Error while reproducing the paper result. #1
Description
Hi, I'm currently studying about digital water marking and came across your paper.
I found your paper very interesting and tried to reproduce the result you guys showed on the papaer.
However during training, after exactly 2000 iterations I'm keep getting an error.
The error message is shown below. I changed some path to ... due to privacy reasons
2%|█████▍ | 10000/400000 [2:02:36<79:41:38, 1.36it/s]
Traceback (most recent call last):
File "/home/.../.../CopyRNeRF-code/run_copyrnerf.py", line 1308, in
mp.spawn(train, args=(2,), nprocs=2, join=True)
File "/home/.../miniconda3/envs/TensoRF/lib/python3.9/site-packages/torch/multiprocessing/spawn.py", line 239, in spawn
return start_processes(fn, args, nprocs, join, daemon, start_method='spawn')
File "/home/.../miniconda3/envs/TensoRF/lib/python3.9/site-packages/torch/multiprocessing/spawn.py", line 197, in start_processes
while not context.join():
File "/home/.../miniconda3/envs/TensoRF/lib/python3.9/site-packages/torch/multiprocessing/spawn.py", line 160, in join
raise ProcessRaisedException(msg, error_index, failed_process.pid)
torch.multiprocessing.spawn.ProcessRaisedException:
-- Process 0 terminated with the following error:
Traceback (most recent call last):
File "/home/.../miniconda3/envs/TensoRF/lib/python3.9/site-packages/torch/multiprocessing/spawn.py", line 69, in _wrap
fn(i, *args)
File "/home/.../.../CopyRNeRF-code/run_copyrnerf.py", line 1269, in train
re = render_path(poses_test, message_test, hwf, K, args.chunk, args.patch_size,
File "/home/.../.../CopyRNeRF-code/run_copyrnerf.py", line 382, in render_path
imageio.imwrite(filename, rgb8)
File "/home/.../miniconda3/envs/TensoRF/lib/python3.9/site-packages/imageio/v2.py", line 397, in imwrite
return file.write(im, **kwargs)
File "/home/.../miniconda3/envs/TensoRF/lib/python3.9/site-packages/imageio/plugins/pillow.py", line 405, in write
raise ValueError("Can't write images with one color channel.")
ValueError: Can't write images with one color channel.
I didn't touch any code instead of changing the mp.spawn arguments to 2 because I'm running on two graphic cards.
Thank you for your time.