Setup

Make sure the Anaconda distribution is installed on your machine.

1. Create and activate a new conda environment with Python 3.10

$ conda create -n <environment_name> python=3.10
$ conda activate <environment_name>

2. Install dependencies

• recommenders - pip install recommenders
• mlflow - conda install mlflow; make sure the installed version is at least 2.16 to avoid problems with Databricks integration that are present in older versions
• tensorflow - pip install tensorflow==2.12.0

3. Run Jupyter Notebook

jupyter notebook

4. Deactivate the conda environment when finished:

$ conda deactivate

To completely remove the conda environment execute:

$ conda remove -n <environment_name> --all

Development

Once the conda <environment_name> has been successfully set up, the only requirement is to activate that environment:

$ conda activate <environment_name>

and (usually) starting the Jupyter Notebook:

jupyter notebook