NVlabs / nvdiffrec

Official code for the CVPR 2022 (oral) paper "Extracting Triangular 3D Models, Materials, and Lighting From Images".
Other
2.13k stars 223 forks source link

can u please make that training visualation also works on linux #28

Closed Manni1000 closed 2 years ago

Manni1000 commented 2 years ago

Visualize training progress (only supported on Windows)

jmunkberg commented 2 years ago

As a WAR, please note that you can increase the frequency of saved training images, by increasing the frequency in the config, e.g., "save_interval": 10, to dump an image every 10 iterations.

We typically run headless in docker on Linux servers, and deliberately removed the OpenGL window to minimize dependencies.

It should be straightforward to enable it on your local workstation if needed. The display code is here: https://github.com/NVlabs/nvdiffrec/blob/main/render/util.py#L378