Commit 14ea6d5a authored by A. Unique TensorFlower's avatar A. Unique TensorFlower Committed by TensorFlower Gardener
Browse files

Disable parallelizing over both batch and inner matrix dimensions in CPU... 

Disable parallelizing over both batch and inner matrix dimensions in CPU BatchMatmul, since this can lead to a deadlock in the Eigen multi-threaded contraction code.
Tuned the heuristic selecting between parallelizing over batch or inner dimensions.

PiperOrigin-RevId: 170861489
parent 6425dbd1
Show whitespace changes
Inline Side-by-side
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Please to comment