This repository contains materials for the tutorial: GPU-Accelerated Data Science for PyData Users

For the in-person version of this tutorial we will use NVIDIA Brev

You can run this tutorial on Google Colab. With a basic free account, you'll have access to:

• An interactive Python environment with GPU support
• Pre-installed RAPIDS libraries (cuDF and cuML)

To run each notebook:

1. Click on the corresponding link below to open it in Google Colab
2. Change the runtime type to T4 GPU
3. Save your changes

Notebook	Link
0 Welcome and Setup
1 Intro to cuDF
2 cudf.pandas
3 Intro to cuML
4 cuml.accel

Notebook	Link
cudf polars engine
cuml clustering models

If you have access to a GPU you can run this locally:

In a terminal:

git clone https://github.com/rapidsai-community/tutorial

Once inside the repository:

conda env create -f local-env.yaml
conda activate rapids-tutorial

During this tutorial we will use different datasets, you can get them all by running the cell below.

python data_setup.py --pydata-vt