Add package 'docker:image-classification-tf-py-ubuntu-18.04' with CUDA

ctuning / ck-mlperf

This repository is outdated! Join the open MLPerf workgroup to participate in the development of the next generation of automation workflows for MLPerf benchmarks:

https://bit.ly/mlperf-edu-wg

BSD 3-Clause "New" or "Revised" License

32 stars 23 forks source link

Add package 'docker:image-classification-tf-py-ubuntu-18.04' with CUDA #31

Closed bellycat77 closed 4 years ago

bellycat77 commented 5 years ago

add CUDA support to image

psyhtest commented 5 years ago

@bellycat77 By the way, please see if you can draw some inspiration from @guschmue's Dockerfile.gpu. (I don't see him installing cuDNN though.)

guschmue commented 5 years ago

No cudnn install because it comes from the nvidia image:

FROM nvidia/cuda:10.0-cudnn7-devel-ubuntu16.04

psyhtest commented 5 years ago

Thanks @guschmue!

@bellycat77 Could you please try adding another CK entry e.g. docker:image-classification-tf-py.nvidia-cuda10.1-cudnn7-devel-ubuntu18.04 using:

FROM nvidia/cuda:10.1-cudnn7-devel-ubuntu18.04

from the NVIDIA Docker hub.

psyhtest commented 4 years ago

We now base our Docker images that require CUDA support on NVIDIA image e.g. see docker:object-detection-tf-py.tensorrt.ubuntu-18.04