-
Notifications
You must be signed in to change notification settings - Fork 39
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
Save best model #365
Save best model #365
Conversation
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## dev #365 +/- ##
==========================================
+ Coverage 91.48% 94.06% +2.57%
==========================================
Files 12 12
Lines 1022 1027 +5
==========================================
+ Hits 935 966 +31
+ Misses 87 61 -26 ☔ View full report in Codecov by Sentry. |
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.
Looks good. One thing to add is to give a warning to the user that the save_top_k
setting has been removed. The current deprecation warning can be modified to report removal in addition to renaming of settings (e.g. by setting the value to None
and modifying the warning a bit further down).
Additionally, this should be mentioned in the changelog.
* prediction output in model eval mode * eliminate eval command, introduce -e flag for predict command * adapted unit test to new model runner and model functionality * updated documentation * removed log and result files * Generate new screengrabs with rich-codex * Update paper reference (#361) * Bug report template (#360) * bug report template * punctuation, hardware description item * Restrict NumPy to pre-2.0 (#344) * Restrict NumPy to pre-2.0 * Update changelog * Update paper reference (#361) --------- Co-authored-by: Lilferrit <[email protected]> * upgrade codecove to v4 (#364) * implemen eval mode at model runner level, fix unit test * CLI documentation * Generate new screengrabs with rich-codex * requested changes * Generate new screengrabs with rich-codex * evaluation test cases * file warnings, evaluation tests * fixed ubuntu specific test case bug * verify annotated mgf files * verify annotated mgf files * Generate new screengrabs with rich-codex * Save best model (#365) * save best model * save best model * updated unit tests * remove save top k config item * added save_top_k to deprecated config options * changelog entry * test case, formatting * requested changes * prediction output in model eval mode * eliminate eval command, introduce -e flag for predict command * adapted unit test to new model runner and model functionality * updated documentation * removed log and result files * implemen eval mode at model runner level, fix unit test * CLI documentation * Bug report template (#360) * bug report template * punctuation, hardware description item * Restrict NumPy to pre-2.0 (#344) * Restrict NumPy to pre-2.0 * Update changelog * Update paper reference (#361) --------- Co-authored-by: Lilferrit <[email protected]> * requested changes * evaluation test cases * file warnings, evaluation tests * fixed ubuntu specific test case bug * verify annotated mgf files * AnnotatedSpectrumIndex type error * requested changes, changelog entry --------- Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Wout Bittremieux <[email protected]>
Save model with the lowest validation loss to the checkpoint file "best.ckpt". This is implemented by locking
save_top_k
on the validation loss callback to 1 and eliminating thesave_top_k
option from the Casanovo config.