Describe the bug
I am not sure where part fails, but I do think this is related to newer TF versions.
To Reproduce
When I ran scripts/run_local.sh nets/resnet_at_cifar10_run.py. I got the following error with TF version 1.14 (and thereafter), whereas no problem at all with 1.13:
ValueError: Tensor("Const:0", shape=(1, 1, 3), dtype=float32) must be from the same graph as Tensor("data/ExperimentalParallelInterleaveDataset:0", shape=(), dtype=variant).
Expected behavior
Do you have plan for Tensorflow version 2.0 support?
Desktop (please complete the following information):
Describe the bug I am not sure where part fails, but I do think this is related to newer TF versions.
To Reproduce
When I ran
scripts/run_local.sh nets/resnet_at_cifar10_run.py
. I got the following error with TF version 1.14 (and thereafter), whereas no problem at all with 1.13:Expected behavior Do you have plan for Tensorflow version 2.0 support?
Desktop (please complete the following information):