Liking cljdoc? Tell your friends :D

clojure-tensorflow.gradients


apply-gradientsclj

(apply-gradients variables gradients)
source

depends-on?clj

(depends-on? independent dependent)
source

get-inputsclj

source

get-op-by-nameclj

(get-op-by-name n)
source

get-registered-gradientclj

(get-registered-gradient node)
source

gradientclj

(gradient y x)
source

gradientsclj

(gradients y)
(gradients y & xs)

The symbolic gradient of y with respect to xs. For example, if we wanted to calculate the gradient of our cost function with respect to our weight, we could use.

The symbolic gradient of y with respect to xs.
For example, if we wanted to calculate the gradient of our
cost function with respect to our weight, we could use.
sourceraw docstring

numerical-gradientsclj

(numerical-gradients y & xs)

Calculate the approximate gradient of y with respect to x This is best used for testing differentiated gradients.

Calculate the approximate gradient of y with respect to x
This is best used for testing differentiated gradients.
sourceraw docstring

pathsclj

(paths from to)

Get all paths from one op to another

Get all paths from one op to another
sourceraw docstring

register-gradientclj

(register-gradient op-type functions)
source

registered-gradientsclj

source

relevant-variablesclj

(relevant-variables op)
source

cljdoc is a website building & hosting documentation for Clojure/Script libraries

× close