pytorch

e6befbe8 - Add flag to optionally average output attention weights across heads (#70055)

Add flag to optionally average output attention weights across heads (#70055) Summary: Fixes https://github.com/pytorch/pytorch/issues/47583 Pull Request resolved: https://github.com/pytorch/pytorch/pull/70055 Reviewed By: bhosmer Differential Revision: D33457866 Pulled By: jbschlosser fbshipit-source-id: 17746b3668b0148c1e1ed8333227b7c42f1e3bf5