feifeifeiliu / probtalk

Towards Variable and Coordinated Holistic Co-Speech Motion Generation, CVPR 2024
https://feifeifeiliu.github.io/probtalk/
45 stars 1 forks source link

how to split the data? #8

Open Xiangyue-Zhang opened 1 month ago

Xiangyue-Zhang commented 1 month ago

I didn't see any tutorial on how to split the data based on your '''/data_utils/split/*.npy''' files. Do I need to copy the data processing step from the repositories 'talkshow'?