PinMemoryTransform¶
- class torchrl.envs.transforms.PinMemoryTransform[源代码]¶
调用 tensordict 的 pin_memory,以便在 CUDA 设备上写入。
- forward(next_tensordict: TensorDictBase) TensorDictBase ¶
读取输入 tensordict,并对选定的键应用转换。
默认情况下,此方法
直接调用
_apply_transform()
。不调用
_step()
或_call()
。
此方法不会在任何时候在 env.step 中调用。但是,它会在
sample()
中调用。注意
forward
还使用dispatch
与常规关键字参数一起工作,将参数名称转换为键。示例
>>> class TransformThatMeasuresBytes(Transform): ... '''Measures the number of bytes in the tensordict, and writes it under `"bytes"`.''' ... def __init__(self): ... super().__init__(in_keys=[], out_keys=["bytes"]) ... ... def forward(self, tensordict: TensorDictBase) -> TensorDictBase: ... bytes_in_td = tensordict.bytes() ... tensordict["bytes"] = bytes ... return tensordict >>> t = TransformThatMeasuresBytes() >>> env = env.append_transform(t) # works within envs >>> t(TensorDict(a=0)) # Works offline too.