gnn-model-registry

To version and catalog GNN models automatically, run the model registry step to discover GNN files in your pipelines, extract metadata like authorship and licenses, and populate a centralized catalog with deterministic hashing and timestamps.

When registering GNN model files, the extracted metadata includes authorship, licenses, versions, and dates to support governance and reproducibility across Active Inference projects.

A centralized model registry helps with GNN model discovery by indexing files into a persistent catalog, enabling efficient queries and lifecycle management across different projects and teams.

Yes, you can use this model registry for GNN lifecycle management across multiple pipelines, as it applies deterministic hashing and timestamps to parse, version, and index models consistently.

You need deterministic hashing and timestamps for GNN model versioning to ensure consistent tracking of model files, enabling accurate discovery and governance in Active Inference pipelines.