Skip to content
体验新版
项目
组织
正在加载...
登录
切换导航
打开侧边栏
xxadev
tensorflow
提交
d57572e9
T
tensorflow
项目概览
xxadev
/
tensorflow
与 Fork 源项目一致
从无法访问的项目Fork
通知
3
Star
1
Fork
0
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
0
列表
看板
标记
里程碑
合并请求
0
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
T
tensorflow
项目概览
项目概览
详情
发布
仓库
仓库
文件
提交
分支
标签
贡献者
分支图
比较
Issue
0
Issue
0
列表
看板
标记
里程碑
合并请求
0
合并请求
0
Pages
分析
分析
仓库分析
DevOps
Wiki
0
Wiki
成员
成员
收起侧边栏
关闭侧边栏
动态
分支图
创建新Issue
提交
Issue看板
前往新版Gitcode,体验更适合开发者的 AI 搜索 >>
提交
d57572e9
编写于
9月 02, 2017
作者:
M
Martin Wicke
提交者:
TensorFlower Gardener
9月 02, 2017
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
Merge changes from github.
PiperOrigin-RevId: 167401527
上级
ddba1e0a
变更
33
隐藏空白更改
内联
并排
Showing
33 changed file
with
134 addition
and
84 deletion
+134
-84
README.md
README.md
+8
-0
RELEASE.md
RELEASE.md
+1
-1
WORKSPACE
WORKSPACE
+4
-4
configure.py
configure.py
+5
-5
tensorflow/cc/gradients/nn_grad.cc
tensorflow/cc/gradients/nn_grad.cc
+8
-0
tensorflow/cc/gradients/nn_grad_test.cc
tensorflow/cc/gradients/nn_grad_test.cc
+8
-0
tensorflow/compiler/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc
...piler/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc
+1
-1
tensorflow/contrib/cmake/tf_tests.cmake
tensorflow/contrib/cmake/tf_tests.cmake
+2
-0
tensorflow/contrib/gdr/BUILD
tensorflow/contrib/gdr/BUILD
+1
-0
tensorflow/contrib/gdr/gdr_memory_manager.h
tensorflow/contrib/gdr/gdr_memory_manager.h
+1
-6
tensorflow/contrib/layers/__init__.py
tensorflow/contrib/layers/__init__.py
+1
-0
tensorflow/contrib/layers/python/layers/optimizers.py
tensorflow/contrib/layers/python/layers/optimizers.py
+3
-2
tensorflow/contrib/learn/python/learn/learn_io/data_feeder.py
...orflow/contrib/learn/python/learn/learn_io/data_feeder.py
+15
-11
tensorflow/contrib/makefile/Makefile
tensorflow/contrib/makefile/Makefile
+6
-6
tensorflow/contrib/makefile/README.md
tensorflow/contrib/makefile/README.md
+2
-1
tensorflow/core/distributed_runtime/base_rendezvous_mgr.cc
tensorflow/core/distributed_runtime/base_rendezvous_mgr.cc
+15
-12
tensorflow/core/framework/op_kernel.h
tensorflow/core/framework/op_kernel.h
+1
-1
tensorflow/core/profiler/g3doc/advise.md
tensorflow/core/profiler/g3doc/advise.md
+2
-2
tensorflow/core/profiler/g3doc/command_line.md
tensorflow/core/profiler/g3doc/command_line.md
+3
-3
tensorflow/core/profiler/g3doc/options.md
tensorflow/core/profiler/g3doc/options.md
+4
-4
tensorflow/core/profiler/g3doc/profile_memory.md
tensorflow/core/profiler/g3doc/profile_memory.md
+1
-1
tensorflow/core/profiler/g3doc/profile_model_architecture.md
tensorflow/core/profiler/g3doc/profile_model_architecture.md
+4
-4
tensorflow/core/profiler/g3doc/profile_time.md
tensorflow/core/profiler/g3doc/profile_time.md
+6
-6
tensorflow/docs_src/install/install_linux.md
tensorflow/docs_src/install/install_linux.md
+3
-3
tensorflow/docs_src/install/install_windows.md
tensorflow/docs_src/install/install_windows.md
+4
-2
tensorflow/examples/speech_commands/README.md
tensorflow/examples/speech_commands/README.md
+1
-1
tensorflow/python/feature_column/feature_column.py
tensorflow/python/feature_column/feature_column.py
+1
-1
tensorflow/python/feature_column/feature_column_test.py
tensorflow/python/feature_column/feature_column_test.py
+14
-0
tensorflow/stream_executor/device_description.h
tensorflow/stream_executor/device_description.h
+2
-2
tensorflow/stream_executor/kernel.h
tensorflow/stream_executor/kernel.h
+1
-1
tensorflow/tools/ci_build/update_version.py
tensorflow/tools/ci_build/update_version.py
+3
-2
tensorflow/tools/pip_package/BUILD
tensorflow/tools/pip_package/BUILD
+1
-0
third_party/sqlite.BUILD
third_party/sqlite.BUILD
+2
-2
未找到文件。
README.md
浏览文件 @
d57572e9
...
...
@@ -36,7 +36,15 @@ and discussion, and please direct specific questions to [Stack Overflow](https:/
People who are a little more adventurous can also try our nightly binaries:
**Nightly pip packages**
*
We are pleased to announce that TensorFlow now offers nightly pip packages
under the
[
tf-nightly
](
https://pypi.python.org/pypi/tf-nightly
)
project on pypi.
Simply run
`pip install tf-nightly`
in a clean environment to install the nightly
tensorflow build. We currently only support CPU-only packages on Linux and Mac.
GPU packages on all platforms and Windows CPU-only packages will arrive soon!
**Individual whl files**
*
Linux CPU-only:
[
Python 2
](
https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=cpu-slave/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.3.0-cp27-none-linux_x86_64.whl
)
(
[build
history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=cpu-slave)) /
[
Python 3.4
](
https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=cpu-slave/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.3.0-cp34-cp34m-linux_x86_64.whl
)
(
[build
history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=cpu-slave/)) /
[
Python 3.5
](
https://ci.tensorflow.org/view/Nightly/job/nightly-python35-linux-cpu/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.3.0-cp35-cp35m-linux_x86_64.whl
)
(
[build
history](https://ci.tensorflow.org/view/Nightly/job/nightly-python35-linux-cpu/))
*
Linux GPU:
[
Python 2
](
https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=gpu-linux/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow_gpu-1.3.0-cp27-none-linux_x86_64.whl
)
(
[build
history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=gpu-linux/)) /
[
Python 3.4
](
https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=gpu-linux/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow_gpu-1.3.0-cp34-cp34m-linux_x86_64.whl
)
(
[build
history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=gpu-linux/)) /
[
Python 3.5
](
https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3.5,label=gpu-linux/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow_gpu-1.3.0-cp35-cp35m-linux_x86_64.whl
)
(
[build
history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-linux-gpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3.5,label=gpu-linux/))
*
Mac CPU-only:
[
Python 2
](
https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=mac-slave/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.3.0-py2-none-any.whl
)
(
[build
history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=mac-slave/)) /
[
Python 3
](
https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=mac-slave/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-1.3.0-py3-none-any.whl
)
(
[build
history](https://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=mac-slave/))
...
...
RELEASE.md
浏览文件 @
d57572e9
...
...
@@ -46,7 +46,7 @@ See also [TensorBoard 0.1.4](https://github.com/tensorflow/tensorboard/releases/
*
Display feed values with the
`print_feed`
or
`pf`
command and clickable links in the curses UI.
*
Runtime profiler at the op level and the Python source line level with the
`run -p`
command.
*
Initial release of the statistical distribution library
`tf.distributions`
.
*
GPU kernels and speed improvements for
for
unary
`tf.where`
and
`tf.nn.top_k`
.
*
GPU kernels and speed improvements for unary
`tf.where`
and
`tf.nn.top_k`
.
*
Monotonic Attention wrappers added to
`tf.contrib.seq2seq`
.
*
Added
`tf.contrib.signal`
, a library for signal processing primitives.
*
Added
`tf.contrib.resampler`
, containing CPU and GPU ops for differentiable resampling of images.
...
...
WORKSPACE
浏览文件 @
d57572e9
...
...
@@ -2,11 +2,11 @@ workspace(name = "org_tensorflow")
http_archive
(
name
=
"io_bazel_rules_closure"
,
sha256
=
"
bc41b80486413aaa551860fc37471dbc0666e1dbb5236fb6177cb83b0c105846
"
,
strip_prefix
=
"rules_closure-
dec425a4ff3faf09a56c85d082e4eed05d8ce38f
"
,
sha256
=
"
25f5399f18d8bf9ce435f85c6bbf671ec4820bc4396b3022cc5dc4bc66303609
"
,
strip_prefix
=
"rules_closure-
0.4.2
"
,
urls
=
[
"http://mirror.bazel.build/github.com/bazelbuild/rules_closure/archive/
dec425a4ff3faf09a56c85d082e4eed05d8ce38f.tar.gz"
,
# 2017-06-02
"https://github.com/bazelbuild/rules_closure/archive/
dec425a4ff3faf09a56c85d082e4eed05d8ce38f
.tar.gz"
,
"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"
,
],
)
...
...
configure.py
浏览文件 @
d57572e9
...
...
@@ -143,7 +143,7 @@ def run_shell(cmd, allow_non_zero=False):
def
cygpath
(
path
):
"""Convert path from posix to windows."""
return
run_shell
([
'cygpath'
,
'-m'
,
path
]
)
return
os
.
path
.
abspath
(
path
).
replace
(
'
\\
'
,
'/'
)
def
get_python_path
(
environ_cp
,
python_bin_path
):
...
...
@@ -196,7 +196,7 @@ def setup_python(environ_cp, bazel_version):
environ_cp
[
'PYTHON_BIN_PATH'
]
=
''
# Convert python path to Windows style before checking lib and version
if
is_cygwin
():
if
is_
windows
()
or
is_
cygwin
():
python_bin_path
=
cygpath
(
python_bin_path
)
# Get PYTHON_LIB_PATH
...
...
@@ -219,7 +219,7 @@ def setup_python(environ_cp, bazel_version):
python_major_version
=
get_python_major_version
(
python_bin_path
)
# Convert python path to Windows style before writing into bazel.rc
if
is_cygwin
():
if
is_
windows
()
or
is_
cygwin
():
python_lib_path
=
cygpath
(
python_lib_path
)
# Set-up env variables used by python_configure.bzl
...
