Skip to content

Smoothing fails for constant scalars #1496

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
Closed
Dapid opened this issue Oct 8, 2018 · 2 comments
Closed

Smoothing fails for constant scalars #1496

Dapid opened this issue Oct 8, 2018 · 2 comments

Comments

@Dapid
Copy link

Dapid commented Oct 8, 2018 •
edited
Loading

I have a network training where the learning rate didn't change:
base

When applying smoothing, I get the following artefacts:
smooth

Note that when pushing smoothing all the way to 1 the line disappears completely, since it gets smoothed down to 0.

More info:
TensorBoard 1.10.0 and 1.11.0. (Screenshots correspond to the former).
Linux Fedora 27
Python 3.6
Observed in both Firefox and Chrome
@nfelt
Copy link
Contributor

nfelt commented Oct 10, 2018

Duplicate of #786

@nfelt nfelt marked this as a duplicate of #786 Oct 10, 2018
@nfelt
Copy link
Contributor

nfelt commented Oct 10, 2018

Yes, unfortunately this is a known issue with the current smoothing. Please follow #786.

@nfelt nfelt closed this as completed Oct 10, 2018
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
@nfelt @Dapid