bschifferer

Yes that worked around works for me - thanks

I dont think we need an extra example for Column Tagging API. I integrated the feature in the scaling criteo example.

I benchmarked NVTabular+TensorFlow for the criteo example and numbers do not as expected. I think we need a follow-up / profile investigation in NVTabular latency numbers. 

* Added column tagging API feature * Fixed TensorFlow example to use numerical input features

@karlhigley @jperez999 have you started some example questions?

@marcromeyn is this PR still relevant? Are you active working on it? Can we close it OR move it to need prioritization/clarification?

I thought we integrate them into the Merlin/models/examples similar to `https://github.com/NVIDIA-Merlin/Transformers4Rec/tree/main/examples/tutorial`

@rnyak Can you add your comments/requirements, as well?

As an success criteria, we need to have benchmarks for each of the point above: - How does throughput change? (E.g. TF Keras vs. SOK. vs TFDE vs. reduced precision...

@sararb do you know if the API update is finished and we can implement this example?