Model Registry#

The Model Registry microservice provides a centralized repository that facilitates the management of AI models.

Overview#

The Model Registry microservice provides a centralized repository that facilitates the management of AI models. It plays a crucial role in the machine learning life cycle, providing a structured storage environment for models and their metadata. It stores each model’s details, such as its version, precision, target device, score and more. This allows for easy tracking of each model over time and ensures that each model is always readily available for deployment.

It is particularly useful for developers as it enables effective model management. Developers can use the registry to compare different models and select the most suitable model for deployment. The registry provides transparency as the details of each model are stored and can be easily accessed.

It also promotes collaboration among users. Since the registry is a centralized repository, it can be accessed by different applications, services, or developers enabling them to store, and review models collectively. The model registry is an essential tool for developers deploying machine learning models, as it streamlines model management, fosters collaboration, and ultimately, aids in improving model deployments.

  • Programming Language: Python* 3

How It Works#

The Model Registry microservice works by serving as a centralized repository for models where, their versions, and metadata are stored. The software behind the microservice is designed to handle the storage, versioning, and metadata management of each model. It also provides functionalities for storing, searching and retrieving model artifacts via a RESTful API.

The software fulfills the promise described in the Overview via its various components.

Model Registry - High Level Architecture Diagram

REST API Endpoints#

The REST API endpoints serve as the primary interface for interacting with the model registry microservice. These endpoints allow users to perform various operations such as registering new models, updating existing models, retrieving model metadata, and deleting models.

MLflow* Model Registry#

The MLflow* Model Registry is a centralized repository used to manage machine learning models. By integrating MLflow*, the microservice stores models in a structured manner, making it easy to manage and deploy models reliably.

Interface - Intel® Geti™#

The Intel® Geti™ interface is designed to facilitate seamless integration with an Intel® Geti™ server. By providing a direct connection, this ensures that users can easily manage their models in a deployed environment within the Intel ecosystem, taking advantage of its advanced features and performance optimizations.

Interface - Object Storage#

This interface enables the microservice to interact with an object storage solution for the storage and retrieval of model files and associated artifacts.

Data Storage - Relational Database Container#

The Relational Database Container is responsible for storing structured data related to the models, such as metadata, and version history.

Data Storage - Object Storage Container#

The Object Storage Container is used to store unstructured data, such as model binaries and other files.

Data Storage - Docker* Volume#

The Docker* Volume is used to persist data generated and used by the containers, ensuring that data remains available even if the containers are restarted or recreated.

Learn More#

  • Get started with the microservice using the Get Started Guide.

  • Follow step-by-step examples to become familiar with the core functionality of the microservice, in Tutorials.