make_composite_from_td

torchrl.envs.make_composite_from_td(data, *, unsqueeze_null_shapes: bool = True, dynamic_shape: bool = False)

從 tensordict 建立 Composite 例項，假設所有值都是無界的。

引數:

data (tensordict.TensorDict) – 要對映到 Composite 的 tensordict。

關鍵字引數:

• unsqueeze_null_shapes (bool, optional) – 如果為 True，則每個空形狀將被 unsqueeze 為 (1,)。預設為 True。

• dynamic_shape (bool, optional) – 如果為 True，則所有張量都將被假定沿著最後一個維度具有動態形狀。預設為 False。

示例

>>> from tensordict import TensorDict
>>> data = TensorDict({
...     "obs": torch.randn(3),
...     "action": torch.zeros(2, dtype=torch.int),
...     "next": {"obs": torch.randn(3), "reward": torch.randn(1)}
... }, [])
>>> spec = make_composite_from_td(data)
>>> print(spec)
Composite(
    obs: UnboundedContinuous(
         shape=torch.Size([3]), space=None, device=cpu, dtype=torch.float32, domain=continuous),
    action: UnboundedContinuous(
         shape=torch.Size([2]), space=None, device=cpu, dtype=torch.int32, domain=continuous),
    next: Composite(
        obs: UnboundedContinuous(
             shape=torch.Size([3]), space=None, device=cpu, dtype=torch.float32, domain=continuous),
        reward: UnboundedContinuous(
             shape=torch.Size([1]), space=ContinuousBox(low=Tensor(shape=torch.Size([]), device=cpu, dtype=torch.float32, contiguous=True), high=Tensor(shape=torch.Size([]), device=cpu, dtype=torch.float32, contiguous=True)), device=cpu, dtype=torch.float32, domain=continuous), device=cpu, shape=torch.Size([])), device=cpu, shape=torch.Size([]))
>>> assert (spec.zero() == data.zero_()).all()