aclCreateFloatArray

Function Usage

Creates an aclFloatArray object as the input parameter for single-operator API execution.

aclFloatArray is a framework-defined array structure used to manage and store floating-point data. You can directly use it without paying attention to its internal implementation.

Prototype

aclFloatArray *aclCreateFloatArray(const float *value, uint64_t size)

Parameters

Parameter

Input/Output

Description

value

Input

Float pointer on the host. Its value will be copied to the aclFloatArray.

size

Input

Length of a floating-point array. The value is a positive integer.

Returns

Created aclFloatArray on success; else, nullptr.

Constraints

  • This API must be used together with aclDestroyFloatArray. They are used to create and destroy the aclFloatArray, respectively.
  • The size of the aclFloatArray can be obtained by calling aclGetFloatArraySize.

Examples

The following code examples are for reference only and are not intended for direct copying and execution:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
//Create an aclFloatArray.
std::vector<float> scalesData = {1.0, 1.0, 2.0, 2.0};
aclFloatArray *scales = aclCreateFloatArray(scalesData.data(),scalesData.size());
...
// Use the aclFloatArray as the input parameter for single-operator API execution.
auto ret = aclxxXxxGetWorkspaceSize(srcTensor, scales, ..., outTensor, ..., &workspaceSize, &executor);
ret = aclxxXxx(...);
...
//Destroy the aclFloatArray.
ret = aclDestroyFloatArray(scales);