Protege Runtime

The Protege Runtime module provides the core infrastructure for executing models and backends in production or research environments. It offers an abstraction layer over inference, logging, configuration, and other runtime services.

✨ Features

• Runtime Abstractions

  • Interfaces for model loading, prediction, and lifecycle control.
  • Decouples model logic from system-specific implementations.

• Prediction & IO Utilities

  • Standardized input/output schemas for predictions.
  • Tools for validation, error handling, and telemetry.

• Backend Integration

  • Compatible with batch, streaming, or real-time execution environments.

• Lightweight & Extensible

  • Designed to be composable with other packages like protege-engine or protege-pipelines.

🛠️ Use Cases

• Running models in production pipelines.
• Testing inference workflows locally.
• Standardizing prediction formats across model versions.

---