Failed to load weights
Hi @aloisio,
Thanks for sharing the code and weights. I'm trying to run the evaluation in NYU, and I can't seem to load the weights to preprocess the dataset.
$ python spawn_preproc.py NYU R333_MMNet_SUPERV-NYU-depth+rgb+normals_da
Loading weights weights/R333_MMNet_SUPERV-NYU-depth+rgb+normals_da
Traceback (most recent call last):
File "spawn_preproc.py", line 303, in <module>
main()
File "spawn_preproc.py", line 299, in main
preproc()
File "spawn_preproc.py", line 194, in preproc
model.load_state_dict(torch.load(wpath))
File "/home/vscode/miniconda3/envs/spawn-torch/lib/python3.8/site-packages/torch/nn/modules/module.py", line 2152, in load_state_dict
raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
RuntimeError: Error(s) in loading state_dict for BiModalRDFNetLW:
Missing key(s) in state_dict: "encoder1.conv1.weight", "encoder1.bn1.weight", "encoder1.bn1.bias", "encoder1.bn1.running_mean", "encoder1.bn1.running_var", "encoder1.layer1.0.conv1.weight", "encoder1.layer1.0.bn1.weight", "encoder1.layer1.0.bn1.bias", "encoder1.layer1.0.bn1.running_mean", "encoder1.layer1.0.bn1.running_var", "encoder1.layer1.0.conv2.weight", "encoder1.layer1.0.bn2.weight", "encoder1.layer1.0.bn2.bias", "encoder1.layer1.0.bn2.running_mean", "encoder1.layer1.0.bn2.running_var", "encoder1.layer1.0.conv3.weight",
...
Note: I had to rename the folder from weigths to weights, since there was a typo. But should it work after that change?
Thanks, Raul.