very verbose Pytorch warnings

[sammlapp]

Lines 172 and 287 of transforms.py produce thousands and thousands of warning messages during training and inference

birdset/datamodule/components/transforms.py:172: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).