examples/kfp_pipeline: add datapreproc component + a two stage pipeline

[d4l3k]

This adds a simple two stage pipeline that can run on KFP. The first stage downloads the tiny imagenet dataset and normalizes the data. The second stage then loads that dataset and trains on it.

Test plan:

python datapreproc/main.py --input_path http://localhost:8000/tiny-imagenet-200.zip --output_path s3://torchx-test/tristanr/data_path/

docker build -t torchx/examples .
docker run --env-file /tmp/aws-env -it --name torchx-classy-vision --rm torchx-classy-vision:latest python3 datapreproc/main.py --output_path s3://torchx-test/tristanr/data_path/
docker tag torchx/examples:latest 495572122715.dkr.ecr.us-west-2.amazonaws.com/torchx/examples:latest
docker push 495572122715.dkr.ecr.us-west-2.amazonaws.com/torchx/examples:latest
./kfp_pipeline.py --data_path s3://torchx-test/tristanr/data_path/ --output_path s3://torchx-test/tristanr/output_path/

[facebook-github-bot]

@d4l3k has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

[facebook-github-bot]

@d4l3k has updated the pull request. You must reimport the pull request before landing.

[facebook-github-bot]

@d4l3k has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

[facebook-github-bot]

@d4l3k has updated the pull request. You must reimport the pull request before landing.

[facebook-github-bot]

@d4l3k has updated the pull request. You must reimport the pull request before landing.

[facebook-github-bot]

@d4l3k has updated the pull request. You must reimport the pull request before landing.

[facebook-github-bot]

@d4l3k has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

[facebook-github-bot]

@d4l3k merged this pull request in 523eb1a.