CreateTensor

Applicability

Product	Supported or Not
Atlas A3 training products/Atlas A3 inference products	√
Atlas A2 training products/Atlas A2 inference products	√
Atlas 200I/500 A2 inference products	√
Atlas inference products	√
Atlas training products	√

Function Usage

Creates a tensor of the specified data type and tensor shape.

Prototype

template <typename T>
std::unique_ptr<Tensor> CreateTensor(const std::vector<T> &value, const std::vector<int64_t> &dims, DataType dt, Format format = FORMAT_ND);

template <typename T>
std::unique_ptr<Tensor> CreateTensor(const T *value, const int64_t *dims, int64_t dim_num, DataType dt, Format format = FORMAT_ND);

Parameters

Parameter	Input/Output	Description
T	Input	Tensor data type.
value	Input	Pointer to the tensor data.
dims	Input	Pointer to the tensor dimension array.
dim_num	Input	Number of tensor dimensions.
dt	Input	Tensor data type.
format	Input	Tensor format. The default value is FORMAT_ND.

Returns

Parameter	Type	Description
-	std::unique_ptr<Tensor>	Smart pointer to the created tensor. If the operation fails, nullptr is returned.

Constraints

None

Examples

Template 1:

EsGraphBuilder builder("graph_name");
std::vector<float> data = {1.0f, 2.0f, 3.0f};
std::vector<int64_t> dims = {3};
auto tensor = builder.CreateTensor<float>(data, dims, ge::DT_FLOAT);

Template 2:

std::vector<float> data = {1.0f, 2.0f, 3.0f};
std::vector<int64_t> dims = {3};
auto tensor = ge::es::CreateTensor<float>(data.data(), dims.data(), dims.size(), ge::DT_FLOAT);

Parent topic: EsGraphBuilder