unbind¶

paddle. unbind ( input, axis=0 ) [source]

Removes a tensor dimension, then split the input tensor into multiple sub-Tensors.

Parameters

• input (Tensor) – The input variable which is an N-D Tensor, data type being bool, float16, float32, float64, int32 or int64.

• axis (int32|int64, optional) – A scalar with type int32|int64 shape [1]. The dimension along which to unbind. If \(axis < 0\), the dimension to unbind along is \(rank(input) + axis\). Default is 0.

Returns

list(Tensor), The list of segmented Tensor variables.

Example

import paddle

# input is a Tensor which shape is [3, 4, 5]
input = paddle.rand([3, 4, 5])

[x0, x1, x2] = paddle.unbind(input, axis=0)
# x0.shape [4, 5]
# x1.shape [4, 5]
# x2.shape [4, 5]

[x0, x1, x2, x3] = paddle.unbind(input, axis=1)
# x0.shape [3, 5]
# x1.shape [3, 5]
# x2.shape [3, 5]
# x3.shape [3, 5]