Cli commands


kedro mlflow init: this command is needed to initalize your project. You cannot run any other commands before you run this one once. It performs 2 actions: - creates a mlflow.yml configuration file in your conf/local folder - replace the src/PYTHON_PACKAGE/ file by an updated version of the template. If your template has been modified since project creation, a warning will be raised. You can either run kedro mlflow init --force to ignore this warning (but this will erase your or set hooks manually.

init has two arguments:

  • --env which enable to specifiy another environment where the mlflow.yml should be created (e.g, base)

  • --force which overrides the mlflow.yml if it already exists and replaces it with the default one. Use it with caution!


kedro mlflow ui: this command opens the mlflow UI (basically launches the mlflow ui command )

ui accepts the port and host arguments of mlflow ui command. The default values used will be the ones defined in the mlflow.yml configuration file under the ui.

If you provide the arguments at runtime, they wil take priority over the mlflow.yml, e.g. if you have:

# mlflow.yml
    localhost: ""
    port: "5001"


kedro mlflow ui --port=5002

will open the ui on port 5002.


kedro mlflow modelify: this command converts a kedro pipeline to a mlflow model and logs it in mlflow. It enables distributing the kedro pipeline as a standalone model and leverages all mlflow serving capabilities (as an API).

modelify accepts the following arguments :

  • --pipeline, -p: The name of the kedro pipeline name registered in that you want to convert to a mlflow model.

  • --input-name, -i: The name of the kedro dataset (in catalog.yml) which is the input of your pipeline. It contains the data to predict on.

  • --infer-signature : A boolean which indicates if the signature of the input data should be inferred for mlflow or not.

  • --infer-input-example : A boolean which indicates if the input_example of the input data should be inferred for mlflow or not

  • --run-id, -r : The id of the mlflow run where the model will be logged. If unspecified, the command creates a new run.

  • --run-name: The name of the mlflow run where the model will be logged. Defaults to "modelify".

  • --copy-mode : The copy mode to use when replacing each dataset by a MemoryDataset. Either a string (applied all datasets) or a dict mapping each dataset to a copy_mode.

  • --artifact-path" : The artifact path of mlflow.pyfunc.log_model, see

  • --code-path : The code path of mlflow.pyfunc.log_model, see

  • --conda-env : “The conda environment of mlflow.pyfunc.log_model, see

  • --registered-model-name : The registered_model_name of mlflow.pyfunc.log_model, see

  • --await-registration-for: The await_registration_for of mlflow.pyfunc.log_model, see*

  • --pip-requirements : The pip_requirements of mlflow.pyfunc.log_model, see

  • --extra-pip-requirements : The extra_pip_requirements of mlflow.pyfunc.log_model, see