Answer by Kilian Batzner for Feed Placeholder with another Placeholder using...
You can modify the DNN placeholders to take the production input placeholder by default. Thus, you can still feed them custom values anytime you want. session = tf.Session() production_input =...
View ArticleAnswer by Jonathan R for Feed Placeholder with another Placeholder using...
Create a input-placeholder and use this as the input to both networks: Eg DNN1: hidden1 = (input * W1) + b1 DNN2: hidden1 = (input * W1) + b1 Can you do this?
View ArticleFeed Placeholder with another Placeholder using tensorflow
The following situation: There is a production server (C++) running tensorflow. The code there expects to have exactly one input Placeholder, that gets fed similar to feed_dict:{'input':...}. I can't...
View Article
More Pages to Explore .....