Clarify checkpoint training differences

Document bundled prompt-conditioned demos

Require prompt-conditioned VocalRender inference

Link CrawlSinger-OS and document model variants

docs: add audio demo link

Emphasize VocalRender architecture in model card

Update VocalRender model card and inference guide

Document independent variant subfolders

Document VocalRender model variants

Document VocalRender and VocalRender-Pro variants

Rename model to VocalRender-Pro

initial commit