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.

Sign up for GitHub

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

Jump to bottom

accuracy UNET #6

Open
luantunez opened this issue Apr 15, 2021 · 1 comment
Open

accuracy UNET #6

luantunez opened this issue Apr 15, 2021 · 1 comment

Comments

@luantunez
Copy link

luantunez commented Apr 15, 2021 •
edited
Loading

Hello and thank you for sharing your work.
I am having trouble since the metrics of my BCNN with custom 2D images are not improving at all.
Can you think of a reason?
I am using
Tensorflow 2.4.1
Tensorflow-probability 0.12.1

I have tried several models all across the web, and I am having the same problem. The data shouldn´t be the problem, since a regular CNN trains perfectly.
@tmlabonte
Copy link
Contributor

tmlabonte commented Apr 27, 2021 •
edited
Loading

Performance is not guaranteed in TensorFlow 2, since a lot changed from TensorFlow 1 to 2 and I'm not sure how the changes interact with TensorFlow Probability. As a general tip, you may want to adjust the learning rate or KL annealing schedule to fit your dataset.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@tmlabonte @luantunez