Model registry
The Aidin model registry is a central catalog for your models: versions, metadata, artifacts, and lifecycle state, shared across notebooks, pipelines, and serving.
What the registry gives you
Section titled “What the registry gives you”- Model versions — each trained artifact recorded as a version with lineage back to the work that produced it.
- Metadata — parameters, metrics, and descriptions attached to versions.
- Lifecycle states — organize versions through stages such as development, staging, and production.
- Deployment source — registered versions are the input to managed inference endpoints.
How you use it
Section titled “How you use it”The typical flow is: train in a notebook or pipeline → register the version → promote it through lifecycle states → deploy it to an endpoint.
Access and tenancy
Section titled “Access and tenancy”The registry is scoped to your project like the rest of your workspace: you see your project’s models, isolated from other tenants, using your normal dashboard sign-in. No separate registry credentials are required.