Transfer AutoML Translation Model between projects without retraining [197430185]

Assigned

Feature Request

Status Update

No update yet.

Description

pa...@fao.org

created issue #1

Aug 23, 2021 07:23AM

What you would like to accomplish: I would be able to export/import an AutoML Translation trained model into different projects. Now I have two different projects under the same organization, one is for QA and one for PRODUCTION. I would like to promote the AutoML Translation model I have in the QA environment to the PRODUCTION environment without the need to train again the model from scratch.

How this might work: Implement the export and import functionality for AutoML Translation

If applicable, reasons why alternative solutions are not sufficient: The alternative means to train again from scratch the models

Other information (workarounds you have tried, documentation consulted, etc): -

Comments

se...@google.com <se...@google.com> #2Aug 23, 2021 02:17PM

Assigned to gc...@google.com.

I have informed our engineering team of this feature request. There is currently no ETA for its implementation.

A current workaround would be to check the returned "boundingPoly" [1] "vertices" for the returned "textAnnotations". If the calculated rectangle's heights > widths, than your image is sideways.

[1]

https://cloud.google.com/vision/reference/rest/v1/images/annotate#boundingpoly

[Deleted User] <[Deleted User]> #3Sep 1, 2021 03:34PM

I also need this problem solved :)

Issue 197430185

Description

Issue summary

Comments

se...@google.com <se...@google.com> #2Aug 23, 2021 02:17PM

[Deleted User] <[Deleted User]> #3Sep 1, 2021 03:34PM

Add comment

Issue metadata