fix bug expand(torch.FloatTensor{[1, 1, 1, 33, 33]}, size=[1, -1, -1,…

[qkhanh711]

… -1]): the number of sizes provided (4) must be greater or equal to the number of dimensions in the tensor (5)

📝 Description

• Fix bug when I'm trying to run PatchCore and PaDiM
• 🛠️ Fixes [Bug]: patchcore pytorch error #1798

✨ Changes

Select what type of change your PR is:

• 🐞 Bug fix (non-breaking change which fixes an issue)
• 🔨 Refactor (non-breaking change which refactors the code base)
• 💥 Breaking change (fix or feature that would cause existing functionality to not work as expected)

✅ Checklist

Before you submit your pull request, please make sure you have completed the following steps:

• 📋 I have summarized my changes in the CHANGELOG and followed the guidelines for my type of change (skip for minor changes, documentation updates, and test enhancements).
• 📚 I have made the necessary updates to the documentation (if applicable).
• 🧪 I have written tests that support my changes and prove that my fix is effective or my feature works (if applicable).

For more information about code review checklists, see the Code Review Checklist.

[samet-akcay]

Hi @qkhanh711, can you provide a more descriptive title and description to the PR?

[qkhanh711]

Hi @qkhanh711, can you provide a more descriptive title and description to the PR?

It happened when I clone anomalib repository. After installing, I have ran the 201_fastflow.ipynb (anomalib/notebooks/200_models/)
and change the model from Fastflow to Patchcore
And I got this error:

`---------------------------------------------------------------------------
RuntimeError Traceback (most recent call last)
Cell In[7], line 1
----> 1 model = Patchcore(
2 coreset_sampling_ratio=0.001
3 )

File ~/Code/computer_vision/EvaluateAD/anomalib/src/anomalib/models/image/patchcore/lightning_model.py:51, in Patchcore.init(self, backbone, layers, pre_trained, coreset_sampling_ratio, num_neighbors)
41 def init(
42 self,
43 backbone: str = "wide_resnet50_2",
(...)
47 num_neighbors: int = 9,
48 ) -> None:
49 super().init()
---> 51 self.model: PatchcoreModel = PatchcoreModel(
52 backbone=backbone,
53 pre_trained=pre_trained,
54 layers=layers,
55 num_neighbors=num_neighbors,
56 )
57 self.coreset_sampling_ratio = coreset_sampling_ratio
58 self.embeddings: list[torch.Tensor] = []

File ~/Code/computer_vision/EvaluateAD/anomalib/src/anomalib/models/image/patchcore/torch_model.py:54, in PatchcoreModel.init(self, layers, backbone, pre_trained, num_neighbors)
...
---> 68 self.kernel = self.kernel.expand(self.channels, -1, -1, -1)
69 self.border_type = border_type
70 self.padding = padding

RuntimeError: expand(torch.FloatTensor{[1, 1, 1, 33, 33]}, size=[1, -1, -1, -1]): the number of sizes provided (4) must be greater or equal to the number of dimensions in the tensor (5)
Output is truncated. View as a scrollable element or open in a text editor. Adjust cell output settings...`

[ashwinvaidya17]

@qkhanh711 are you using the latest main branch? I am not able to reproduce the issue locally.
These are my relevant changes to 201_fastflow.ipynb

datamodule = MVTec(
    root=dataset_root,
    category="bottle",
    image_size=[256, 256],
    train_batch_size=32,
    eval_batch_size=32,
    num_workers=0,
    task=task,
)

model = Patchcore(backbone="wide_resnet50_2")

engine = Engine(
    pixel_metrics="AUROC",
    accelerator="auto",  # \<"cpu", "gpu", "tpu", "ipu", "hpu", "auto">,
    devices=1,
    logger=False,
)

[qkhanh711]

@qkhanh711 are you using the latest main branch? I am not able to reproduce the issue locally. These are my relevant changes to 201_fastflow.ipynb

datamodule = MVTec(
    root=dataset_root,
    category="bottle",
    image_size=[256, 256],
    train_batch_size=32,
    eval_batch_size=32,
    num_workers=0,
    task=task,
)

model = Patchcore(backbone="wide_resnet50_2")

engine = Engine(
    pixel_metrics="AUROC",
    accelerator="auto",  # \<"cpu", "gpu", "tpu", "ipu", "hpu", "auto">,
    devices=1,
    logger=False,
)

yes, I'm using the main branch of https://github.com/openvinotoolkit/anomalib.git and a met this.
anw, thank you for your answer.

ut with the original repository, I still encountered this bug

[windhorse777]

do you solve this problem? I enchanted the same bug when running the 000 geting__started

[qkhanh711]

do you solve this problem? I enchanted the same bug when running the 000 geting__started
in line 68 of anomalib/src/anomalib/models/components/filters/blur.py
turn self.kernel = self.kernel.expand(self.channels, -1, -1, -1) to self.kernel = self.kernel.expand(self.channels, -1, -1, -1, -1)

[samet-akcay]

#1944 addressed this issue. This is no longer needed. Thanks a lot for reporting and proposing a fix.