rodrigo-arenas / Sklearn-genetic-opt

ML hyperparameters tuning and features selection, using evolutionary algorithms.
https://sklearn-genetic-opt.readthedocs.io
MIT License
289 stars 73 forks source link

Workflow for mlflow added #74

Closed Turtle24 closed 11 months ago

Turtle24 commented 2 years ago

I think I might have a solution for MLflow finally. I've been working with docker quite a bit lately so I think this might work. Tell me what you think.

codecov[bot] commented 2 years ago

Codecov Report

Merging #74 (1de2dd4) into master (e733cb7) will decrease coverage by 6.74%. The diff coverage is n/a.

:exclamation: Current head 1de2dd4 differs from pull request most recent head 525ad4f. Consider uploading reports for the commit 525ad4f to get more accurate results Impacted file tree graph

@@             Coverage Diff             @@
##            master      #74      +/-   ##
===========================================
- Coverage   100.00%   93.25%   -6.75%     
===========================================
  Files           19       19              
  Lines          727      727              
===========================================
- Hits           727      678      -49     
- Misses           0       49      +49     
Impacted Files Coverage Δ
sklearn_genetic/callbacks/loggers.py 59.61% <0.00%> (-40.39%) :arrow_down:
sklearn_genetic/callbacks/base.py 75.00% <0.00%> (-25.00%) :arrow_down:
sklearn_genetic/space/space.py 80.24% <0.00%> (-19.76%) :arrow_down:
sklearn_genetic/callbacks/early_stoppers.py 89.70% <0.00%> (-10.30%) :arrow_down:
sklearn_genetic/callbacks/validations.py 90.00% <0.00%> (-10.00%) :arrow_down:

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update e733cb7...525ad4f. Read the comment docs.

rodrigo-arenas commented 2 years ago

Hi, I think is a good idea, there are a couple things I'd to comment:

Turtle24 commented 2 years ago

Hi, I think is a good idea, there are a couple things I'd to comment:

  • If we get it working on Docker, I think the test_cleanup function would be unnecessary, so we can remove it

  • The test themselves are failing, I think is because the defined docker port is not being reached, please check the execution log in here

Thanks!

Hey, yeah you're absolutely right. I'll remove the cleanup and I think I'll use testcontainers to get the server port, there's a compose class that'll actually run the compose.yml for us, which is nice.

No worries!

rodrigo-arenas commented 2 years ago

Ey thanks, I see now the tests pass, whici is great! I just have to check why the test coverage is decreasing

Sorry for the delayed answer, I was a bit far away a couple weeks

rodrigo-arenas commented 11 months ago

Closing for now as test have been successful without docker, just to keep dependencies and test as simple as possible