InterpretML

InterpretML is the open-source Python toolkit providing a unified API to compare glassbox models (like EBM) and blackbox explainers (LIME, SHAP) for full model transparency.

InterpretML is your essential, open-source Python toolkit for machine learning interpretability: a unified framework for responsible AI. It delivers two core capabilities: training inherently intelligible **glassbox** models, notably the high-accuracy Explainable Boosting Machine (EBM), and applying **blackbox** explainers like LIME and SHAP to analyze any opaque system. The unified API allows practitioners to efficiently compare multiple methods, generating both global (overall model behavior) and local (individual prediction) explanations. Use InterpretML to debug models, meet auditing requirements, and build necessary trust with stakeholders.

https://interpret.ml