Skip to content

Conversation

@bschifferer
Copy link
Contributor

Fix some bug related to serving NVT + TensorFlow with Ragged Tensors.

This is an alternative to #316

@bschifferer bschifferer changed the title Fix Sessionbased NVT + TensorFlow Serving [WIP] Fix Sessionbased NVT + TensorFlow Serving Apr 6, 2023
@github-actions
Copy link

github-actions bot commented Apr 6, 2023

Documentation preview

https://nvidia-merlin.github.io/systems/review/pr-317

@karlhigley karlhigley added this to the Merlin 23.04 milestone Apr 7, 2023
@karlhigley karlhigley added the bug Something isn't working label Apr 7, 2023
@karlhigley karlhigley added the breaking Breaking change label Apr 7, 2023
@karlhigley
Copy link
Contributor

This is a breaking change because it removes support for converting ragged Workflow outputs to dense via sparse_max. I think we should go ahead with it, because we're currently in the process of replacing sparse_max with an actual padding operator anyway.

@karlhigley karlhigley changed the title [WIP] Fix Sessionbased NVT + TensorFlow Serving Convert Workflow outputs to Triton-friendly format using TensorTable Apr 7, 2023
@karlhigley karlhigley merged commit 679b54c into NVIDIA-Merlin:main Apr 7, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

breaking Breaking change bug Something isn't working

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants