Skip to content

gpuopenanalytics/demo-docker-dl

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 

Repository files navigation

README

Code

To see how to use pytorch tensors and cuda kernels to avoid transfering GDF data to RAM when using GOAI with pytorch, check out the following code (which is the core of what is being run in this demo)...

notebook-demo-docker/demo/notebooks/mapd_to_pygdf_to_pytorch_tensors.ipynb

Docker Build

To build the docker image, go into the ./notebook-demo-docker and run:

docker build -t conda_cuda_base:latest ./base
docker build -t cudf:latest ./demo

Run Docker

nvidia-docker run -p 8888:8888 -ti cudf:latest

This launches mapd and the notebook automatically.

Login to the notebook with your browser by following the URL printed on the terminal.

Run Notebook

Open mapd_to_pygdf_to_pytorch_tensors.ipynb and hit "Run All" to test. This notebook should run to the end without error.

Diagnostic

To run on specific GPUs, use NV_GPU.

For example:

NV_GPU=0 nvidia-docker run -p 8888:8888 -ti cudf:latest

About

demo docker image for deep learning

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published