RAPIDS Demo Container

Get started with our preconfigured RAPIDS demo container, featuring several demo notebooks using cuDF, cuML, cuGraph, Dask, and XGBoost

  1. RAPIDS Demo Container
    1. Current Version
      1. RAPIDS 0.15 - 26 August 2020
    2. Former Version
      1. RAPIDS 0.14 - 3 June 2020
    3. Image Types
    4. Image Tag Naming Scheme
    5. Prerequisites
    6. Usage
      1. Start Container and Notebook Server
        1. Preferred - Docker CE v19+ and nvidia-container-toolkit
        2. Legacy - Docker CE v18 and nvidia-docker2
      2. Use JupyterLab to Explore the Notebooks
      3. Custom Data and Advanced Usage
        1. Preferred - Docker CE v19+ and nvidia-container-toolkit
        2. Legacy - Docker CE v18 and nvidia-docker2
      4. Access Documentation within Notebooks
    7. More Information
    8. Where can I get help or file bugs/requests?

Current Version

RAPIDS 0.15 - 26 August 2020

Versions of libraries included in the 0.15 images:

  • cuDF v0.15, cuML v0.15, cuGraph v0.15, RMM v0.15, cuSpatial v0.15, cuSignal v0.15
    • IMPORTANT: CUDA 10.0 & Python 3.6 support ended in v0.14; v0.15 includes CUDA 11.0 & Python 3.8 support
    • NOTE: See RAPIDS Notices for release changes for clx & cuxfilter as well as other recent changes

Former Version

RAPIDS 0.14 - 3 June 2020

Versions of libraries included in the 0.14 images:

Image Types

The RAPIDS images are based on nvidia/cuda, and are intended to be drop-in replacements for the corresponding CUDA images in order to make it easy to add RAPIDS libraries while maintaining support for existing CUDA applications.

RAPIDS images come in three types, distributed in two different repos:

The rapidsai/rapidsai repo contains the following:

  • base - contains a RAPIDS environment ready for use.
    • TIP: Use this image if you want to use RAPIDS as a part of your pipeline.
  • runtime - extends the base image by adding a notebook server and example notebooks.
    • TIP: Use this image if you want to explore RAPIDS through notebooks and examples.

The rapidsai/rapidsai-dev repo adds the following:

  • devel - contains the full RAPIDS source tree, pre-built with all artifacts in place, and the compiler toolchain, the debugging tools, the headers and the static libraries for RAPIDS development.
    • TIP: Use this image to develop RAPIDS from source.

Image Tag Naming Scheme

The tag naming scheme for RAPIDS images incorporates key platform details into the tag as shown below:

0.15-cuda10.1-runtime-ubuntu18.04-py3.7
 ^       ^    ^        ^         ^
 |       |    type     |         python version
 |       |             |
 |       cuda version  |
 |                     |
 RAPIDS version        linux version

To get the latest RAPIDS version of a specific platform combination, simply exclude the RAPIDS version. For example, to pull the latest version of RAPIDS for the runtime image with support for CUDA 10.1, Python 3.7, and Ubuntu 18.04, use the following tag:

cuda10.1-runtime-ubuntu18.04-py3.7

Many users do not need a specific platform combination but would like to ensure they’re getting the latest version of RAPIDS, so as an additional convenience, a tag named simply latest is also provided which is equivalent to cuda10.1-runtime-ubuntu16.04-py3.7.

Prerequisites

  • NVIDIA Pascal™ GPU architecture or better
  • CUDA 10.1/10.2/11.0 with a compatible NVIDIA driver
  • Ubuntu 16.04/18.04 or CentOS 7
  • Docker CE v18+
  • nvidia-docker v2+

Usage

Start Container and Notebook Server

Preferred - Docker CE v19+ and nvidia-container-toolkit

$ docker pull rapidsai/rapidsai:cuda10.1-runtime-ubuntu18.04-py3.7
$ docker run --gpus all --rm -it -p 8888:8888 -p 8787:8787 -p 8786:8786 \
         rapidsai/rapidsai:cuda10.1-runtime-ubuntu18.04-py3.7

NOTE: This will open a shell with JupyterLab running in the background on port 8888 on your host machine.

Legacy - Docker CE v18 and nvidia-docker2

$ docker pull rapidsai/rapidsai:cuda10.1-runtime-ubuntu18.04-py3.7
$ docker run --runtime=nvidia --rm -it -p 8888:8888 -p 8787:8787 -p 8786:8786 \
         rapidsai/rapidsai:cuda10.1-runtime-ubuntu18.04-py3.7

NOTE: This will open a shell with JupyterLab running in the background on port 8888 on your host machine.

Use JupyterLab to Explore the Notebooks

Notebooks can be found in the following directories within the 0.15 container:

  • /rapids/notebooks/clx - CLX demo notebooks
  • /rapids/notebooks/cugraph - cuGraph demo notebooks
  • /rapids/notebooks/cuml - cuML demo notebooks
  • /rapids/notebooks/cusignal - cuSignal demo notebooks
  • /rapids/notebooks/cuxfilter - cuXfilter demo notebooks
  • /rapids/notebooks/xgboost - XGBoost demo notebooks

For a full description of each notebook, see the README in the notebooks repository.

Custom Data and Advanced Usage

You are free to modify the above steps. For example, you can launch an interactive session with your own data:

Preferred - Docker CE v19+ and nvidia-container-toolkit

$ docker run --gpus all --rm -it -p 8888:8888 -p 8787:8787 -p 8786:8786 \
         -v /path/to/host/data:/rapids/my_data \
                  rapidsai/rapidsai:cuda10.1-runtime-ubuntu18.04-py3.7

Legacy - Docker CE v18 and nvidia-docker2

$ docker run --runtime=nvidia --rm -it -p 8888:8888 -p 8787:8787 -p 8786:8786 \
         -v /path/to/host/data:/rapids/my_data \
                  rapidsai/rapidsai:cuda10.1-runtime-ubuntu18.04-py3.7

This will map data from your host operating system to the container OS in the /rapids/my_data directory. You may need to modify the provided notebooks for the new data paths.

Access Documentation within Notebooks

You can check the documentation for RAPIDS APIs inside the JupyterLab notebook using a ? command, like this:

[1] ?cudf.read_csv

This prints the function signature and its usage documentation. If this is not enough, you can see the full code for the function using ??:

[1] ??pygdf.read_csv

Check out the RAPIDS documentation for more detailed information and a RAPIDS cheat sheet.

More Information

Check out the RAPIDS and XGBoost API docs.

Learn how to setup a mult-node cuDF and XGBoost data preparation and distributed training environment by following the mortgage data example notebook and scripts.

Where can I get help or file bugs/requests?

Please submit issues with the container to this GitHub repository: https://github.com/rapidsai/docs

For issues with RAPIDS libraries like cuDF, cuML, RMM, or others file an issue in the related GitHub project.

Additional help can be found on Stack Overflow or Google Groups.