torch.nn.utils.rnn.unpack_sequence

torch.nn.utils.rnn.unpack_sequence(packed_sequences)

将 PackedSequence 解包成一个包含可变长度张量的列表。

packed_sequences 应为 PackedSequence 对象。

示例

>>> from torch.nn.utils.rnn import pack_sequence, unpack_sequence
>>> a = torch.tensor([1, 2, 3])
>>> b = torch.tensor([4, 5])
>>> c = torch.tensor([6])
>>> sequences = [a, b, c]
>>> print(sequences)
[tensor([1, 2, 3]), tensor([4, 5]), tensor([6])]
>>> packed_sequences = pack_sequence(sequences)
>>> print(packed_sequences)
PackedSequence(data=tensor([1, 4, 6, 2, 5, 3]), batch_sizes=tensor([3, 2, 1]), sorted_indices=None, unsorted_indices=None)
>>> unpacked_sequences = unpack_sequence(packed_sequences)
>>> print(unpacked_sequences)
[tensor([1, 2, 3]), tensor([4, 5]), tensor([6])]

参数

packed_sequences (PackedSequence) – PackedSequence 对象。

返回值

一个 Tensor 对象列表

返回类型

List[Tensor]