Open-Sora-Plan
Open-Sora-Plan copied to clipboard
Why not reuse mixkit/pixabay/pexels in v1.1.0 for 1.2.0 training?
HI there, thanks a lot for the great work! I am wondering whether there is reason not to use the original (3000 hours) data collected in v1.1.0 training for 1.2.0 training? Does this mean Panda data is better than these three datasets? Thank you very much!