You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Jul 1, 2023. It is now read-only.
Currently matmul invokes the batch matmul op whenever the input matrices have rank higher than 2. However, batch matmul takes adjoint flags as input instead of transpose flags. This means that when we support complex number types we should be passing conjugate(lhs) instead of lhs to the batch matmul op if transpose for lhs is true.
The text was updated successfully, but these errors were encountered:
Sign up for freeto subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Imported from JIRA https://bugs.swift.org/browse/TF-629
Currently matmul invokes the batch matmul op whenever the input matrices have rank higher than 2. However, batch matmul takes
adjoint
flags as input instead oftranspose
flags. This means that when we support complex number types we should be passingconjugate(lhs)
instead oflhs
to the batch matmul op iftranspose
forlhs
istrue
.The text was updated successfully, but these errors were encountered: