data => datasets
│
src
├── preprocess
│ └── preprocess.py => preprocessing methods and pipeline
│ └── utils.py => misc methods
└── training
└── main_optuna.py => training script with hyperparameters search
└── models.py => list of all models and hyperparameters
make install
To launch a training:
cd src/training
python main_optuna.py -n 5
-n
is the number of trials. It is possible to use the flag --drop
to drop certain features.
To view all the experiments:
mlflow ui