There are a lot of ins and outs for setting up the Dockerfile, etc., so I thought I would share and perhaps save someone else the effort.
This is set up to run Tensorflow without using CUDA/GPU, it uses the CPU instead.
There are a couple of things to change:
USERNAME
,TF_VERSION
,PYTHON_VERSION
at the top ofDockerfile
.- You username in
remoteUser
in devcontainer.json.
From there, you should be able to use the Dev Containers extension to Rebuild and Reopen in Container, then run the tensorflow_mnist_test.py file to see if everything works.