pypto.sum

Applicable Products

Ascend 950PR/Ascend 950DT: Supported
Atlas A3 training products/Atlas A3 inference products: Supported
Atlas A2 training products/Atlas A2 inference products: Supported

Description

Accumulates data of a multi-dimensional tensor along a specified dimension.

Define the specified computation dimension as the R-axis (Reduce axis) and the non-specified dimension as the A-axis (Normal axis). As shown in the figure below, for a two-dimensional matrix of shape (2, 3), when the sum is computed along the first dimension, the output is [5, 7, 9]; when computed along the second dimension, the output is [6, 15].

Figure 1 Example of sum computing along the first dimension

Figure 2 Example of sum computing along the last dimension

Prototype

sum(input: Tensor, dim: int, keepdim: bool = False) -> Tensor

Parameters

Parameter	Input/Output	Description
input	Input	Source operand. Supported type: Tensor. Supported data types of Tensor: DT_FP32, DT_BF16, DT_INT32, and DT_INT16. Empty tensors are not supported. The shape supports only 1 to 4 dimensions. The shape size must not exceed 2147483647 (that is, INT32_MAX).
dim	Input	Dimension of reduction. Any single axis is supported.
keepdim	Input	Whether to retain the reduced dimension after reduction. The default value is False.

Return Value

Returns the output tensor. The shape of the output tensor depends on the keepdim parameter.

If the keepdim parameter is True, the reduced dimension is retained after the reduction operation. The shape of the output tensor is the same as that of the input tensor in all dimensions except the dimension specified by dim, where the size is 1.

If the keepdim parameter is False (default), the reduced dimension is removed from the output tensor, while the corresponding dimension in TileShape remains unchanged. Therefore, it is recommended to reset TileShape before calling other operations.

Constraints

The TileShape size must not exceed 64 KB.
Tensor inputs do not support the TileOpFormat.TILEOP_NZ format.

Examples

TileShape Setting Example

Note: Before calling this operation API, set the TileShape through set_vec_tile_shapes.

The dimensions of the TileShape must match that of the input shape.

Example 1: If the input shape is [m, n] and the output is [m, 1], set the TileShape to [m1, n1], where m1 and n1 are used to tile the m and n axes, respectively.

pypto.set_vec_tile_shapes(4, 16)

API Call Example

x = pypto.tensor([2, 3], pypto.DT_FP32)
y = pypto.sum(x, -1, True)

The results are as follows:

Input data x: [[1.0 2.0 3.0],
             [1.0 2.0 3.0]]
Output data y: [[6.0],
             [6.0]]