TensorUInt32Bit.CreateFromBuffer(Int64[], IBuffer) 方法

定义

创建具有给定 形状 的 32 位无符号整数张量对象,并使用 缓冲区中的基础缓冲区 进行后续计算。

public:
 static TensorUInt32Bit ^ CreateFromBuffer(Platform::Array <long long> ^ shape, IBuffer ^ buffer);
 static TensorUInt32Bit CreateFromBuffer(winrt::array_view <long> const& shape, IBuffer const& buffer);
public static TensorUInt32Bit CreateFromBuffer(long[] shape, IBuffer buffer);
function createFromBuffer(shape, buffer)
Public Shared Function CreateFromBuffer (shape As Long(), buffer As IBuffer) As TensorUInt32Bit

参数

shape

Int64[]

long long[]

long[]

缓冲区的大小。

buffer
IBuffer

在评估期间使用的缓冲区。

返回

由缓冲区引用提供的 IBuffer缓冲区支持的 32 位无符号整数张量对象。

Windows 要求

设备系列
Windows 10, version 1903 (在 10.0.18362.0 中引入)
API contract
Windows.AI.MachineLearning.MachineLearningContract (在 v2.0 中引入)

注解

创建此对象期间不会执行任何副本,并且引擎直接使用 缓冲区中的基础缓冲区

在非 CPU 设备上运行时,将复制到视频内存。

适用于