search

mljar / mljar-supervised

Python package for AutoML on Tabular Data with Feature Engineering, Hyper-Parameters Tuning, Explanations and Automatic Documentation
https://mljar.com
MIT License
2.98k stars 393 forks source link

Optuna Mode fails. Optuna has migrated to optuna-integration. #729

Open decmca opened 3 weeks ago

decmca commented 3 weeks ago

Exception in LightgbmObjective Optuna's integration modules for third-party libraries have started migrating from Optuna itself to a package called optuna-integration. The module you are trying to use has already been migrated to optuna-integration. Please install the package by running pip install optuna-integration.

Is it possible to upgrade mljar-supervised so that Optuna functionality is still available after the implementation of optuna-integration?

pplonski commented 3 weeks ago

Hi @decmca,

Yes, it is possible, we will have intern in the summer, I hope this issue will be fixed. In the meantime, please try to use older Optuna version.

May I ask what is your use case?

decmca commented 3 weeks ago

Yeah i used Optuna 3.5.0 (current version with the Optuna-Integration requirement is 3.6.1), maybe set 3.5.0 as the highest version in requirements.txt for now?

I'm new here, just testing out AutoML libraries, I think yours is amazing!

maciekmalachowski commented 1 week ago

Hi @decmca, As you asked, I have changed the code so that Optuna functionality is still available. Among other things, I changed optuna.integration to optuna_integration in xgboost and lightgbm packages and also changed optuna to optuna-integration in requirements.txt, because optuna is included in optuna-integration.

Best wishes