...
@@ -600,7 +600,7 @@ def set_tf_cuda_version(environ_cp):
# Find out where the CUDA toolkit is installed
default_cuda_path
=
_DEFAULT_CUDA_PATH
if
is_cygwin
():
if
is_
windows
()
or
is_
cygwin
():
default_cuda_path
=
cygpath
(
environ_cp
.
get
(
'CUDA_PATH'
,
_DEFAULT_CUDA_PATH_WIN
))
elif
is_linux
():
...
...
@@ -660,7 +660,7 @@ def set_tf_cunn_version(environ_cp):
# unusable. Going through one more level of expansion to handle that.
cudnn_install_path
=
os
.
path
.
realpath
(
os
.
path
.
expanduser
(
cudnn_install_path
))
if
is_cygwin
():
if
is_
windows
()
or
is_
cygwin
():
cudnn_install_path
=
cygpath
(
cudnn_install_path
)
if
is_windows
():
...
...
tensorflow/cc/gradients/nn_grad.cc
浏览文件 @
d57572e9
...
...
@@ -95,6 +95,14 @@ Status SeluGradHelper(const Scope& scope, const Operation& op,
}
REGISTER_GRADIENT_OP
(
"Selu"
,
SeluGradHelper
);
Status
L2LossGrad
(
const
Scope
&
scope
,
const
Operation
&
op
,
const
std
::
vector
<
Output
>&
grad_inputs
,
std
::
vector
<
Output
>*
grad_outputs
)
{
grad_outputs
->
push_back
(
Mul
(
scope
,
op
.
input
(
0
),
grad_inputs
[
0
]));
return
scope
.
status
();
}
REGISTER_GRADIENT_OP
(
"L2Loss"
,
L2LossGrad
);
Status
BiasAddGradHelper
(
const
Scope
&
scope
,
const
Operation
&
op
,
const
std
::
vector
<
Output
>&
grad_inputs
,
std
::
vector
<
Output
>*
grad_outputs
)
{
...
...
tensorflow/cc/gradients/nn_grad_test.cc
浏览文件 @
d57572e9
...
...
@@ -122,6 +122,14 @@ TEST_F(NNGradTest, SeluGrad) {
RunTest
(
x
,
x_init_value
,
y
,
shape
);
}
TEST_F
(
NNGradTest
,
L2LossGrad
)
{
TensorShape
x_shape
({
5
,
2
});
TensorShape
y_shape
({
1
});
auto
x
=
Placeholder
(
scope_
,
DT_FLOAT
,
Placeholder
::
Shape
(
x_shape
));
auto
y
=
L2Loss
(
scope_
,
x
);
RunTest
(
x
,
x_shape
,
y
,
y_shape
);
}
TEST_F
(
NNGradTest
,
BiasAddGradHelper
)
{
TensorShape
shape
({
4
,
5
});
TensorShape
bias_shape
({
5
});
...
...
tensorflow/compiler/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc
浏览文件 @
d57572e9
...
...
@@ -389,7 +389,7 @@ StatusOr<string> CompileModuleToPtx(llvm::Module* module,
// Loop unrolling exposes more opportunities for SROA. Therefore, we run SROA
// again after the standard optimization passes [http://b/13329423].
// TODO(jingyue): SROA may further expose more optimization opportunities
,
such
// TODO(jingyue): SROA may further expose more optimization opportunities such
// as more precise alias analysis and more function inlining (SROA may change
// the inlining cost of a function). For now, running SROA already emits good
// enough code for the evaluated benchmarks. We may want to run more
...
...
tensorflow/contrib/cmake/tf_tests.cmake
浏览文件 @
d57572e9
...
...
@@ -82,6 +82,7 @@ function(AddTest)
set_tests_properties
(
${
_AT_TARGET
}
PROPERTIES ENVIRONMENT
"TEST_TMPDIR=
${
tempdir
}
;TEST_SRCDIR=
${
testdir
}
"
)
set_tests_properties
(
${
_AT_TARGET
}
PROPERTIES TIMEOUT
"600"
)
foreach
(
datafile
${
_AT_DATA
}
)
file
(
RELATIVE_PATH datafile_rel
${
tensorflow_source_dir
}
${
datafile
}
)
...
...
@@ -117,6 +118,7 @@ function(AddPythonTests)
if
(
_AT_DEPENDS
)
add_dependencies
(
${
_AT_TARGET
}
${
_AT_DEPENDS
}
)
endif
()
set_tests_properties
(
${
sourcefile
}
PROPERTIES TIMEOUT
"600"
)
endforeach
()
endfunction
(
AddPythonTests
)
...
...
tensorflow/contrib/gdr/BUILD
浏览文件 @
d57572e9
...
...
@@ -62,6 +62,7 @@ tf_cuda_library(
}),
deps
=
[
":gdr_proto_cc"
,
"//tensorflow/core:core_cpu_internal"
,
"//tensorflow/core:framework"
,
"//tensorflow/core:gpu_runtime"
,
"//tensorflow/core:lib"
,
...
...
tensorflow/contrib/gdr/gdr_memory_manager.h
浏览文件 @
d57572e9
...
...
@@ -16,14 +16,9 @@ limitations under the License.
