Describe the bug

The tutorial at (https://docs.pytorch.org/rl/stable/tutorials/pendulum.html#reproducible-experiments-seeding) has the example

def _set_seed(self, seed: Optional[int]) -> None:
    rng = torch.manual_seed(seed)
    self.rng = rng

However, torch.manual does not accept None.

Additional context

As a side note, I wonder what the recommended way to handle randomness in environments is. My own preference would be to create a torch.Generator in _set_seed and then use it for all sources of randomness in _step. To my understanding, calling torch.manual_seed affects the global RNG of pytorch, not only the environment that we call it in.

Checklist

• I have checked that there is no similar issue in the repo (required)
• I have read the documentation (required)
• I have provided a minimal working example to reproduce the bug (required)