From d97fdf0f29decae53224774db95e5ee441e81ee7 Mon Sep 17 00:00:00 2001 From: Amirsina Torfi Date: Thu, 19 Sep 2019 11:11:43 -0400 Subject: [PATCH] Update README.rst --- .../3-neural_network/convolutiona_neural_network/README.rst | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/docs/tutorials/3-neural_network/convolutiona_neural_network/README.rst b/docs/tutorials/3-neural_network/convolutiona_neural_network/README.rst index 837ddb7..f9771b4 100644 --- a/docs/tutorials/3-neural_network/convolutiona_neural_network/README.rst +++ b/docs/tutorials/3-neural_network/convolutiona_neural_network/README.rst @@ -197,9 +197,6 @@ There are different types of variance-scaling initializers. The one we used in is the one proposed by the paper `Understanding the difficulty of training deep feedforward neural networks `__ -and provided by the TensorFlow. is the one proposed by the paper -`Understanding the difficulty of training deep feedforward neural -networks `__ and provided by the TensorFlow. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -489,7 +486,7 @@ get back to it in another post. The image summaries are created which has the duty of visualizing the input elements to the summary tensor. These elements here -are 3 random images from the train data. In The outputs of different layers will be fed to the relevant summary tensor. +are 3 random images from the train data. In the outputs of different layers will be fed to the relevant summary tensor. Finally, some scalar summaries are created in order to track the *training convergence* and *testing performance*. The collections argument in summary definitions is a supervisor which direct