#ifndef GDR_MEMORY_MANAGER_H_
#define GDR_MEMORY_MANAGER_H_
#include "google/protobuf/any.pb.h"
#include "tensorflow/core/lib/core/status.h"
namespace
google
{
namespace
protobuf
{
class
Any
;
}
}
namespace
tensorflow
{
class
Device
;
...
...
tensorflow/contrib/layers/__init__.py
浏览文件 @
d57572e9
...
...
@@ -115,6 +115,7 @@ _allowed_symbols = ['bias_add',
'legacy_linear'
,
'legacy_relu'
,
'OPTIMIZER_CLS_NAMES'
,
'OPTIMIZER_SUMMARIES'
,
'regression_target'
,
'SPARSE_FEATURE_CROSS_DEFAULT_HASH_KEY'
,
'summaries'
]
...
...
tensorflow/contrib/layers/python/layers/optimizers.py
浏览文件 @
d57572e9
...
...
@@ -129,8 +129,9 @@ def optimize_loss(loss,
`None` to use all trainable variables.
name: The name for this operation is used to scope operations and summaries.
summaries: List of internal quantities to visualize on tensorboard. If not
set only the loss and the learning rate will be reported. The
complete list is in OPTIMIZER_SUMMARIES.
set, the loss, the learning rate, and the global norm of the
gradients will be reported. The complete list of possible values
is in OPTIMIZER_SUMMARIES.
colocate_gradients_with_ops: If True, try colocating gradients with the
corresponding op.
increment_global_step: Whether to increment `global_step`. If your model
...
...
tensorflow/contrib/learn/python/learn/learn_io/data_feeder.py
浏览文件 @
d57572e9
...
...
@@ -28,7 +28,6 @@ import six
from
six.moves
import
xrange
# pylint: disable=redefined-builtin
from
tensorflow.python.framework
import
dtypes
from
tensorflow.python.framework
import
ops
from
tensorflow.python.ops
import
array_ops
from
tensorflow.python.platform
import
tf_logging
as
logging
...
...
@@ -44,7 +43,7 @@ def _get_in_out_shape(x_shape, y_shape, n_classes, batch_size=None):
x_is_dict
,
y_is_dict
=
isinstance
(
x_shape
,
dict
),
y_shape
is
not
None
and
isinstance
(
y_shape
,
dict
)
if
y_is_dict
and
n_classes
is
not
None
:
assert
(
isinstance
(
n_classes
,
dict
)
)
assert
isinstance
(
n_classes
,
dict
)
if
batch_size
is
None
:
batch_size
=
list
(
x_shape
.
values
())[
0
][
0
]
if
x_is_dict
else
x_shape
[
0
]
...
...
@@ -322,10 +321,12 @@ class DataFeeder(object):
self
.
_x
=
dict
([(
k
,
check_array
(
v
,
v
.
dtype
))
for
k
,
v
in
list
(
x
.
items
())
])
if
x_is_dict
else
check_array
(
x
,
x
.
dtype
)
self
.
_y
=
None
if
y
is
None
else
\
dict
([(
k
,
check_array
(
v
,
v
.
dtype
))
for
k
,
v
in
list
(
y
.
items
())])
if
x_is_dict
else
check_array
(
y
,
y
.
dtype
)
self
.
_y
=
None
if
y
is
None
else
(
dict
([(
k
,
check_array
(
v
,
v
.
dtype
))
for
k
,
v
in
list
(
y
.
items
())])
if
y_is_dict
else
check_array
(
y
,
y
.
dtype
))
# self.n_classes is not None means we're converting raw target indices to one-hot.
# self.n_classes is not None means we're converting raw target indices
# to one-hot.
if
n_classes
is
not
None
:
if
not
y_is_dict
:
y_dtype
=
(
np
.
int64
...
...
@@ -344,12 +345,15 @@ class DataFeeder(object):
x_shape
,
y_shape
,
n_classes
,
batch_size
)
# Input dtype matches dtype of x.
self
.
_input_dtype
=
dict
([(
k
,
_check_dtype
(
v
.
dtype
))
for
k
,
v
in
list
(
self
.
_x
.
items
())])
if
x_is_dict
\
else
_check_dtype
(
self
.
_x
.
dtype
)
# note: self._output_dtype = np.float32 when y is None
self
.
_output_dtype
=
dict
([(
k
,
_check_dtype
(
v
.
dtype
))
for
k
,
v
in
list
(
self
.
_y
.
items
())])
if
y_is_dict
\
else
_check_dtype
(
self
.
_y
.
dtype
)
if
y
is
not
None
else
np
.
float32
self
.
_input_dtype
=
(
dict
([(
k
,
_check_dtype
(
v
.
dtype
))
for
k
,
v
in
list
(
self
.
_x
.
items
())])
if
x_is_dict
else
_check_dtype
(
self
.
_x
.
dtype
))
# self._output_dtype == np.float32 when y is None
self
.
_output_dtype
=
(
dict
([(
k
,
_check_dtype
(
v
.
dtype
))
for
k
,
v
in
list
(
self
.
_y
.
items
())])
if
y_is_dict
else
(
_check_dtype
(
self
.
_y
.
dtype
)
if
y
is
not
None
else
np
.
float32
))
# self.n_classes is None means we're passing in raw target indices
if
n_classes
is
not
None
and
y_is_dict
:
...
...
tensorflow/contrib/makefile/Makefile
浏览文件 @
d57572e9
...
