diff options
author | Peter Hawkins <phawkins@google.com> | 2018-09-26 13:48:21 -0700 |
---|---|---|
committer | TensorFlower Gardener <gardener@tensorflow.org> | 2018-09-26 13:51:50 -0700 |
commit | 1736e0bbbfdeeba178dff37c970b5a0180ee013f (patch) | |
tree | 390c309b5997a752644d2c50bb4ee5bf8fc1654d /tensorflow/core/BUILD | |
parent | 652ce1aaefdadd04a9905a0788ab26c6fff93658 (diff) |
[TF] Add new internal ops _VarHandlesOp and _ReadVariablesOp.
The purpose of these ops is to fix a latency problem observed for an inference benchmark. Often a inference step starts by reading the value of many (hundreds) of weights. For a resource variable, this requires a VarHandleOp and a ReadVariableOp per variable. Running hundreds of trivial ops can add hundreds of microseconds of latency to the critical path of an inference step. The inter-op latency of the executor can be hundreds of nanoseconds, which rapidly adds up.
This change introduces two fused ops _VarHandlesOp and _ReadVariablesOp that allow us to read many variables in a pair of larger ops, rather than many tiny ops.
PiperOrigin-RevId: 214662338
Diffstat (limited to 'tensorflow/core/BUILD')
-rw-r--r-- | tensorflow/core/BUILD | 9 |
1 files changed, 8 insertions, 1 deletions
diff --git a/tensorflow/core/BUILD b/tensorflow/core/BUILD index bc0bfb793c..d85cb379bb 100644 --- a/tensorflow/core/BUILD +++ b/tensorflow/core/BUILD @@ -1057,7 +1057,6 @@ tf_gen_op_libs( "random_grad", "random_ops", "remote_fused_graph_ops", - "resource_variable_ops", "rpc_ops", "scoped_allocator_ops", "sdca_ops", @@ -1099,6 +1098,14 @@ tf_gen_op_libs( deps = ["//tensorflow/core/kernels:debug_ops"], ) +tf_gen_op_libs( + is_external = False, + op_lib_names = [ + "resource_variable_ops", + ], + deps = [":lib"], +) + # And one for all user ops cc_library( name = "user_ops_op_lib", |