tensordict.nn.make_tensordict

tensordict.nn.make_tensordict(input_dict: Optional[dict[str, Union[torch.Tensor, tensordict._tensorcollection.TensorCollection]]] = None, batch_size: Optional[Union[Sequence[int], Size, int]] = None, device: Optional[Union[device, str, int]] = None, auto_batch_size: Optional[bool] = None, **kwargs: Union[Tensor, TensorCollection]) → TensorDict

从关键字参数或输入字典返回一个创建的 TensorDict。

If batch_size is not specified, returns the maximum batch size possible.

This function works on nested dictionaries too, or can be used to determine the batch-size of a nested tensordict.

参数:

• input_dict (dictionary, optional) – a dictionary to use as a data source (nested keys compatible).

• **kwargs (TensorDict 或 torch.Tensor) – 关键字参数作为数据源（与嵌套键不兼容）。

• batch_size (iterable of int, optional) – a batch size for the tensordict.

• device (torch.device 或 兼容类型, optional) – TensorDict 的设备。

• auto_batch_size (bool, 可选) – 如果设置为 True，则将自动计算 batch size。默认为 False。

示例

>>> input_dict = {"a": torch.randn(3, 4), "b": torch.randn(3)}
>>> print(make_tensordict(input_dict))
TensorDict(
    fields={
        a: Tensor(shape=torch.Size([3, 4]), device=cpu, dtype=torch.float32, is_shared=False),
        b: Tensor(shape=torch.Size([3]), device=cpu, dtype=torch.float32, is_shared=False)},
    batch_size=torch.Size([3]),
    device=None,
    is_shared=False)
>>> # alternatively
>>> td = make_tensordict(**input_dict)
>>> # nested dict: the nested TensorDict can have a different batch-size
>>> # as long as its leading dims match.
>>> input_dict = {"a": torch.randn(3), "b": {"c": torch.randn(3, 4)}}
>>> print(make_tensordict(input_dict))
TensorDict(
    fields={
        a: Tensor(shape=torch.Size([3]), device=cpu, dtype=torch.float32, is_shared=False),
        b: TensorDict(
            fields={
                c: Tensor(shape=torch.Size([3, 4]), device=cpu, dtype=torch.float32, is_shared=False)},
            batch_size=torch.Size([3, 4]),
            device=None,
            is_shared=False)},
    batch_size=torch.Size([3]),
    device=None,
    is_shared=False)
>>> # we can also use this to work out the batch sie of a tensordict
>>> input_td = TensorDict({"a": torch.randn(3), "b": {"c": torch.randn(3, 4)}}, [])
>>> print(make_tensordict(input_td))
TensorDict(
    fields={
        a: Tensor(shape=torch.Size([3]), device=cpu, dtype=torch.float32, is_shared=False),
        b: TensorDict(
            fields={
                c: Tensor(shape=torch.Size([3, 4]), device=cpu, dtype=torch.float32, is_shared=False)},
            batch_size=torch.Size([3, 4]),
            device=None,
            is_shared=False)},
    batch_size=torch.Size([3]),
    device=None,
    is_shared=False)