Skip to content

explain() likely affects PyTorch autograd operations #14

Description

@parrt

PyTorch: use with torch.no_grad() to prevent messing up autograd computations.

Metadata

Metadata

Assignees

No one assigned

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions