accessing the tfrecords is done with slim.
The line "provider = slim.dataset_data_provider.DatasetDataProvider( … " sets up the provider.
The line “[image, label] = provider.get([‘image’, ‘label’])” accesses a tfrecord.
I added a print statement "print( "label “, label )”
It outputs “label Tensor(“Reshape:0”, shape=(), dtype=int64, device=/device:CPU:0)”
I have searched for hours, tried run and eval, nothing seems to give me the value of the feature.
A simple script run with eager enabled and not using a graph or slim provides easy access to the features of the trecord. But inside train_image_classifier without eager and using slim is a different issue.
I am at a loss as to how to get the value.
Can anyone give me a hint, it has to be something simple I am just overlooking.