model-collection

This repository contains a collection of models for testing explainability methods, bias detection methods, and other model evaluation techniques.

Repository structure

Each model should be contained in a separate directory within the repository. The directory should contain the following files:

📁 model_name
│
├── README.md
├── artifact
├── train.py
│
📁 └── data
    │
    └── data_file

• README.md: A description of the model and its purpose.
• artifact: The serialized model artifact. For the same serialisation format the name should be consistent, e.g. model.joblib.
• train.py: The script used to train the model, if applicable.
• data: The directory containing the data used to train the model.

Any requirements for running the model should be included in a common requirements.txt file in the root of the repository.

Changelog

2025-06-05

• Updated dependency: scikit-learn==1.6.1
• Note: New model serialization format.