Skip to content

Tutorial - 20 newsgroups

In this example, we will use the 20 newsgroups text dataset to train a text classification model. This example is inspired by the scikit-learn example for text classification.

Step 1: Logging in to the jupyterhub server

Access the jupyterhub server by navigating to the following URL: 

https://jupyterhub.{IDOML_DOMAIN}

Step 2: Cloning the example repository

Open a terminal in the JupyterLab interface. Run the following commands:

git clone https://github.com/serval-uni-lu/idoml-pipeline-example-text-clf-20news.git

cd idoml-pipeline-example-text-clf-20news

Open the Git graphical interface on the left side of the JupyterLab interface. Click on the clone repository button.

Image title

Paste the following URL: 

https://github.com/serval-uni-lu/idoml-pipeline-example-text-clf-20news.git

Step 3: Running the pipeline

Open the pipeline file text-clf-20news.pipeline in the JupyterLab interface. Then click on the run button to execute the pipeline.

Image title

Then select the IDOML runtime defined previously and click on the OK button.

Step 4: Checkout the pipeline

We can now check out the pipeline by navigating to the airflow server with the following URL: 

http://airflow.{IDOML_DOMAIN}

Note

If it is the first time you are submitting a pipeline to airflow, you will need to wait airflow to sync the DAGs. This can take a few minutes.

In the meantime, you can check the dag file in the repository which airflow is tracking for the pipeline execution.

Step 5: Check the Pipeline execution

The pipeline execution can be checked by navigating to the Minio server with the following URL: 

http://minio.{IDOML_DOMAIN}

Login with the following the user account, the execution logs can be found in the bucket defined previously in the IDOML runtime.

Step 6: Check the ML outcome

After the pipeline execution, we can check the ML outcome by navigating to the MLflow server with the following URL: 

http://idoml.mlflow.{IDOML_DOMAIN}