How to import MLFlow model from a managed folder on Dataiku DSS?

aa_sofia

Hi!

I encounter a problem when importing an MLFlow model saved in a managed folder on Dataiku DSS. I use the following code:

import dataiku
client = dataiku.api_client()
project = client.get_project("PROJECT_ID")

# 1. Create DSS Saved Model
saved_model = project.create_mlflow_pyfunc_model(name, prediction_type)
# 2. Load the MLflow Model as a new version of DSS Saved Model
## from a DSS managed folder:
mlflow_version = saved_model.import_mlflow_version_from_managed_folder('version_id', 'managed_folder_id', path_of_model, 'code-environment-to-use')

Which I found in the doc below :

https://doc.dataiku.com/dss/latest/mlops/mlflow-models/importing.html

However, I don't understand what I should write as a value for the "version_id" and "path_of_model" parameters. For the second one, is it the path of the folder containing the model and other files or the path of the model (binary file in .pth format)? Do you have a concrete example of how using this code please?

When I run the code I get the following error: "libcublas.so.11: cannot open shared object file: No such file or directory". Do you know what that means please?
Thank you in advance for your help!

Have a nice day!

Operating system used: Windows