...
@@ -316,14 +316,14 @@ ifeq ($(TARGET),IOS)
IPHONESIMULATOR_SYSROOT
:=
$(
shell
xcrun
--sdk
iphonesimulator
\
--show-sdk-path
)
IOS_SDK_VERSION
:=
$(
shell
xcrun
--sdk
iphoneos
--show-sdk-version
)
MIN_SDK_VERSION
:=
8
.0
MIN_SDK_VERSION
:=
9
.0
# Override IOS_ARCH with ARMV7, ARMV7S, ARM64, or I386.
IOS_ARCH
:=
X86_64
ifeq
($(IOS_ARCH),ARMV7)
CXXFLAGS
+=
-miphoneos-version-min
=
$(MIN_SDK_VERSION)
\
-arch
armv7
\
-fembed-bitcode
\
-D__thread
=
\
-D__thread
=
thread_local
\
-DUSE_GEMM_FOR_CONV
\
-Wno-c
++11-narrowing
\
-mno-thumb
\
...
...
@@ -347,7 +347,7 @@ ifeq ($(TARGET),IOS)
CXXFLAGS
+=
-miphoneos-version-min
=
$(MIN_SDK_VERSION)
\
-arch
armv7s
\
-fembed-bitcode
\
-D__thread
=
\
-D__thread
=
thread_local
\
-DUSE_GEMM_FOR_CONV
\
-Wno-c
++11-narrowing
\
-mno-thumb
\
...
...
@@ -371,7 +371,7 @@ ifeq ($(TARGET),IOS)
CXXFLAGS
+=
-miphoneos-version-min
=
$(MIN_SDK_VERSION)
\
-arch
arm64
\
-fembed-bitcode
\
-D__thread
=
\
-D__thread
=
thread_local
\
-DUSE_GEMM_FOR_CONV
\
-Wno-c
++11-narrowing
\
-DTF_LEAN_BINARY
\
...
...
@@ -395,7 +395,7 @@ ifeq ($(TARGET),IOS)
-arch
i386
\
-mno-sse
\
-fembed-bitcode
\
-D__thread
=
\
-D__thread
=
thread_local
\
-DUSE_GEMM_FOR_CONV
\
-Wno-c
++11-narrowing
\
-DTF_LEAN_BINARY
\
...
...
@@ -418,7 +418,7 @@ ifeq ($(TARGET),IOS)
CXXFLAGS
+=
-mios-simulator-version-min
=
$(MIN_SDK_VERSION)
\
-arch
x86_64
\
-fembed-bitcode
\
-D__thread
=
\
-D__thread
=
thread_local
\
-DUSE_GEMM_FOR_CONV
\
-Wno-c
++11-narrowing
\
-DTF_LEAN_BINARY
\
...
...
tensorflow/contrib/makefile/README.md
浏览文件 @
d57572e9
...
...
@@ -201,7 +201,8 @@ tensorflow/contrib/makefile/compile_ios_protobuf.sh
Then, you will need to compile the nsync library for iOS:
```
export HOST_NSYNC_LIB=`tensorflow/contrib/makefile/compile_nsync.sh`
```
bash
export
HOST_NSYNC_LIB
=
`
tensorflow/contrib/makefile/compile_nsync.sh
`
export
TARGET_NSYNC_LIB
=
`
tensorflow/contrib/makefile/compile_nsync.sh
-t
ios
`
```
...
...
tensorflow/core/distributed_runtime/base_rendezvous_mgr.cc
浏览文件 @
d57572e9
...
...
@@ -35,14 +35,18 @@ limitations under the License.
namespace
tensorflow
{
static
void
StartAbortRendevous
(
Rendezvous
*
rendez
,
const
Status
&
s
)
{
rendez
->
StartAbort
(
s
);
rendez
->
Unref
();
}
BaseRendezvousMgr
::
BaseRendezvousMgr
(
const
WorkerEnv
*
worker_env
)
:
worker_env_
(
worker_env
)
{}
BaseRendezvousMgr
::~
BaseRendezvousMgr
()
{
for
(
auto
&
p
:
table_
)
{
BaseRemoteRendezvous
*
rendez
=
p
.
second
;
rendez
->
StartAbort
(
errors
::
Aborted
(
"Shutdown"
));
rendez
->
Unref
();
auto
rendez
=
p
.
second
;
StartAbortRendevous
(
rendez
,
errors
::
Aborted
(
"Shutdown"
));
}
}
...
...
@@ -52,7 +56,7 @@ RemoteRendezvous* BaseRendezvousMgr::Find(int64 step_id) {
BaseRemoteRendezvous
*
BaseRendezvousMgr
::
FindOrCreate
(
int64
step_id
)
{
mutex_lock
l
(
mu_
);
Table
::
iterator
iter
=
table_
.
find
(
step_id
);
auto
iter
=
table_
.
find
(
step_id
);
if
(
iter
==
table_
.
end
())
{
auto
rr
=
Create
(
step_id
,
worker_env_
);
iter
=
table_
.
insert
({
step_id
,
rr
}).
first
;
...
...
@@ -64,7 +68,7 @@ BaseRemoteRendezvous* BaseRendezvousMgr::FindOrCreate(int64 step_id) {
void
BaseRendezvousMgr
::
RecvLocalAsync
(
int64
step_id
,
const
Rendezvous
::
ParsedKey
&
parsed
,
Rendezvous
::
DoneCallback
done
)
{
BaseRemoteRendezvous
*
rendez
=
FindOrCreate
(
step_id
);
auto
rendez
=
FindOrCreate
(
step_id
);
using
namespace
std
::
placeholders
;
Rendezvous
::
DoneCallback
done_cb
=
std
::
bind
(
[
rendez
](
Rendezvous
::
DoneCallback
done
,
...
...
@@ -101,15 +105,15 @@ void BaseRendezvousMgr::Cleanup(int64 step_id) {
Rendezvous
*
rendez
=
nullptr
;
{
mutex_lock
l
(
mu_
);
Table
::
iterator
iter
=
table_
.
find
(
step_id
);
auto
iter
=
table_
.
find
(
step_id
);
if
(
iter
!=
table_
.
end
())
{
rendez
=
iter
->
second
;
table_
.
erase
(
iter
);
}
}
if
(
!
rendez
)
return
;
rendez
->
StartAbort
(
errors
::
Aborted
(
"Cleanup "
,
step_id
));
rendez
->
Unref
();
if
(
rendez
)
{
StartAbortRendevous
(
rendez
,
errors
::
Aborted
(
"Cleanup "
,
step_id
));
}
}
void
BaseRendezvousMgr
::
CleanupAll
()
{
...
...
@@ -122,8 +126,7 @@ void BaseRendezvousMgr::CleanupAll() {
table_
.
clear
();
}
for
(
auto
rendez
:
rendezs
)
{
rendez
->
StartAbort
(
errors
::
Aborted
(
"Shutdown"
));
rendez
->
Unref
();
StartAbortRendevous
(
rendez
,
errors
::
Aborted
(
"Shutdown"
));
}
}
...
...
@@ -165,7 +168,7 @@ Status BaseRemoteRendezvous::Initialize(WorkerSession* session) {
session_
=
session
;
std
::
swap
(
deferred_calls
,
deferred_calls_
);
}
for
(
DeferredCall
&
call
:
deferred_calls
)
{
for
(
auto
&
call
:
deferred_calls
)
{
RecvLocalAsyncInternal
(
call
.
parsed
,
std
::
move
(
call
.
done
));
}
return
Status
::
OK
();
...
...
tensorflow/core/framework/op_kernel.h
浏览文件 @
d57572e9
...
...
@@ -310,7 +310,7 @@ class OpKernelConstruction {
FunctionLibraryRuntime
*
function_library
()
const
{
return
flib_
;
}
// The GraphDef version whose behavior we should follow.
const
int
graph_def_version
()
const
{
return
graph_def_version_
;
}
int
graph_def_version
()
const
{
return
graph_def_version_
;
}
// Helper routines for the OP_REQUIRES macros
void
CtxFailure
(
Status
s
);
...
...
tensorflow/core/profiler/g3doc/advise.md
浏览文件 @
d57572e9
...
...
@@ -86,7 +86,7 @@ For example:
*
Checks RecvTensor RPC latency and bandwidth.
*
Checks CPU/Memory utilization of the job.
####AcceleratorUtilization Checker
####
AcceleratorUtilization Checker
*
Checks what percentage of time the accelerator spends on computation.
#### OperationChecker
...
...
@@ -100,7 +100,7 @@ For example:
*
Checks the most expensive graph nodes.
*
Checks the most expensive graph-building Python codes.
####Contribute Your Checker
####
Contribute Your Checker
Follow examples of accelerator_utilization_checker.h
...
...
tensorflow/core/profiler/g3doc/command_line.md
浏览文件 @
d57572e9
...
...
@@ -51,7 +51,7 @@ It defines _checkpoint_variable op type. It also provides checkpointed tensors'
Note: this feature is not well maintained now.
###Start `tfprof`
###
Start `tfprof`
#### Build `tfprof`
...
...
@@ -140,9 +140,9 @@ tfprof>
-output
```
###Examples
###
Examples
####Profile Python Time
####
Profile Python Time
```
shell
# Requires --graph_path --op_log_path
tfprof> code
-max_depth
1000
-show_name_regexes
.
*
model_analyzer.
*
py.
*
-select
micros
-account_type_regexes
.
*
-order_by
micros
...
...
tensorflow/core/profiler/g3doc/options.md
浏览文件 @
d57572e9
##Options
##
Options
###Overview
###
Overview
For all tfprof views, the profiles are processed with the following procedures
...
...
@@ -35,7 +35,7 @@ For all tfprof views, the profiles are processed with the following procedures
4) Finally, the filtered data structure is output in a format depending
on the
`-output`
option.
####Option Semantics In Different View
####
Option Semantics In Different View
options usually have the same semantics in different views. However, some
can vary. For example
`-max_depth`
in scope view means the depth of
name scope
<b>
tree
</b>
. In op view, it means the length of operation
<b>
list
</b>
.
...
...
@@ -68,7 +68,7 @@ output_bytes: The memory output by the operation. It's not necessarily requested
by the current operation. For example, it can be a tensor
forwarded from input to output, with in-place mutation.
###Docs
###
Docs
`-max_depth`
: Show nodes that are at most this number of hops from starting node in the data structure.
...
