# Constants
Like ReluX; but passes all values in the range [-X,X].
No description provided by the author
Rectified linear activation: f(x) = x < 0 ? 0 : x.
Rectified linear activation; where upper maximum is 6.0.
Rectified linear activation; where upper maximum specified by BatchDescriptor::value_max().
No description provided by the author
No description provided by the author
No description provided by the author
The GPU may operate 4x4 matrix FMA.
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
Naming convention: Y <-> row or height X <-> column or width Batch <-> batch, or N Depth <-> feature, or channel TODO(timshen): turn them into cuDNN names, e.g.
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
Naming convention: Y <-> row or height X <-> column or width Output <-> output feature, or N Input <-> input feature, or N TODO(timshen): turn them into cuDNN names, e.g.
No description provided by the author
No description provided by the author
No description provided by the author
# Variables
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
No description provided by the author
# Structs
Generic algorithm representation.
Convolution-specific parameters.
Generic tensor representation.
No description provided by the author
No description provided by the author
# Type aliases
Describes a kind of non-linearity (threshold-like mathematical function).
No description provided by the author
No description provided by the author
Describe the math definition for the conv op.
Describes how a convolution input or output layer's data is formatted.
Specifies the data type used by an operation.
Describes how a convolution filter is laid out in the memory.