Creating registered model versions from common libraries

Verta provides wrapper functions to create Verta Standard Models for commonly used libraries including scikit-learn, xgboost, keras and PyTorch. Note, these wrappers can only be used when model objects are available at the time that the RMV is registered.

  my_model_obj = model_object_created_from_library
  model_version = registered_model.create_standard_model_from_library(
    my_model_obj,
    name="v0",
    environment=Python(["scikit-learn"])
  )

Find examples of creating RegisteredModelVersions via convenience functions here:

• Scikit-learn

• Keras

• PyTorch

• XGBoost