Alexander Riedel

Is this feature supposed to be implemented by github contributors? Are you going to open-source your model, training and dataset for this purpose? Your project sounds a bit more open...

You can set `model.image_threshold.value = ... model.pixel_threshold.value = ... ` before inference Please be aware that these are unnormalized values

Can you show me your testing / inferencing script? I'm pretty sure, calling `trainer.predict` does not affect the model. As there is not `predict` method implemented in the models in...

I cannot reproduce the issue with a Patchcore Model ``` model.adaptive_threshold = False model.pixel_threshold.value = torch.tensor(float(1)) model.image_threshold.value = torch.tensor(float(1)) print(model.pixel_threshold.value, model.image_threshold.value) trainer.predict(model=model, dataloaders=datamodule) print(model.pixel_threshold.value, model.image_threshold.value) ``` ``` tensor(1.) tensor(1.) LOCAL_RANK:...

My suggestion to solving this issue is to save the (validation) preprocessing pipeline to the model so it can be loaded from the model file for inferencing. This can be...

PaDim isn't "trained" but is extracting image features at training time that are stored. If the features of every dataset image have been retrieved, at test time the test set...

The error is probably GPU OOM in both cases and there's not too much you can do about it besides increasing your GPU VRAM or reducing the training set size....

This is because you do not update any weights during the "training" of patchcore but only save features while inferencing your training data (https://arxiv.org/pdf/2106.08265.pdf). Doing this for multiple epochs will...

You could replace the `torch.cdist` by a simple implementation of the euclidian distance like this one: https://discuss.pytorch.org/t/understanding-cdist-function/76296/13 ``` def my_cdist(x1, x2): x1_norm = x1.pow(2).sum(dim=-1, keepdim=True) x2_norm = x2.pow(2).sum(dim=-1, keepdim=True) res...

You basically have to factor your model code like one of the existing models https://github.com/openvinotoolkit/anomalib/tree/development/anomalib/models then you have to add tests and documentation and create a pull request :)