Please add RIFLEx as a community contribution.

[gracezhao1997]

We are grateful for the open-sourcing of such an excellent model. We extended the video with just one line of code: freq[k-1]=(2np.pi)/(Ls). We found that HunyuanVideo performs exceptionally well in video length extrapolation, achieving high-quality videos at 2x temporal extrapolation in a completely training-free manner. By fine-tuning for only 1,000 steps with a batch size of 8 on original-length videos, both dynamic quality and visual quality are further improved. See demos on our project page: https://riflex-video.github.io/. The code is available at: https://github.com/thu-ml/RIFLEx. Thank you very much!

[kathrinawu]

Thanks for your contribution! Awesome work👍 has added in community contribution.