Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Branch 170919783 #13478

Merged
merged 57 commits into from
Oct 4, 2017
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
57 commits
Select commit Hold shift + click to select a range
f94d410
[tf-signal] Add tf.contrib.signal.mfccs_from_log_mel_spectrograms.
rryan Oct 2, 2017
ee4f13d
PiperOrigin-RevId: 170752644
tensorflower-gardener Oct 2, 2017
6d2244e
Improve a text comment related to MonitoredSession's hooks.
isaprykin Oct 2, 2017
0618971
[TF:XLA] Add IdentityN operator.
tensorflower-gardener Oct 2, 2017
553d10c
[TF:XLA] Add support for negative values of "split_dim" argument to S…
hawkinsp Oct 2, 2017
a470779
TF WhereOp now acts more like np.where: extend input types to any num…
ebrevdo Oct 2, 2017
dd94edb
Standardizing device names to the newer /device:<type>:<index> format…
rohan100jain Oct 2, 2017
c55a2e1
Remove deprecated is_training from resnet_arg_scope.
Oct 2, 2017
501253e
Update README.md for tf.contrib.kfac
tensorflower-gardener Oct 2, 2017
c5123fe
Update ops-related pbtxt files.
tensorflower-gardener Oct 2, 2017
320a824
Go: Update generated wrapper functions for TensorFlow ops.
tensorflower-gardener Oct 2, 2017
c280e8c
Move the logic for adding regularization losses to collections
tensorflower-gardener Oct 2, 2017
6ce50b2
Convert inputs to `wasserstein_gradient_penalty` to Tensor.
tensorflower-gardener Oct 2, 2017
9b027db
[tf.data] Use the user-provided type when converting to a NumPy array.
mrry Oct 2, 2017
a2b23b0
Automated g4 rollback of changelist 170758184
ebrevdo Oct 2, 2017
d627ca4
Forwarding out of range errors rather than capturing them within the …
tensorflower-gardener Oct 2, 2017
931268a
Clean up properties of layers.Layer:
tensorflower-gardener Oct 2, 2017
991dea6
[tf-signal] Add a test that windowing, framing, and mel ops are const…
rryan Oct 3, 2017
df3dbba
[tf.data] Internal minor code restructure
tensorflower-gardener Oct 3, 2017
189ccb3
Update Closure Rules dependency to HEAD
jart Oct 3, 2017
b229b06
[tf.contrib.data] Add deprecation decorators to deprecated methods.
mrry Oct 3, 2017
0466135
Fix backwards_compatibility_test broken by rollback of changes to Whe…
tensorflower-gardener Oct 3, 2017
b3d6b40
Adds strong validation on eval metrics returnes by `Estimator.evaluate`
Oct 3, 2017
cb460e4
Add tf.spectral.dct, based on scipy.fftpack.dct.
rryan Oct 3, 2017
955c525
quantize API and copy and modify quantize mangle script to allow open…
Oct 3, 2017
263d025
Add XlaCompiledFunction, a lightweight API for calling XLA computatio…
tensorflower-gardener Oct 3, 2017
0ea4331
Use shape information in constant propagation.
tensorflower-gardener Oct 3, 2017
6425dbd
Update bazel-toolchains repo to use Bazel 0.6.0 toolchain configs.
tensorflower-gardener Oct 3, 2017
14ea6d5
Disable parallelizing over both batch and inner matrix dimensions in …
tensorflower-gardener Oct 3, 2017
448de13
a) Added a new op tf.linalg.slogdet, the equivalent of numpy.linalg.s…
tensorflower-gardener Oct 3, 2017
c724691
Java: Updated release notes to include some recent contributions.
asimshankar Oct 3, 2017
3e7ac6d
Use the -l flag of nvidia-smi rather than watch "nvidia-smi".
tensorflower-gardener Oct 3, 2017
fbdb366
Allow "." in list item names of Args/Returns/Raises blocks
MarkDaoust Oct 3, 2017
2db3e32
Added CheckpointSavingListeners into Estimator.train interface. This …
ispirmustafa Oct 3, 2017
7020f17
Correct names for contrib summaries.
alextp Oct 3, 2017
0e286d3
Bugfix: tf.random_gamma incorrectly handles non-batch, scalar draws.
jvdillon Oct 3, 2017
cb291f3
Bugfix: tf.contrib.distributions.Affine incorrectly computed log-det-…
jvdillon Oct 3, 2017
68ec8b8
Bugfix: tf.random_gamma incorrectly handles non-batch, scalar draws.
jvdillon Oct 3, 2017
9d93a11
Bugfix: tf.contrib.distributions.Affine incorrectly computed log-det-…
jvdillon Oct 3, 2017
84b1d6d
Split restore_variables_on_create out of tfe.Saver
tensorflower-gardener Oct 3, 2017
0be0671
Don't use dlsym to resolve symbols in the CPU JIT
Oct 3, 2017
941a498
Made sure the save/restore op is preserved during optimization
benoitsteiner Oct 3, 2017
57c5613
Makes custom_gradient work in graph mode.
alextp Oct 3, 2017
31058d7
[XLA] Add documentation for ReducePrecision HLO instruction.
tensorflower-gardener Oct 3, 2017
e9180e5
Adds the following ops to android_extended_ops_group2:
tensorflower-gardener Oct 3, 2017
9be9649
[tf.data] Fix docstring descriptions.
tensorflower-gardener Oct 3, 2017
3f57902
Convert cells to OO-based to reduce call() overhead
tensorflower-gardener Oct 3, 2017
a19d80d
[CMake] Add tf_cc_while_loop to the list of objects in tf_shared_lib.…
mrry Oct 3, 2017
8fb14b1
get_variable in graph_callable returns Variable objects and not Tensors.
alextp Oct 3, 2017
0a11eaf
Internal Variant API allowing registering Variants to be copied from/…
ebrevdo Oct 3, 2017
7d62e1d
[tf.data] Fix typo in docstring.
mrry Oct 3, 2017
1a04342
Make Saver work with GPU
tensorflower-gardener Oct 3, 2017
2dab9fd
Update labels docstring to match.
tensorflower-gardener Oct 3, 2017
65ae3e9
Automated g4 rollback of changelist 170892257
gunan Oct 3, 2017
66df43d
Handle the absence of a fresh eval checkpoint in `run_local`.
isaprykin Oct 3, 2017
435b31b
Automated g4 rollback of changelist 170892257
gunan Oct 3, 2017
2d98bd9
Merge commit for internal changes
Oct 3, 2017
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
3 changes: 3 additions & 0 deletions RELEASE.md
Original file line number Diff line number Diff line change
@@ -1,6 +1,9 @@
# Release 1.4.0

