pytorch

57b23c61 - In the CUDA implementation of erfinv, erfinv() should be used for double (#25337)

In the CUDA implementation of erfinv, erfinv() should be used for double (#25337) Summary: This best preserves accuracy, while erfinvf() should be used for half and float. This is also consistent with the implementation before the migration: https://github.com/pytorch/pytorch/issues/24943 Pull Request resolved: https://github.com/pytorch/pytorch/pull/25337 Differential Revision: D17102333 Pulled By: zou3519 fbshipit-source-id: 5178cff534cf5f10d86ab04d4b6c1779ffedf49e