Commit 1bd82e09 authored by Tim Shen's avatar Tim Shen Committed by TensorFlower Gardener
Browse files

Simplify some cuda_dnn logic: 

* De-duplicate the accumulator type logic. Currently we have two ways to
specify them, "AccumulatorType" as template argument, or GetConvComputeType.
Removed GetConvComputeType.
* Simplified GetCudnnDataType.

PiperOrigin-RevId: 222144108
parent 0d2344c1
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