GetDynamicInputTensorRange
Description
Obtains the pointer to a dynamic input tensor range based on the input index in the operator prototype definition.
Prototype
using TensorRange = Range<Tensor>
const TensorRange *GetDynamicInputTensorRange(const size_t ir_index, const size_t relative_index) const
Parameters
Parameter |
Input/Output |
Description |
|---|---|---|
ir_index |
Input |
Input index in the operator IR prototype definition, starting from 0. |
relative_index |
Input |
Relative index after the input is instantiated. For example, if three inputs are instantiated for a DYNAMIC_INPUT, the value range of relative_index is [0, 2]. |
Returns
Pointer to the tensor range. If ir_index or relative_index is invalid, a null pointer is returned.
Restrictions
If the input is not set to data dependency, when this API is called to obtain a tensor range, only the correct shape, format, and datatype information can be obtained from the tensor, and the actual tensor data address (the obtained address is nullptr) cannot be obtained.
Example
const auto infer_shape_range_func = [](gert::InferShapeRangeContext *context) -> graphStatus {
auto input_shape_range = context->GetDynamicInputTensorRange(0U, 0U);
auto output_shape_range = context->GetOutputShapeRange(0U);
*output_shape_range-->GetMin() = input_shape_range->GetMin()->GetStorageShape();
*output_shape_range-->GetMax() = input_shape_range->GetMax()->GetStorageShape();
return GRAPH_SUCCESS;
};