k_clip

Interface to 'Keras' <https://keras.io>, a high-level neural
networks 'API'. 'Keras' was developed with a focus on enabling fast experimentation,
supports both convolution based networks and recurrent networks (as well as
combinations of the two), and runs seamlessly on both 'CPU' and 'GPU' devices.

Tomasz Kalinowski

keras

R Interface to 'Keras'

Daniel Falbel

JJ Allaire

François Chollet

RStudio 

Google 

Yuan Tang

Wouter Van Der Bijl

Martin Studer

Sigrid Keydana

k_clip function

<dl><dt>x</dt>
<dd>Tensor or variable.</dd>
<dt>min_value</dt>
<dd>Float or integer.</dd>
<dt>max_value</dt>
<dd>Float or integer.</dd></dl>

Arguments

This function is part of a set of Keras backend functions that enable
lower level access to the core operations of the backend tensor engine
(e.g. TensorFlow, CNTK, Theano, etc.).
You can see a list of all available backend functions here:
<a href="https://tensorflow.rstudio.com/reference/keras/index.html#backend">https://tensorflow.rstudio.com/reference/keras/index.html#backend</a>.

Keras Backend

Element-wise value clipping. — k_clip

<dl>

<dt>x</dt>
<dd>Tensor or variable.</dd>


<dt>min_value</dt>
<dd>Float or integer.</dd>


<dt>max_value</dt>
<dd>Float or integer.</dd>

</dl>

This function is part of a set of Keras backend functions that enable
lower level access to the core operations of the backend tensor engine
(e.g. TensorFlow, CNTK, Theano, etc.).
You can see a list of all available backend functions here:
<a href='https://tensorflow.rstudio.com/reference/keras/index.html#backend'>https://tensorflow.rstudio.com/reference/keras/index.html#backend</a>.

k_clip: Element-wise value clipping.

Description

Usage

Value

Arguments

Keras Backend