Layout is to describe how data is organized within an N-dimention tensor. It is composed of upper cases, lower cases and numbers, where upper case indicates a primal axis and the corresponding lower case with factor size indicates the subordinate axis. For example, NCHW16c can describe a 5-D tensor of [batch_size, channel, height, width, channel_block]. Here subordinate axis channel_block=16 is the factor size of the primal axis C (channel). Layout for scalar is defined, while both its name and axes have size 0. More...

#include <data_layout.h>

Inheritance diagram for tvm::tir::LayoutNode:

Collaboration diagram for tvm::tir::LayoutNode:

Public Member Functions
	TVM_FFI_DECLARE_OBJECT_INFO_FINAL ("tir.Layout", LayoutNode, Object)

Static Public Member Functions
static void	RegisterReflection ()

Public Attributes
ffi::String	name
	string representation of layout, "" for scalar. More...

ffi::Array< tir::IterVar >	axes
	specify each axis of the layout, in which the variable name is the name of the axis. The IterVar's extent indicates the size of the axis, it is a variable for a primal axis, but a constant for a subordinate axis. Empty for scalar's layout. More...

Detailed Description

Layout is to describe how data is organized within an N-dimention tensor. It is composed of upper cases, lower cases and numbers, where upper case indicates a primal axis and the corresponding lower case with factor size indicates the subordinate axis. For example, NCHW16c can describe a 5-D tensor of [batch_size, channel, height, width, channel_block]. Here subordinate axis channel_block=16 is the factor size of the primal axis C (channel). Layout for scalar is defined, while both its name and axes have size 0.

Member Function Documentation

◆ RegisterReflection()

static void tvm::tir::LayoutNode::RegisterReflection ( )

inlinestatic

◆ TVM_FFI_DECLARE_OBJECT_INFO_FINAL()

tvm::tir::LayoutNode::TVM_FFI_DECLARE_OBJECT_INFO_FINAL	(	"tir.Layout"	,
		LayoutNode	,
		Object
	)

Member Data Documentation

◆ axes

ffi::Array<tir::IterVar> tvm::tir::LayoutNode::axes

specify each axis of the layout, in which the variable name is the name of the axis. The IterVar's extent indicates the size of the axis, it is a variable for a primal axis, but a constant for a subordinate axis. Empty for scalar's layout.

◆ name

ffi::String tvm::tir::LayoutNode::name

string representation of layout, "" for scalar.

The documentation for this class was generated from the following file:

include/tvm/tir/data_layout.h

Public Member Functions

Static Public Member Functions