xla
xla copied to clipboard
EmbeddingDenseBackward: Remove `padding_idx` cast to `double`
Fixes #9392
Thanks for the change! I am going to run the CI and see if it breaks anything
Can you also please add a test for embedding backward similar to this one https://github.com/pytorch/xla/blob/193f7049ef56639122b11c03add44512aee9f66a/test/test_operations.py#L738
Done.
Please, rerun the CI
Seems like everything has passed. The failing job is broken in master, if I understand correctly.
The E2E tests have been failing for some time