tensorclass

class tensordict.tensorclass(autocast: bool = False)

一个用于创建 tensorclass 类别的装饰器。

tensorclass 类别是专门的 dataclass 实例，可以开箱即用地执行一些预定义的张量操作，例如索引、项目赋值、重塑、转换为设备或存储以及许多其他操作。

示例

>>> from tensordict import tensorclass
>>> import torch
>>> from typing import Optional
>>>
>>> @tensorclass
... class MyData:
...     X: torch.Tensor
...     y: torch.Tensor
...     z: str
...     def expand_and_mask(self):
...         X = self.X.unsqueeze(-1).expand_as(self.y)
...         X = X[self.y]
...         return X
...
>>> data = MyData(
...     X=torch.ones(3, 4, 1),
...     y=torch.zeros(3, 4, 2, 2, dtype=torch.bool),
...     z="test"
...     batch_size=[3, 4])
>>> print(data)
MyData(
    X=Tensor(torch.Size([3, 4, 1]), dtype=torch.float32),
    y=Tensor(torch.Size([3, 4, 2, 2]), dtype=torch.bool),
    z="test"
    batch_size=[3, 4],
    device=None,
    is_shared=False)
>>> print(data.expand_and_mask())
tensor([])

也可以将 tensorclasses 实例嵌套在彼此之中

示例： >>> from tensordict import tensorclass >>> import torch >>> from typing import Optional >>> >>> @tensorclass … class NestingMyData: … nested: MyData … >>> nesting_data = NestingMyData(nested=data, batch_size=[3, 4]) >>> # 尽管数据存储为 TensorDict，但类型提示帮助我们 >>> # 将数据适当地转换为正确的类型 >>> assert isinstance(nesting_data.nested, type(data))