Developer Guide for Intel® Data Analytics Acceleration Library 2018
A uniform initializer is an initializer algorithm to initialize a p-dimensional tensor W ∈ R n1 x ... x np with random numbers uniformly distributed on the interval [a, b), where a, b ∈ R and a < b.
In addition to common parameters of the initializer interface, a uniform initializer has the following parameters:
Parameter |
Default Value |
Description |
|
---|---|---|---|
a |
-0.5 |
The left bound of the interval, a. |
|
b |
0.5 |
The right bound of the interval, b. |
|
DEPRECATED:seed |
777 |
NoteThis parameter is deprecated and will be removed in a future release. |