FilesLast commitLast update
23 days ago
23 days ago
23 days ago
13 days ago
README

RefIdentity

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: The function is the same as Identity—the output tensor y has the same shape, data type, and content as the input tensor x. The difference is that the input of RefIdentity is a reference tensor that can be overwritten by ref-type operators, and is used for reference (ref) type data.

Parameter Description

Parameter Name Input/Output/Attribute Description Data Type Data Format
x Input Input tensor. float16, float, double, int8, int16, int32, int64, uint8, uint16, uint32, uint64, bool, bfloat16 ND
y Output Tensor that has the same shape, data type, and content as the input x. Same as x ND

Constraint Description

  • The data type and shape of the input x and the output y are the same.
  • Dynamic shape and dynamic rank are supported.

Invocation Description

The RefIdentity operator has no prototype (REG_OP) or inference implementation yet, and no invocation sample is provided. If you want to contribute a complete implementation of this operator, refer to Contribution Process.