NVTabular Example Notebooks

In this library, we provide a collection of Jupyter notebooks, which demonstrates the functionality of NVTabular.

Inventory

Getting Started with NVTabular: Get started with NVTabular by processing data on the GPU.
Advanced NVTabular workflow: Understand NVTabular in more detail by defining more advanced workflows and learn about different operators
Running on multiple GPUs or on CPU: Run NVTabular in different environments, such as multi-GPU or CPU-only mode.

In addition, NVTabular is used in many of our examples in other Merlin libraries. You can explore more complex processing pipelines in following examples:

Running the Example Notebooks

You can run the example notebooks by installing NVTabular and other required libraries. Alternatively, Docker containers are available from the NVIDIA GPU Cloud (NGC) at http://ngc.nvidia.com/catalog/containers/ with pre-installed versions. Depending on which example you want to run, you should use any one of these Docker containers:

merlin-hugectr (contains NVTabular with HugeCTR)
merlin-tensorflow (contains NVTabular with TensorFlow)
merlin-pytorch (contains NVTabular with PyTorch)

Beginning with the 22.06 release, each container includes the software for training models and performing inference.

To run the example notebooks using Docker containers, perform the following steps:

Pull and start the container by running the following command:
```
docker run --gpus all --rm -it \
  -p 8888:8888 -p 8797:8787 -p 8796:8786 --ipc=host \
  <docker container> /bin/bash
```
The container opens a shell when the run command execution is completed. Your shell prompt should look similar to the following example:
```
root@2efa5b50b909:
```

Start the JupyterLab server by running the following command:

jupyter-lab --allow-root --ip='0.0.0.0'

View the messages in your terminal to identify the URL for JupyterLab. The messages in your terminal show similar lines to the following example:

Or copy and paste one of these URLs:
http://2efa5b50b909:8888/lab?token=9b537d1fda9e4e9cadc673ba2a472e247deee69a6229ff8d
or http://127.0.0.1:8888/lab?token=9b537d1fda9e4e9cadc673ba2a472e247deee69a6229ff8d

Open a browser and use the 127.0.0.1 URL provided in the messages by JupyterLab.
After you log in to JupyterLab, navigate to the /nvtabular directory to try out the example notebooks.

Troubleshooting

If you experience any trouble running the example notebooks, check the latest troubleshooting documentation.