Rapids End to End ML Demo

Contents

1. Motivate rapids [ show coverage of modern data science tools ]

2. Generate a synthetic dataset

• 2.1 - Split into train and test set
• 2.2 - Visualize sub-datasets

3. ETL

• 3.1 - Load data [ csv read ]
• 3.2 - Transform data [ standard scaler ]

4. Model Building

• 4.1 - Train CPU and GPU XGBoost classifier models
• 4.2 - Use trained models for inference
• 4.3 - Compare accuracy
• 4.4 - Visualize sample boosted trees & model predictions

5. Extensions

• 5.1 - Create an ensemble with a clustering model [ DBScan ]
• 5.2 - Export data to DeepLearning Framework [ PyTorch ]

Install & Run Demo

Video Walkthrough Link

1 -- clone repository

git clone https://github.com/miroenev/rapids && cd rapids

2 -- build container [ takes 5-10 minutes ]

sudo docker build -t rapids-demo:latest .

3 -- launch/run the container [ auto starts jupyter notebook ]

sudo docker run --runtime=nvidia -it --rm -p 8888:8888 -p 8787:8787 rapids-demo:latest

4 -- connect to notebook

i) navigate browser to IP of machine running container at port 8888
    e.g., http://127.0.0.1:8888

ii) in the rapids folder launch the notebook titled 
    rapids_ml_workflow_demo.ipynb