Prediction does not run with torch 1.3

[iparask]

I am trying to execute iceberg_seals.predicting with the following stack:

(SealsPytorch1) [paraskev@gpu045 unit.000066]$ python
Python 3.5.2 (default, Dec 16 2016, 08:52:16)
[GCC 4.8.5 20150623 (Red Hat 4.8.5-4)] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
>>> import torchvision
>>> import tensorboardX
>>> torch.__version__
'1.3.0'
>>> torchvision.__version__
'0.4.1'
>>> tensorboardX.__version__
'1.9'
>>>

and I am getting the following error:

(SealsPytorch1) [paraskev@gpu045 unit.000066]$ iceberg_seals.predicting "--input_image" "WV03_20141120084907_1040010004AD7600_14NOV20084907-P1BS-500268558060_01_P001_u08rf3031.tif" "--model_architecture" "UnetCntWRN" "--hyperparameter_set" "A" "--training_set" "test_vanilla" "--test_folder" "/pylon5/mc3bggp/paraskev/SealsTest/test_image" "--model_path" "./" "--output_folder" "./WV03_20141120084907_1040010004AD7600_14NOV20084907-P1BS-500268558060_01_P001_u08rf3031"
Traceback (most recent call last):
  File "/pylon5/mc3bggp/paraskev/SealsPytorch1/bin/iceberg_seals.predicting", line 87, in <module>
    ['num_workers_train'])
  File "/pylon5/mc3bggp/paraskev/SealsPytorch1/lib/python3.5/site-packages/iceberg_seals/search/predicting/predict_sealnet.py", line 140, in predict_patch
    for data in dataloader:
  File "/pylon5/mc3bggp/paraskev/SealsPytorch1/lib/python3.5/site-packages/torch/utils/data/dataloader.py", line 819, in __next__
    return self._process_data(data)
  File "/pylon5/mc3bggp/paraskev/SealsPytorch1/lib/python3.5/site-packages/torch/utils/data/dataloader.py", line 846, in _process_data
    data.reraise()
  File "/pylon5/mc3bggp/paraskev/SealsPytorch1/lib/python3.5/site-packages/torch/_utils.py", line 385, in reraise
    raise self.exc_type(msg)
RuntimeError: Caught RuntimeError in DataLoader worker process 0.
Original Traceback (most recent call last):
  File "/pylon5/mc3bggp/paraskev/SealsPytorch1/lib/python3.5/site-packages/torch/utils/data/_utils/worker.py", line 178, in _worker_loop
    data = fetcher.fetch(index)
  File "/pylon5/mc3bggp/paraskev/SealsPytorch1/lib/python3.5/site-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "/pylon5/mc3bggp/paraskev/SealsPytorch1/lib/python3.5/site-packages/torch/utils/data/_utils/fetch.py", line 44, in <listcomp>
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "/pylon5/mc3bggp/paraskev/SealsPytorch1/lib/python3.5/site-packages/iceberg_seals/search/utils/dataloaders/data_loader_test.py", line 95, in __getitem__
    sample = self.transform(sample)
  File "/pylon5/mc3bggp/paraskev/SealsPytorch1/lib/python3.5/site-packages/torchvision/transforms/transforms.py", line 61, in __call__
    img = t(img)
  File "/pylon5/mc3bggp/paraskev/SealsPytorch1/lib/python3.5/site-packages/torchvision/transforms/transforms.py", line 166, in __call__
    return F.normalize(tensor, self.mean, self.std, self.inplace)
  File "/pylon5/mc3bggp/paraskev/SealsPytorch1/lib/python3.5/site-packages/torchvision/transforms/functional.py", line 217, in normalize
    tensor.sub_(mean[:, None, None]).div_(std[:, None, None])
RuntimeError: output with shape [1, 224, 224] doesn't match the broadcast shape [3, 224, 224]

(SealsPytorch1) [paraskev@gpu045 unit.000066]$

[bspitzbart]

Is this related to issue #65?

[iparask]

I am not sure. I think not, because it has to do with the dimensions of the output matrix.