Update torch.nn.quantizable.MultiHeadAttention docstring (#53106)
Summary:
Apply the same fix as PR https://github.com/pytorch/pytorch/pull/49950
Pull Request resolved: https://github.com/pytorch/pytorch/pull/53106
Reviewed By: zou3519
Differential Revision: D26752234
Pulled By: albanD
fbshipit-source-id: 5c924319b8365da4d3d2ba2206e2586e23e718f0
