MLflow integration for H2O.
Load a H2O model from a local file (if run_id is None) or a run.
This function expects there is a h2o instance initialised with h2o.init().
Loads an H2O model from the directory at
pathas a Python Function model. Note that this method calls
h2o.init(...), so the right version of h2o(-py) must be in the environment. The arguments given to
h2o.init(...)can be customized in
path/h2o.yamlunder the key
log_model(h2o_model, artifact_path, **kwargs)
Log a H2O model as an MLflow artifact for the current run.
save_model(h2o_model, path, conda_env=None, mlflow_model=<mlflow.models.Model object>, settings=None)
Save a H2O model to a path on the local file system.
- h2o_model – H2O model to be saved.
- path – Local path where the model is to be saved.
- mlflow_model – MLflow model config this flavor is being added to.