Export a model to CoreML

Trained CatBoost models can be exported to CoreML.

Choose the implementation for more details.

Python package

Method

Parameters

format

Description

The output format of the model.

Set this parameter to “coreml”.

Method

export_parameters

Parameters

Additional format-dependent parameters for:

Apple CoreML

Possible values (all are strings):

  • prediction_type. Possible values are probability and raw.
  • coreml_description
  • coreml_model_version
  • coreml_model_author
  • coreml_model_license
ONNX-ML
  • onnx_graph_name
  • onnx_domain
  • onnx_model_version
  • onnx_doc_string

See the ONNX-ML parameters reference for details.

PMML

Possible values (all are strings):

  • pmml_copyright
  • pmml_description
  • pmml_model_version

See the PMML parameters reference for details.

The following values are supported for Apple CoreML:

  • prediction_type. Possible values are probability and raw.

  • coreml_description

  • coreml_model_version

  • coreml_model_author

  • coreml_model_license

Refer to the example for more details.

Command-line version

The following command keys can be specified for the corresponding commands and are used when the model is trained:

Command

catboost fit

Command keys

--model-format

Key description

A comma-separated list of output model formats.

Set the value of this key to “json”.