google / google/jaxopt

JAXOPT Projected Gradient

Open
#582 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
1.1k
Forks
76
Avg merge
2d 21h
Merged PRs (30d)
1

Description

Hello, I have been using JAXOPT Projected Gradient approach with projection box constraints. And I would like to get more output from the algorithm than it is providing such as I would like to have to loss value at each iteration as output. I believe I can set verbose to False to display the error at each iteration but instead I'd like it as output.

Secondly, I believe the tolerance that is passed to the Proximal Gradient class is not used anywhere and so if that is the case, this might be a bug as no matter what tolerance I set, it would not matter. Could you please look into it? Thanks.

Irrespective of what tolerance I put, it runs to the max iterations and so I looked at this and found that tolerance is not used at all:
https://github.com/google/jaxopt/blob/501cc208c2493395fbe8026b963e7867397403db/jaxopt/_src/proximal_gradient.py#L153

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.