Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Added cflow algorithm #47

Merged
merged 24 commits into from
Jan 9, 2022
Merged

Added cflow algorithm #47

merged 24 commits into from
Jan 9, 2022

Conversation

blakshma
Copy link
Contributor

No description provided.

@samet-akcay
Copy link
Contributor

@blakshma pipeline fails because of missing einops module. I guess you need to add it to requirements/base.txt. You probably need to add freia as well

Copy link
Collaborator

@ashwinvaidya17 ashwinvaidya17 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Minor comment but rest looks good.

anomalib/models/cflow/backbone.py Outdated Show resolved Hide resolved
@blakshma
Copy link
Contributor Author

blakshma commented Dec 25, 2021

Thank you for updating requirements @samet-akcay

@blakshma
Copy link
Contributor Author

@samet-akcay the pipeline fails while testing load model for cflow. Is this something you have encountered before? Any leads?

@samet-akcay
Copy link
Contributor

samet-akcay commented Dec 26, 2021

@samet-akcay the pipeline fails while testing load model for cflow. Is this something you have encountered before? Any leads?

@blakshma , you could look at the details here, which shows why it failed. It's mainly becuase of a CUDA memory issue while running C-Flow.

@blakshma
Copy link
Contributor Author

blakshma commented Jan 6, 2022

All tests are passed in both of my local system. Running into OOM error with pre-merge check.
image

@samet-akcay
Copy link
Contributor

It's quite likely that this is due to hardware differences between your machine and the mlops machine on which we run the CI. We'll investigate whether evaluation the metrics on cpu would remedy the issue.

@@ -0,0 +1,3 @@
# Real-Time Unsupervised Anomaly Detection via Conditional Normalizing Flows

This is the implementation of the [CW-AD](https://arxiv.org/pdf/2107.12571v1.pdf) paper.
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do you have sample results? It would be nice to include them here for consistency with other tasks. Here is an example https://github.com/openvinotoolkit/anomalib/blob/development/anomalib/models/stfpm/README.md
Otherwise, can you create a ticket for it to address this in a different PR.

@samet-akcay samet-akcay self-requested a review January 9, 2022 20:30
@samet-akcay samet-akcay merged commit 80a3de2 into development Jan 9, 2022
@samet-akcay samet-akcay deleted the algo/barath/cflow branch January 9, 2022 20:31
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants