Closed tarrade closed 4 years ago
No idea where the issue is coming from. One issue could be from AI Platform training
Discussion with GCP support and Tensorflow team https://github.com/tensorflow/tensorflow/issues/39913
this issue is now fixed but many more related to GCP. Ops need to be declared after TPUCluster Resolver.
No idea where the issue is coming from. One issue could be from AI Platform training