regularized - TensorFlow: How to change keep_prob for Dropout without using a feed_dict -

i have build tensorflow model works training , test batches provided input queues. thus, not explicitly feeding data training using standard feed_dict. nevertheless, need implement dropout requires keep_prob placeholder turn off dropout during testing.

i can't find how solve without separate model. there suggestions?

thanks

my question obsolete, made complicated.

one can still feed values placeholders through feed_dict when calling sess.run() though there input queue providing training examples directly.