## Major Features And Improvements
* Java:
* Generics (e.g., `Tensor<Integer>`) for improved type-safety (courtesy @andrewcmyers).
* Support for multi-dimensional string tensors.

## Bug Fixes and Other Changes
* `tf.nn.rnn_cell.DropoutWrapper` is now more careful about dropping out LSTM
Expand Down
8 changes: 4 additions & 4 deletions WORKSPACE
Original file line number Diff line number Diff line change
Expand Up @@ -2,11 +2,11 @@ workspace(name = "org_tensorflow")

http_archive(
name = "io_bazel_rules_closure",
sha256 = "25f5399f18d8bf9ce435f85c6bbf671ec4820bc4396b3022cc5dc4bc66303609",
strip_prefix = "rules_closure-0.4.2",
sha256 = "110fe68753413777944b473c25eed6368c4a0487cee23a7bac1b13cc49d3e257",
strip_prefix = "rules_closure-4af89ef1db659eb41f110df189b67d4cf14073e1",
urls = [
"http://mirror.bazel.build/github.com/bazelbuild/rules_closure/archive/0.4.2.tar.gz", # 2017-08-29
"https://github.com/bazelbuild/rules_closure/archive/0.4.2.tar.gz",
"http://mirror.bazel.build/github.com/bazelbuild/rules_closure/archive/4af89ef1db659eb41f110df189b67d4cf14073e1.tar.gz",
"https://github.com/bazelbuild/rules_closure/archive/4af89ef1db659eb41f110df189b67d4cf14073e1.tar.gz", # 2017-08-28
],
)

