Why is translation set to 0?

[Tenvence]

Thanks for your great work.

In line 181 of the inference.py file, the following code concatenates the rotation and translation features. However, the translation feature is explicitly set to zero by multiplying it with 0:

driven_pose_feat_embed = torch.cat([driven_pose_feat['rotation'], driven_pose_feat['translation'] * 0], dim=-1)

I am concerned about why the translation component is effectively ignored in this embedding.