joeylitalien / noise2noise-pytorch

PyTorch Implementation of Noise2Noise (Lehtinen et al., 2018)
MIT License
297 stars 59 forks source link

OpenEXR Error while running the train files. #5

Open sneha-gathani opened 5 years ago

sneha-gathani commented 5 years ago

Traceback (most recent call last): File "train.py", line 7, in from datasets import load_dataset File "/Users/snehagathani/Desktop/noise/src/datasets.py", line 9, in from utils import load_hdr_as_tensor File "/Users/snehagathani/Desktop/noise/src/utils.py", line 12, in import OpenEXR ImportError: dlopen(/anaconda3/lib/python3.6/site-packages/OpenEXR.cpython-36m-darwin.so, 2): Symbol not found: __ZN7Imf_2_314TypedAttributeISsE13readValueFromERNS_7IStreamEii Referenced from: /anaconda3/lib/python3.6/site-packages/OpenEXR.cpython-36m-darwin.so Expected in: flat namespace in /anaconda3/lib/python3.6/site-packages/OpenEXR.cpython-36m-darwin.so

This error comes for all the training files.

joeylitalien commented 5 years ago

Hi Sneha,

The OpenEXR module is only used for the experimental Monte Carlo denoising, you can safely comment the part of the code that deals with it. Or you can try re-installing OpenEXR. A quick google at your problem revealed that you are not the first with this issue but I am unsure how to solve it on Windows. In any case, it is not essential to the default N2N code.

Cheers, J.