Andrew James
Andrew James
cc @rec might be interesting to see how this diff looks wrt the type completeness delta summary.
A note for reviewers: the failure in xla_op1 shard appears to be unrelated, the tests for the new feature are in xla_op3
+1 to adding CODEOWNER entries. > 1. Add a comment saying "Do not implement checks that can be enabled by check_sparse_tensor_invariants" to sparse tensor constructor implementations where one should want...
I can take a look at this it is the same issue I was dealing with.