hiring-challenge

The project uses python3.10 (important for some typing syntax).

• checkpoints/ : A checkpoint for the fight_classifier.model.image_based_model.ProjFromFeatures model. The notebook 2. Image-based classifiers.ipynb shows how to load it.
• dataset/ : The dataset where I load the data. I used DVC to synchronize across computers, but this requires access to my personal s3 account.
• fight_classifier/ : Contains the python package with all of the code, as well as the notebooks used to describe the work.
• joblib_random_forest: I tried to save a naive model (the random forest of notebook 1. Images augmentation and Clever-Hans effect.ipynb) with joblib. I could not do the same with the interesting model. I recommend to install the fight_classifier module and run the notebook 2. Image-based classifiers.ipynb in order to run the model.