FilesLast commitLast update
12 days ago
2 months ago
2 months ago
7 days ago
7 days ago
2 months ago
7 days ago
20 days ago
1 month ago
5 days ago
README

StridedSliceGrad

English | 简体中文

Product Support Status

Product Supported
Ascend 950PR&950DT products √
Atlas A3 products √
Atlas A2 products √
Atlas 200I/500 A2 inference products ×
Atlas inference products √
Atlas training products √

Function Description

  • Operator function: maps the gradient of a sub-tensor back to the corresponding position in the original tensor.

Parameter Description

Parameter Name Input/Output/Attribute Description Data Type Data Format
shape Input Input tensor INT32, INT64 ND
begin Input Start position INT32, INT64 ND
end Input End position INT32, INT64 ND
strides Input Step INT32, INT64 ND
dy Input Gradient FLOAT16, BFLOAT16, FLOAT, DOUBLE, INT8, INT16, INT32, INT64, UINT8, UINT16, UINT32, UINT64, COMPLEX32, COMPLEX64 ND
begin_mask Input attribute Input format INT -
end_mask Input attribute Input format INT -
ellipsis_mask Input attribute Input format INT -
new_axis_mask Input attribute Input format INT -
shrink_axis_mask Input attribute Input format INT -
output Output Padded gradient tensor FLOAT16, BFLOAT16, FLOAT, DOUBLE, INT8, INT16, INT32, INT64, UINT8, UINT16, UINT32, UINT64, COMPLEX32, COMPLEX64 ND

Constraint Description

None

Invocation Description

Invocation Method Invocation Sample Description
Graph-mode invocation test_geir_strided_slice_grad Invokes the StridedSliceGrad operator by building a graph through the operator IR.
aclnn API test_aclnn_strided_slice_grad Invokes the StridedSliceGrad operator through the aclnnStridedSliceGrad API.