Public Repository

Last pushed: 5 months ago
Short Description
Tensorflow with some additional dependencies installed
Full Description

It basically follows these steps:

In order to make changes to a docker image and have them persist, do the following.

  1. Start a container, without the -rm flag so that it remains after you exit. Like:

nvidia-docker run --net=host -v /home/ash/Documents/python/testing:/testing_mapped -it gcr.io/tensorflow/tensorflow:latest-gpu /bin/bash

  1. Make any changes you want to persist inside the container. For the TensorFlow image, we want the following.

Install python-tk from apt-get, so we can do run tracing
apt-get update
apt-get install python-tk
Install scikit-image
apt-get install python-skimage
Append the path for libcutl
echo 'export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/extras/CUPTI/lib64' >> ~/.profile
echo 'export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/extras/CUPTI/lib64' >> ~/.bashrc

Docker Pull Command
Owner
ashwhall

Comments (0)