Intel(R) Math Kernel Library for Deep Neural Networks (Intel(R) MKL-DNN)  0.13
Performance library for Deep Learning
Modules
Primitive operations

Modules

 Attributes
 An extension for controlling primitive behavior.
 
 Batch Normalization
 A primitive to perform batch normalization

\[dst[n][c][h][w] = \gamma[c] \frac{src[n][c][h][w] - \mu[c]} {\sqrt{\sigma[c] + eps}} + \beta[c],\]

.

 
 Common primitive operations
 
 Concat
 A primitive to concatenate data by arbitrary dimension.
 
 Convolution
 A primitive to compute convolution using different algorithms.
 
 Convolution followed by ReLU
 A merged primitive to compute a convolution followed by relu.
 
 Eltwise
 A primitive to compute element wise operations like parametric rectifier linear unit (ReLU).
 
 Inner product
 A primitive to compute an inner product.
 
 LRN
 A primitive to perform local response normalization (LRN) across or within channels.
 
 Memory
 A primitive to describe data.
 
 Pooling
 A primitive to perform max, min, or average pooling.
 
 ReLU (deprecated, use Eltwise instead)
 A primitive to compute a parametric rectifier linear unit (ReLU).
 
 Reorder
 A primitive to copy data between memory formats.
 
 Softmax
 A primitive to perform softmax.
 
 Sum
 A primitive to sum data.
 

Detailed Description