"AWS token invalid" error when running

[alexander-myltsev]

Thanks for the demo.

First, I found a command as follows that actually runs:

.env/bin/python multi_step_mnist.py --with kube:cpu=3,memory=4000,image=tensorflow/tensorflow:latest-py3 --no-pylint --environment=conda --datastore=s3  run --num_training_examples 10000

When I run it, I get the error:

S3 datastore operation _put_s3_object failed (An error occurred (InvalidToken) when calling the PutObject operation: The provided token is malformed or otherwise invalid.). Retrying 7 more times..

How should I properly generate AWS Access token to make it valid for your demo?