Download Jupyter Notebook Examples and instructions - Teradata AI Studio

Teradata® AI Studio - Open Python Framework

Product
Teradata AI Studio
Release Number
1.3
Published
July 2026
ft:locale
en-US
ft:lastEdition
2026-07-28
dita:mapPath
rmu1782935249910.ditamap
dita:ditavalPath
ayr1485454803741.ditaval
dita:id
cgn1782251337635
For more information about Open Python Framework and Ray clusters see:
  1. Go to Teradata AI Studio - Open Python Framework.
  2. Download ( ) OPF-Ray_Demo_Notebooks.zip.
    The zip file contains:
    • GettingStarted_With_Ray.ipynb

      Provides a hands-on introduction to the Teradata Ray Management Service SDK. You will connect to the Ray Management Service and walk through the complete Ray cluster lifecycle: discovering available node groups, provisioning a new cluster, polling it until it is ready, inspecting its details, and finally tearing it down.

    • opf_dask_on_ray.ipnyb

      Demonstrates how to use the Open Python Framework to run Dask DataFrame workloads on a Ray cluster. Demonstrates ETL patterns such as derived columns, filtering, group-by aggregatios, rolling time windows, dimension joins, custom partition transforms, and final report generation.

    • opf_ray_rapids.ipynb

      Demonstrates how to use NVIDIA RAPIDS (cuDF, cuML) with the Ray distributed computing framework. It connects to a remote Ray cluster, verifies GPU availability, runs GPU-accelerated dataframe operations, trains and evaluates cuML models, and launches multiple GPU tasks in parallel from a notebook orchestrator.

    • opf_ray_tune_model.ipynb

      Demonstrates distributed hyperparameter tuning with Ray Tune over the Ray client while loading training data from a Teradata table through the teradatasql Python driver.

    • opf_teradata_access_example.ipynb

      Ray Batch Inference Example with Data Access from Teradata. Demonstrates an end-to-end distributed batch classification pipeline. Includes reads and writes from Teradata.

    • opf_xgboost_distributed_training.ipynb

      Demonstrates distributed, data-parallel model training using XGBoost on Ray Train.