Pack
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: packs a list of tensors in values into a tensor that is one dimension higher than each tensor in values, by packing along the axis dimension. Given a list of length N of tensors with shape (A, B, C); if axis == 0, the output tensor will have shape (N, A, B, C).
Parameter Description
| Parameter Name | Input/Output/Attribute | Description | Data Type | Data Format |
|---|---|---|---|---|
| x | Input | List of tensors to be concatenated. | COMPLEX128, COMPLEX64, BOOL, DOUBLE, FLOAT32, FLOAT16, BFLOAT16, INT16, INT32, INT64, INT8, UINT8, UINT16, UINT32, UINT64, HIFLOAT8, FLOAT8_E5M2, FLOAT8_E4M3FN | ND |
| axis | Optional attribute | Specifies the dimension along which tensors are packed. | INT | - |
| N | Optional attribute | Specifies the number of tensors to be packed. | INT | - |
| y | Output | Output result. | COMPLEX128, COMPLEX64, BOOL, DOUBLE, FLOAT32, FLOAT16, BFLOAT16, INT16, INT32, INT64, INT8, UINT8, UINT16, UINT32, UINT64, HIFLOAT8, FLOAT8_E5M2, FLOAT8_E4M3FN | ND |
- Atlas training products, Atlas inference products, and Atlas 200I/500 A2 inference products: BFLOAT16, HIFLOAT8, FLOAT8_E5M2, and FLOAT8_E4M3FN are not supported.
- Atlas A2 products and Atlas A3 products: HIFLOAT8, FLOAT8_E5M2, and FLOAT8_E4M3FN are not supported.
Constraint Description
None.
Invocation Description
| Invocation Method | Sample Code | Description |
|---|---|---|
| aclnn API | test_aclnn_stack | Invokes the Pack operator through the aclnnStack API. |