point_cloud_classification.ipynb: TypeError: initialize() got an unexpected keyword argument 'device'

[msieb1]

In the example notebook, the dataset class does not accept a device argument. How does one cast the loader to cuda in this example?:

val_loader = DataLoader(ModelNet(modelnet_path, categories=categories,
                                 split='test',transform=transform, device=device),
                        batch_size=12)

yields

 in __init__(self, preprocessing_transform, preprocessing_params, transform, no_progress, *args, **kwargs)
     89             no_progress (bool): disable tqdm progress bar for preprocessing.
     90         """
---> 91         self.initialize(*args, **kwargs)
     92         if preprocessing_transform is not None:
     93             desc = 'Applying preprocessing'

TypeError: initialize() got an unexpected keyword argument 'device'

If run without device arg:

engine = ClassificationEngine(PointNet(num_classes=len(categories)),
                              train_loader, val_loader, device=device)
engine.fit()

yields

RuntimeError: Input type (torch.FloatTensor) and weight type (torch.cuda.FloatTensor) should be the same

[msieb1]

What works is adding a cuda cast to the data via transforms and label via _get_attributes works (albeit hacky).

[Jean-Francois-Lafleche]

Thank you @msieb1 for helping us make Kaolin better! If you still had trouble getting started with that example, PR #264 fixes the issues you found.

[Caenorst]

Closing as this is related to v0.1