Model Wrappers

The model wrapper modules convert trained machine learning models from popular Python libraries into callables compatible with URANIE’s Relauncher module. This enables you to use your trained models within URANIE workflows.

PyTorch Models

Wrap PyTorch model for URANIE using pytorch_wrapper.uranie_evaluator().

See: PyTorch Wrapper for detailed documentation and examples.

Scikit-learn Models

Wrap scikit-learn estimators for URANIE using sklearn_wrapper.uranie_evaluator().

See: Scikit-learn Wrapper for detailed documentation and examples.

TensorFlow Models

Wrap TensorFlow/Keras models for URANIE using tensorflow_wrapper.uranie_evaluator().

See: TensorFlow Wrapper for detailed documentation and examples.

This section includes the following pages:

• pytorch_wrapper
  • Examples
  • Implementation Details
  • Supported Model Types
  • Supported Output Types
• sklearn_wrapper
  • Examples
  • Implementation Details
  • Supported Model Types
  • Supported Output Types
• tensorflow_wrapper
  • Examples
  • Implementation Details
  • Supported Model Types
  • Supported Output Types