ConverJens

@kennethyang404 Awsome! Anything else needed from me?

@jiyongjung0 You are also set as a reviewer on this PR but i don't know if you also have to approve?

> @ConverJens We only need one approval, but I'm happy to approve this again. :) Now tfx team will move this PR into our internal review system and start the...

@bli00 You can read and deserialise these tf examples using the schema that Transform provides for you as well. Visualising this schema is usually a quick way of gaining understanding...

@dineshvelmuruga TF serving isn't a destination, it's an application. You decide where TF serving is to look for model, be it S3, GS, local disc or a PCV in Kubernetes....

@dineshvelmuruga I can't say for sure, I haven't tested but it seems like it: https://www.tensorflow.org/io/tutorials/azure. TFX itself doesn't have explicit storage support but rather relies on Tensorflows fs support. At...

@dineshvelmuruga There are just two things that needs to be done: 1. Set the required env vars 2. point to an s3 location For 1. you need to set: S3_ENDPOINT:...

@dineshvelmuruga What do you mean "read the image"? Do you mean read data with ExampleGen from S3? If you set the env vars on your pipeline as I specified, then...

@dineshvelmuruga Depends on how you run it. If running in KFP, set the appropriate pipeline_operator_funcs in your KubeflowDagRunnerConfig when compiling your pipeline.

@dineshvelmuruga Your setup looks ok. The only thing that you should probably do is to set your env var S3_USE_HTTPS=0 since you aren't using https and you should also specify...