DMtet fine-tune in remote service

[lingtengqiu]

Description

When I run command

python main.py -O --text "A baby bunny sitting on top of a stack of pancakes" --workspace trial_dmtet --dmtet --iters 5000 --init_ckpt trial/checkpoints/df.pth

to obtain the fine-tune DMtet mesh, I had the following problem.

[F glutil.cpp:338] eglInitialize() failed

Steps to Reproduce

""

Expected Behavior

""

Environment

PyTorch1.2.0, CUDA11.3

[ashawkey]

@lingtengqiu Hi, this is related to failure of OpenGL installation, you may try to reinstall the CUDA driver, or use nvidia-docker as suggested in #131.

[lingtengqiu]

Thanks

[zhongjinluo]

Replace all dr.RasterizeCudaContext(output_db=False) to dr.RasterizeCudaContext() in nerf/renderer.py as suggested in sicxu/Deep3DFaceRecon_pytorch#2 (comment)