"?