IssueTracker

The outputs format differ between global and regional endpoints when call the AI Platform prediction API.
Predictions on the global endpoint returns the name of the output layer and the regional endpoint doesn’t.

This issue has been reported on this Stackoverflow post. And it happens both with the gcloud command and using the library (See [1]).

gcloud examples from the post (results are similar when using the Python library):

$ gcloud ai-platform predict --model=my_model_id --json-request=data.json --region=europe-west1
Using endpoint [https://europe-west1-ml.googleapis.com/]
[[5.64439188e-06, 1.11136234e-09, 4.66703168e-05, 1.34729596e-08,...]]

$ gcloud ai-platform predict --model=my_model_id --json-request=data.json
Using endpoint [https://ml.googleapis.com/]
DENSE_1 [5.644391876558075e-06, 1.1113623354930269e-09, 4.6670316805830225e-05,...]

From the list of endpoints, I have already tested the following:

• europe-west1
• asia-east1
• us-east1
• australia-southeast1

They don’t return the output layer’s name like the global endpoint does.