experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[Slice] Reduce variance in benchmark.

Browse Source 
This change switches the various Slice benchmarks to run using tf.data's
single-threaded executor. This eliminates noise from thread scheduling, which
is often larger than the actual execution time for some of the benchmarks.

PiperOrigin-RevId: 298638978
Change-Id: Id9472df0eea4a2ec56a3b88154b8069e92bc11dc
This commit is contained in:
Derek Murray 2020-03-03 10:36:55 -08:00 committed by TensorFlower Gardener
parent e4dc8c0ded
commit 8da96e9b52
2 changed files with 7 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
tensorflow/core/kernels/BUILD
View File

@ -2280,6 +2280,7 @@ tf_cc_test(
"//tensorflow/core:test",
"//tensorflow/core:test_main",
"//tensorflow/core:testlib",
"//tensorflow/core/kernels/data:single_threaded_executor",
],
)

7
tensorflow/core/kernels/slice_op_test.cc
View File

@ -22,6 +22,7 @@ limitations under the License.
#include "tensorflow/core/framework/tensor.h"
#include "tensorflow/core/framework/types.h"
#include "tensorflow/core/framework/types.pb.h"
#include "tensorflow/core/graph/algorithm.h"
#include "tensorflow/core/graph/node_builder.h"
#include "tensorflow/core/graph/testlib.h"
#include "tensorflow/core/kernels/ops_testutil.h"
@ -62,10 +63,14 @@ static void SliceHelper(int iters, int size) {
.Input(test::graph::Constant(g, sizes))
.Attr("T", dt)
.Finalize(g, &node));
FixupSourceAndSinkEdges(g);
testing::BytesProcessed(static_cast<int64>(iters) * kDim * size * sizeof(T));
testing::StartTiming();
test::Benchmark("cpu", g).Run(iters);
test::Benchmark("cpu", g, nullptr, nullptr, nullptr,
"SINGLE_THREADED_EXECUTOR")
.Run(iters);
testing::UseRealTime();
}