Optuna

[nikk-nikaznan]

To be reviewed after the validation PR
Implementing Optuna to hopefully able to get the best hyper parameters for our problem (especially learning rate)

• For now having options for number of epochs and LR.
• log in the mlflow the best params and the validation precision of that.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 27.08333% with 35 lines in your changes missing coverage. Please review.

Project coverage is 36.83%. Comparing base (c372e30) to head (b58c731).

Files	Patch %	Lines
crabs/detection_tracking/optuna_fn.py	28.57%	25 Missing ⚠️
crabs/detection_tracking/models.py	14.28%	6 Missing ⚠️
crabs/detection_tracking/train_model.py	33.33%	4 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #136      +/-   ##
==========================================
- Coverage   37.12%   36.83%   -0.30%     
==========================================
  Files          19       20       +1     
  Lines        1309     1352      +43     
==========================================
+ Hits          486      498      +12     
- Misses        823      854      +31     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.