UofM EECS 545 Fall 2020 Project: Domain Exploration Through Artificial Curiosity
The original docker image can be accessed with a docker pull tsender/tensorflow-gpu (note, you may need to login to pull the image). If you make any changes to the dockerfile to run it properly on your own system, you maydo so. After you make your changes, just build the dockerfile on your system with your docker username and the tag latest-gpu
cd dockerfiles/
docker build -t <username>/tensorflow:latest-gpu .
Note 1: This dockerfile includes openCV and you may want to lower the number of threads to use so it doesn't use all of your CPU power when building.
Note 2: I (Ted) will be running this exact dockerfile. If you need additional dependencies, let me know so I can add them.
Place the provided sample shell script in the SAME level as this project folder on your system. Then, run the docker file with
sh run_tf_docker_sample.sh eecs545_artificial_curiosity
The shell script is configured to handle volume mapping for various directories. Depending on your system, you may need to remove/add some volumes.