hsplit¶

paddle. hsplit ( x, num_or_indices, name=None ) [source]

Split the input tensor into multiple sub-Tensors along the horizontal axis, which is equivalent to paddle.tensor_split with axis=1 when x ‘s dimension is larger than 1, or equivalent to paddle.tensor_split with axis=0 when x ‘s dimension is 1.

Parameters

• x (Tensor) – A Tensor whose dimension must be greater than 0. The data type is bool, bfloat16, float16, float32, float64, uint8, int32 or int64.

• num_or_indices (int|list|tuple) – If num_or_indices is an int n, x is split into n sections. If num_or_indices is a list or tuple of integer indices, x is split at each of the indices.

• name (str, optional) – The default value is None. Normally there is no need for user to set this property. For more information, please refer to Name .

Returns

list[Tensor], The list of segmented Tensors.

Examples

>>> import paddle

>>> # x is a Tensor of shape [8]
>>> x = paddle.rand([8])
>>> out0, out1 = paddle.hsplit(x, num_or_indices=2)
>>> print(out0.shape)
[4]
>>> print(out1.shape)
[4]

>>> # x is a Tensor of shape [7, 8]
>>> x = paddle.rand([7, 8])
>>> out0, out1 = paddle.hsplit(x, num_or_indices=2)
>>> print(out0.shape)
[7, 4]
>>> print(out1.shape)
[7, 4]

>>> out0, out1, out2 = paddle.hsplit(x, num_or_indices=[1, 4])
>>> print(out0.shape)
[7, 1]
>>> print(out1.shape)
[7, 3]
>>> print(out2.shape)
[7, 4]