...
tensorflow/core/profiler/g3doc/profile_memory.md
浏览文件 @
d57572e9
##Profile Memory
##
Profile Memory
It is generally a good idea to visualize the memory usage in timeline.
It allows you to see the memory consumption of each GPU over time.
...
...
tensorflow/core/profiler/g3doc/profile_model_architecture.md
浏览文件 @
d57572e9
##Profile Model Architecture
##
Profile Model Architecture
*
[
Profile Model Parameters
](
#profile-model-parameters
)
*
[
Profile Model Float Operations
](
#profile-model-float-operations
)
###Profile Model Parameters
###
Profile Model Parameters
<b>
Notes:
</b>
`VariableV2`
operation type might contain variables created by TensorFlow
...
...
@@ -39,9 +39,9 @@ param_stats = tf.profiler.profile(
sys.stdout.write('total_params: %d\n' % param_stats.total_parameters)
```
###Profile Model Float Operations
###
Profile Model Float Operations
####Caveats
####
Caveats
For an operation to have float operation statistics:
...
...
tensorflow/core/profiler/g3doc/profile_time.md
浏览文件 @
d57572e9
##Profile Time
##
Profile Time
*
[
Times in TensorFlow and tfprof
](
#times-in-tensorflow-and-tfprof
)
*
[
Profile by Python Code
](
#profile-by-python-code
)
...
...
@@ -7,7 +7,7 @@
*
[
Profile by Name Scope
](
#profile-by-name-scope
)
###Times in TensorFlow and tfprof
###
Times in TensorFlow and tfprof
When we run a model, Tensorflow schedules and runs the nodes (operations)
in the graph. An operation can be placed on an accelerator or on CPU.
...
...
@@ -37,7 +37,7 @@ When an operation is placed on CPU, it will completely run on CPU. Hence,
should be 0.
###Profile by Python Code
###
Profile by Python Code
```
python
# In code view, the time of each line of Python code is the aggregated
# times of all operations created by that line.
...
...
@@ -112,7 +112,7 @@ Set ```-output timeline:outfile=<filename>``` to generate timeline instead of st
</left>
###Profile by Operation Type
###
Profile by Operation Type
```
python
# In op view, you can view the aggregated time of each operation type.
tfprof
>
op
-
select
micros
,
occurrence
-
order_by
micros
...
...
@@ -138,7 +138,7 @@ MatMul 618.97ms (63.56%, 16.51%), |/job:worker/replica:0/
```
###Profile by Graph
###
Profile by Graph
Usually, use graph view to generate a timeline to visualize the result.
...
...
@@ -163,7 +163,7 @@ Open a Chrome browser, enter URL chrome://tracing and load the timeline file.
******************************************************
```
###Profile by Name Scope
###
Profile by Name Scope
Usually scope view allows you to pin point the problematic places if you
have properly named your operations with tf.name_scope or tf.variable_scope.
...
...
tensorflow/docs_src/install/install_linux.md
浏览文件 @
d57572e9
...
...
@@ -151,10 +151,10 @@ Take the following steps to install TensorFlow with Virtualenv:
(tensorflow)$ <b>pip install --upgrade tensorflow-gpu</b> # for Python 2.7 and GPU
(tensorflow)$ <b>pip3 install --upgrade tensorflow-gpu</b> # for Python 3.n and GPU</pre>
If the preceding command succeeds, skip Step
5
. If the preceding
command fails, perform Step
5
.
If the preceding command succeeds, skip Step
6
. If the preceding
command fails, perform Step
6
.
5.
(Optional) If Step 4
failed (typically because you invoked a pip version
6.
(Optional) If Step 5
failed (typically because you invoked a pip version
lower than 8.1), install TensorFlow in the active virtualenv environment
by issuing a command of the following format:
...
...
tensorflow/docs_src/install/install_windows.md
浏览文件 @
d57572e9
...
...
@@ -71,12 +71,14 @@ Use that package at your own risk.
## Installing with native pip
If
the following version
of Python is not installed on your machine,
If
one of the following versions
of Python is not installed on your machine,
install it now:
*
[
Python 3.5.x 64-bit from python.org
](
https://www.python.org/downloads/release/python-352/
)
*
[
Python 3.6.x 64-bit from python.org
](
https://www.python.org/downloads/release/python-362/
)
Note that Python 3.5.x comes with the pip3 package manager, which is the
-TensorFlow supports Python 3.5.x and 3.6.x on Windows.
Note that Python 3 comes with the pip3 package manager, which is the
program you'll use to install TensorFlow.
To install TensorFlow, start a terminal. Then issue the appropriate
...
...
tensorflow/examples/speech_commands/README.md
浏览文件 @
d57572e9
# Speech Commands Example
This is a basic speech recognition example. For more information, see the
tutorial at http
://tensorflow.org
/tutorials/audio_recognition.
tutorial at http
s://www.tensorflow.org/versions/master
/tutorials/audio_recognition.
tensorflow/python/feature_column/feature_column.py
浏览文件 @
d57572e9
...
...
@@ -2473,7 +2473,7 @@ class _IndicatorColumn(_DenseColumn,
weighted_column
=
sparse_ops
.
sparse_merge
(
sp_ids
=
id_tensor
,
sp_values
=
weight_tensor
,
vocab_size
=
self
.
_variable_shape
[
-
1
]
)
vocab_size
=
int
(
self
.
