Fernando Pérez-García

Thanks. I think that it's detrimental for you and for PyTorch that the provided example is not representative of the original model's performance. Could you update the example with a...

> The problem is that it would require loading the entire 4D MRI volume in the code sample which is run as a test every time there is a PR...

For reference, here's a Colab notebook in which I manage to run the segmentation (using TorchIO): https://colab.research.google.com/drive/1XprcAZ59dAYBGFjawgb7h9mx0Y9N6pTj?usp=sharing I assumed a spacing of 1 x 1 x 9 mm and LPS...

By the way, the image I used, assuming 32 bits per voxel, takes 18 MB in memory. The frequency of PRs on pytorch/hub doesn't seem very high, so I think...

Yes, you could potentially use TorchIO for that visualization. But it would probably be easier (or more powerful) to use SimpleITK and then matplotlib.

At each iteration, [`batch_size` windows are extracted from the queue](https://niftynet.readthedocs.io/en/dev/config_spec.html#queue-length), they're fed to the network and the weights are updated using the gradient of the loss function with respect to...

Maybe your window size is too small. Check my tutorial: https://gist.github.com/fepegar/1fb865494cb44ac043c3189ec415d411 Also, you can start by trying to fit only one image instead of the whole dataset? https://twitter.com/karpathy/status/1013244313327681536

What is [`interp_order`](https://niftynet.readthedocs.io/en/dev/config_spec.html#interp-order)? The docs say "Interpolation order of the input data", I think that's not clear. When I searched `sitkNearest` in the repo, the only occurrence is in `RandomElasticDeformationLayer`....

What is the purpose of this issue?

Thanks for adding some info :+1: