Code stuck while pretraining

[rishavbb]

Hi,

Thanks for the amazing work. I am trying to use your repo for a last few days. I have made sure I am using Ubuntu 18.04, CUDA 11.3 and PyTorch 1.12.1 and install all the necessary libraries. I have done the preprocessing part of the video. However, whenever I am running this command for pretraining bash scripts/pretrain_con.sh data/pretrain output/<project_name> <GPU_ID>
It is exactly freezing at this line 177 of pretrain_mouth.py
render_pkg = render_motion_mouth_con(viewpoint_cam, gaussians, motion_net, scene.gaussians_2, motion_net_face, pipe, background, personalized=True)

Initially I thought it is because of insufficient GPU size, but then I switched to a system having a GPU size of 40 GBs. I tried running it multiple times and it gets stuck at exactly the above mentioned line.

[Fictionarry]

Hi, actually I have not met such a problem yet. GPU memory does not seem to be a problem, as its typical occupancy should be about 2GB. What if just run the inference with the provided weights? I guess it may be caused by some dependency installation problems, especially as it occurs at the rendering step that should not involve any other parts but the CUDA rasterizer built previously.