Maximum
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: returns a new tensor composed of the maximum values of the two input tensors.
-
Computation formula:
out[i]=max(x1[i],x2[i])out[i] = max(x_1[i], x_2[i])
Parameter Description
| Parameter Name | Input/Output/Attribute | Description | Data Type | Data Format |
|---|---|---|---|---|
| x1 | Input | Input parameter for the Maximum computation, which is x1 in the formula. | FLOAT, FLOAT16, BFLOAT16, DOUBLE, INT8, UINT8, INT32, INT64 | ND |
| x2 | Input | Input parameter for the Maximum computation, which is x2 in the formula. | FLOAT, FLOAT16, BFLOAT16, DOUBLE, INT8, UINT8, INT32, INT64 | ND |
| y | Output | Element-wise maximum result of the tensors, which is out in the formula. | FLOAT, FLOAT16, BFLOAT16, DOUBLE, INT8, UINT8, INT32, INT64 | ND |
Constraint Description
None
Invocation Description
| Invocation Method | Invocation Sample | Description |
|---|---|---|
| aclnn invocation | test_aclnn_maximum | Invokes the aclnnMaximum operator through the aclnnMaximum API. |
| aclnn invocation | test_aclnn_maxn | Invokes the aclnnMaxN operator through the aclnnMaxN API. |