_variable_shape
[
-
1
])
)
return
sparse_ops
.
sparse_tensor_to_dense
(
weighted_column
)
dense_id_tensor
=
sparse_ops
.
sparse_tensor_to_dense
(
...
...
tensorflow/python/feature_column/feature_column_test.py
浏览文件 @
d57572e9
...
...
@@ -3206,6 +3206,20 @@ class IndicatorColumnTest(test.TestCase):
with
_initialized_session
():
self
.
assertAllEqual
([[
0
,
0
,
1
],
[
1
,
0
,
0
]],
indicator_tensor
.
eval
())
def
test_transform_with_weighted_column
(
self
):
# Github issue 12557
ids
=
fc
.
categorical_column_with_vocabulary_list
(
key
=
'ids'
,
vocabulary_list
=
(
'a'
,
'b'
,
'c'
))
weights
=
fc
.
weighted_categorical_column
(
ids
,
'weights'
)
indicator
=
fc
.
indicator_column
(
weights
)
features
=
{
'ids'
:
constant_op
.
constant
([
'c'
,
'b'
,
'a'
],
shape
=
(
1
,
3
)),
'weights'
:
constant_op
.
constant
([
2.
,
4.
,
6.
],
shape
=
(
1
,
3
))
}
indicator_tensor
=
_transform_features
(
features
,
[
indicator
])[
indicator
]
with
_initialized_session
():
self
.
assertAllEqual
([[
6.
,
4.
,
2.
]],
indicator_tensor
.
eval
())
def
test_linear_model
(
self
):
animal
=
fc
.
indicator_column
(
fc
.
categorical_column_with_identity
(
'animal'
,
num_buckets
=
4
))
...
...
tensorflow/stream_executor/device_description.h
浏览文件 @
d57572e9
...
...
@@ -82,7 +82,7 @@ class DeviceDescription {
// Returns the limit on the number of simultaneously resident blocks
// on a multiprocessor.
const
uint64
blocks_per_core_limit
()
const
{
return
blocks_per_core_limit_
;
}
uint64
blocks_per_core_limit
()
const
{
return
blocks_per_core_limit_
;
}
// Returns the limit on the total number of threads that can be launched in a
// single block; i.e. the limit on x * y * z dimensions of a ThreadDim.
...
...
@@ -141,7 +141,7 @@ class DeviceDescription {
uint64
device_memory_size
()
const
{
return
device_memory_size_
;
}
// Returns the device's core clock rate in GHz.
const
float
clock_rate_ghz
()
const
{
return
clock_rate_ghz_
;
}
float
clock_rate_ghz
()
const
{
return
clock_rate_ghz_
;
}
// Returns whether ECC is enabled.
bool
ecc_enabled
()
const
{
return
ecc_enabled_
;
}
...
...
tensorflow/stream_executor/kernel.h
浏览文件 @
d57572e9
...
...
@@ -302,7 +302,7 @@ class KernelArgIterator {
//
// Returns a default-constructed KernelArg if there is no next argument.
KernelArg
next
()
{
KernelArg
result
;
KernelArg
result
=
{}
;
if
(
!
has_next
())
{
return
result
;
}
else
if
((
shmem_indices_iter_
!=
shmem_indices_end_
)
&&
...
...
tensorflow/tools/ci_build/update_version.py
浏览文件 @
d57572e9
...
...
@@ -276,8 +276,9 @@ def check_for_lingering_string(lingering_string):
"""Check for given lingering strings."""
formatted_string
=
lingering_string
.
replace
(
"."
,
r
"\."
)
try
:
linger_strs
=
subprocess
.
check_output
(
[
'grep'
,
'-rnoH'
,
formatted_string
,
TF_SRC_DIR
]).
split
(
"
\n
"
)
linger_str_output
=
subprocess
.
check_output
(
[
"grep"
,
"-rnoH"
,
formatted_string
,
TF_SRC_DIR
])
linger_strs
=
linger_str_output
.
decode
(
"utf8"
).
split
(
"
\n
"
)
except
subprocess
.
CalledProcessError
:
linger_strs
=
[]
...
...
tensorflow/tools/pip_package/BUILD
浏览文件 @
d57572e9
...
...
@@ -84,6 +84,7 @@ py_binary(
"//tensorflow/python/saved_model"
,
"//tensorflow/python:spectral_ops_test_util"
,
"//tensorflow/python/tools:tools_pip"
,
"//tensorflow/python/eager:eager_pip"
,
# These targets don't build on Windows yet. Exclude them for now.
# "//tensorflow/contrib/ndlstm",
# "//tensorflow/contrib/slim",
...
...
third_party/sqlite.BUILD
浏览文件 @
d57572e9
...
...
@@ -2,9 +2,9 @@
# Sqlite3 library. Provides utilities for interacting
# with sqlite3 databases.
licenses(["
notice"]) # BSD/MIT-like license
licenses(["
unencumbered"]) # Public Domain
exports_files(["LICENSE"])
#
exports_files(["LICENSE"])
cc_library(
name = "sqlite",
...
...
编辑
预览
Markdown
is supported
0%
请重试
或
添加新附件
.
添加附件
取消
You are about to add
0
people
to the discussion. Proceed with caution.
先完成此消息的编辑!
取消
想要评论请
注册
或
登录