reference/generated/torchrl.render.load_checkpoint

Run in Google Colab

Colab

Download Notebook

Notebook

View on GitHub

GitHub

load_checkpoint

torchrl.render.load_checkpoint(path: str | Path, map_location: str | device = 'cpu', *, weights_only: bool | None = None) → Any

Loads a local PyTorch checkpoint.

Parameters:

• path – Local checkpoint path.

• map_location – Device mapping passed to torch.load().

• weights_only – Whether payloads are restricted to safe weight-only types. Unified checkpoints default to True. Legacy payloads retain their historical False default for compatibility.

Returns:

The checkpoint payload.