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
Avoid maximum limit when mlflow saves information #3651
Conversation
@xadrianzetx Thanks. It's perfect! |
Cool. Will check it out over the weekend. |
Codecov Report
@@ Coverage Diff @@
## master #3651 +/- ##
==========================================
- Coverage 90.66% 90.65% -0.01%
==========================================
Files 162 162
Lines 12585 12597 +12
==========================================
+ Hits 11410 11420 +10
- Misses 1175 1177 +2
📣 Codecov can now indicate which changes are the most critical in Pull Requests. Learn more |
@contramundum53 Could you review this PR if you have time? |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Code looks great, but I was wondering if we (or anyone else using mlflow
APIs) should care about implementation details such as batch sizes or max tag length. This makes us tightly coupled and might result in additional maintenance cost. Imo, ideally all this should be handled by mlflow
internally, and at quick glance it seems like there are some existing solutions.
Approach from this PR is fine for now, but it could be worth to reconsider later on. What do you think @nzw0301?
run_dict = run.to_dictionary() | ||
|
||
# The `tags` contains param's distributions and other information too, such as trial number. | ||
print(run_dict["data"]["tags"]) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Sorry, this line should be removed.
Thank you for sharing your opinion and I agree with you; this might be able to be handled by mlflow's python API. So let me create a feature request issue on mlflow repo. If the issue sounds reasonable to them, I'll add a TODO comment to this PR (or as a follow-up). |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Awesome! Seems like mlflow
maintainers agree with your proposition, so I guess this one could be merged to unblock and removed later once changes are published mlflow
side. LGTM!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM!
Motivation
Takeover #3530
Description of the changes