Oliver Simons

Thanks for moving the issue to the correct repo ! > which is is why you cannot refer to objects constructed inside the `axis` before `\end{axis}`. I am confused by...

Hi, we also tested around with anomaly segmentation (provided e.g. by Autoencoder), and need the masks for evaluating these methods (results didn't make it into the paper and padim beat...

I took the time to provide a brief [MWE](https://github.com/ftilmann/latexdiff/files/7978123/MWE_latexdiff_booktabs.zip). Calling `latexdiff old.tex new.tex > diff.tex` results in a broken `diff.tex` with `\DIFaddbeginFL \cmidrule\DIFaddFL{(lr)}{\DIFaddFL{1-3}}` being the cause (should be `\cmidrule(lr)\DIFaddFL{1-3}`). Version...

What is your preamble ? And what was your latexdiff version ? I could reproduce the error with ``` $ /usr/bin/latexdiff --version This is LATEXDIFF 1.2.1 (Algorithm::Diff 1.15 fast, Perl...

I would also love to see something like this !

Possibly related due to `scatter` flag inconsistencies: #480 #479

As state in the error, ONNX does not yet support `cdist`, see https://github.com/onnx/onnx/issues/2442

Just a quick update from my side: support for `torch.cdist` was added in PyTorch 1.12 by partitioning/refactorign `cdist` via supported ops in onnx (see [this PR](https://github.com/pytorch/pytorch/pull/75278), so export to onnx...

I just overrode those requirements manually, and this seemed to work out locally (at least for my testings with EfficientNets + onnx export of patchcore). So you could always do...

Since PrecisionRecallCurve works on all data/is computed on the complete data by default in torchmetrics, you will have to hoid all your datapoints in memory. For large datasets, you can...