Johannes Leuschner

Configure GPU index for 'astra_cuda', select GPU currently used by PyTorch in OperatorModule

5

This pull request implements two new features: - Choice of GPU for 'astra_cuda' via settable property ``'gpu_index'`` in ``RayTransformBase`` - Automatically select GPU currently used by PyTorch in ``OperatorModule`` by...

torch.nn.DataParallel with OperatorModule wrapping 'astra_cuda'-RayTransform malfunctioning

6

There seem to be problems when using the `'astra_cuda'` backend wrapped by a `RayTransform` operator again wrapped by `odl.contrib.torch.operator.OperatorModule` when trying to distribute the model on multiple GPUs using `torch.nn.DataParallel`....

The iterative reconstructor in `from dival.reconstructors.odl_reconstructors` inherit from `dival.reconstructors.reconstructor.IterativeReconstructor`, and thus should reflect the hyper parameter `iterations`. Instead they currently use an unrelated attribute `niter`. This should be merged, ideally...