Principle:Interpretml Interpret EBM JSON Deserialization
| Field | Value |
|---|---|
| Sources | Repo: InterpretML |
| Domains | Model_Persistence, Interoperability |
| Updated | 2026-02-07 |
Overview
A deserialization procedure that reconstructs a fitted EBM model from a JSON representation for deployment and inference.
Description
EBM JSON Deserialization takes a JSON dictionary (produced by to_jsonable) and restores all model attributes on an empty EBM instance, making it ready for prediction and explanation. It handles version checking, type conversion from Python native types back to numpy arrays, and validation of the JSON structure.
Usage
Use this when loading a previously serialized EBM for deployment, inference, or further analysis.
Theoretical Basis
The inverse of serialization: reconstructs F(x) = B_0 + Sum f_t(x_St) by restoring all bin definitions and score lookup tables from JSON.
Note: This function is prefixed with UNTESTED_ indicating it is considered experimental by the maintainers.