Expand Down
1 change: 1 addition & 0 deletions tensorflow/BUILD
Original file line number Diff line number Diff line change
Expand Up @@ -400,6 +400,7 @@ filegroup(
"//tensorflow/contrib/nn:all_files",
"//tensorflow/contrib/opt:all_files",
"//tensorflow/contrib/predictor:all_files",
"//tensorflow/contrib/quantize:all_files",
"//tensorflow/contrib/receptive_field:all_files",
"//tensorflow/contrib/reduce_slice_ops:all_files",
"//tensorflow/contrib/remote_fused_graph/pylib:all_files",
Expand Down
25 changes: 18 additions & 7 deletions tensorflow/cc/ops/const_op.cc
Original file line number Diff line number Diff line change
Expand Up @@ -19,19 +19,17 @@ limitations under the License.
namespace tensorflow {
namespace ops {

Output Const(const Scope& scope, const Input::Initializer& val) {
namespace {
template <typename T>
Output ConstHelper(const Scope& scope, const T& value, DataType dtype) {
if (!scope.ok()) return Output();
if (!val.status.ok()) {
scope.UpdateStatus(val.status);
return Output();
}

Node* ret;
Graph* graph = scope.graph();
const string unique_name = scope.GetUniqueNameForOp("Const");
auto builder = NodeBuilder(unique_name, "Const")
.Attr("value", val.tensor)
.Attr("dtype", val.tensor.dtype());
.Attr("value", value)
.Attr("dtype", dtype);
scope.UpdateBuilder(&builder);
scope.UpdateStatus(builder.Finalize(graph, &ret));
if (!scope.ok()) return Output();
Expand All @@ -41,6 +39,19 @@ Output Const(const Scope& scope, const Input::Initializer& val) {

return Output(ret);
}
} // namespace

Output Const(const Scope& scope, const Input::Initializer& val) {
if (!val.status.ok()) {
scope.UpdateStatus(val.status);
return Output();
}
return ConstHelper(scope, val.tensor, val.tensor.dtype());
}

Output ConstFromProto(const Scope& scope, const TensorProto& proto) {
return ConstHelper(scope, proto, proto.dtype());
}

NodeBuilder::NodeOut AsNodeOut(const Scope& scope, const Input& inp) {
if (!inp.status().ok()) {
Expand Down
2 changes: 2 additions & 0 deletions tensorflow/cc/ops/const_op.h
Original file line number Diff line number Diff line change
Expand Up @@ -28,6 +28,8 @@ namespace ops {

Output Const(const Scope& scope, const Input::Initializer& val);

Output ConstFromProto(const Scope& scope, const TensorProto& proto);

NodeBuilder::NodeOut AsNodeOut(const Scope& scope, const Input& inp);

template <typename T>
Expand Down
14 changes: 14 additions & 0 deletions tensorflow/cc/ops/const_op_test.cc
Original file line number Diff line number Diff line change
Expand Up @@ -100,6 +100,20 @@ TEST(ConstOpTest, WithExplicitShape) {
ExpectNodeEqual<string>(d.node(), {"1", "2", "3", "4", "5", "6"}, {2, 3});
}

TEST(ConstOpTest, FromProto) {
Scope root = Scope::NewRootScope();
TensorProto proto;
proto.set_dtype(DT_DOUBLE);
TensorShape({2, 2}).AsProto(proto.mutable_tensor_shape());
for (int i = 0; i < 4; ++i) {
proto.add_double_val(static_cast<double>(i));
}
auto c = ops::ConstFromProto(root, proto);
TF_CHECK_OK(root.status());
EXPECT_EQ(c.op().output_type(0), DT_DOUBLE);
ExpectNodeEqual<double>(c.node(), {0.0, 1.0, 2.0, 3.0}, {2, 2});
}

TEST(ConstOpTest, InvalidInitializer) {
Scope root = Scope::NewRootScope();
ops::Const(root, {{2.0}, {"df"}});
Expand Down