Change 109730179
Add support for selecting partition strategy in tf.nn.embedding_lookup and related ops, and allow unequally-sized shards to be used as input.
Change 109729548
TensorFlow: add RELEASE.md notes for 0.6.0.
Change 109728185
Make seq2seq_test non-flaky by setting python and numpy random seed.
Change 109725913
Refactor slot creation in optimizers and moving averages to separate file
Change 109718024
TensorFlow: reduce runtime of seq2seq_test from ~30s to ~18s.
Change 109712251
More performance improvement for convnet on GPU.
+ Switch forward convolution format to NCHW.
+ Allocate scratch space for forward- and backward- convolutions.
+ Users can use "TF_CUDNN_WORKSPACE_LIMIT_IN_MB" to configure the scratch space
limit. The default limit in 1GB.
Change 109710898
Added extract_sub_graph utility function
Base CL: 109731609
Change 109695551
Update FAQ
Change 109694725
Add a gradient for resize_bilinear op.
Change 109694505
Don't mention variables module in docs
variables.Variable should be tf.Variable.
Change 109658848
Adding an option to create a new thread-pool for each session.
Change 109640570
Take the snapshot of stream-executor.
+ Expose an interface for scratch space allocation in the interface.
Change 109638559
Let image_summary accept uint8 input
This allows users to do their own normalization / scaling if the default
(very weird) behavior of image_summary is undesired.
This required a slight tweak to fake_input.cc to make polymorphically typed
fake inputs infer if their type attr is not set but has a default.
Unfortunately, adding a second valid type to image_summary *disables* automatic
implicit conversion from np.float64 to tf.float32, so this change is slightly
backwards incompatible.
Change 109636969
Add serialization operations for SparseTensor.
Change 109636644
Update generated Op docs.
Change 109634899
TensorFlow: add a markdown file for producing release notes for our
releases. Seed with 0.5.0 with a boring but accurate description.
Change 109634502
Let histogram_summary take any realnumbertype
It used to take only floats, not it understands ints.
Change 109634434
TensorFlow: update locations where we mention python 3 support, update
them to current truth.
Change 109632108
Move HSV <> RGB conversions, grayscale conversions, and adjust_* ops back to tensorflow
- make GPU-capable version of RGBToHSV and HSVToRGB, allows only float input/output
- change docs to reflect new size constraints
- change HSV format to be [0,1] for all components
- add automatic dtype conversion for all adjust_* and grayscale conversion ops
- fix up docs
Change 109631077
Improve optimizer exceptions
1. grads_and_vars is now a tuple, so must be wrapped when passed to format.
2. Use '%r' instead of '%s' for dtype formatting
Base CL: 109697989
Change 109628097
Fix gcc 4.8.1 compile
Modified from patch by @assolini here:
https://github.com/tensorflow/tensorflow/issues/405
Change 109624275
Make preview frame ImageReader global so that it does not get GC'd.
This may fix an issue with connecting to the camera on some devices where the underlying Surface is prematurely cleaned up (http://stackoverflow.com/questions/33437961/android-camera-2-api-bufferqueue-has-been-abandoned).
Change 109620599
- improved test a little to make it easier to understand as it serves as an
example for users
Change 109614953
TensorFlow: update tutorials/howtos to point to correct
location of files, show python example in addition to bazel.
Change 109612732
TensorFlow: move reading_data into examples, change data dir
to /tmp/data. Validated that they all run, but these
probably need a selftest at some point.
Change 109608695
Apply 'gate_gradients' only when there is more than one real gradients.
Change 109605014
There are 3 obvious places to start using TensorFlow. 2/3 of the starting points do not have a link to the installation instructions.
Change 109604287
Make the `tf.reshape` shape function more restrictive.
Previously, it did not raise a construction-time error if the input
shape and the new shape were incompatible; now it detects this and
raises a `ValueError`.
Change 109603375
TensorFlow: Move word2vec_basic.py from g3doc/ to examples/
There are no additional libraries this uses, so nothing else
needs to be done
Change 109601289
TensorBoard tag 3
Change 109600908
Decrease number of scalar values stored by TensorBoard.
10k is more than displays nicely.
Change 109599464
Fix "smart restart" functionality in TensorBoard (it throws away dead data)
After restarts, a file_version event is created that always has step 0.
We need to ignore this.
Change 109597667
Switch to using /dev/urandom for TensorFlow randomness.
Using /dev/random leads to slowdown when running in an environment
with poor access to an entropy source (such as some VMs). /dev/urandom
has more predictable performance, and we don't require
cryptographically secure random number generation, so a PRNG is good
enough.
Also removes the use of the RNG in DirectSession construction. This
was being used to generate a session handle, which is not necessary
(since a DirectSession owns its devices, we don't need a unique handle
to key the OpSegment objects registered with the various devices).
This addresses bugs that have been reported on the mailing list and
Stack Overflow.
Change 109596906
Add an is_unsigned property to dtype
Change 109596830
Remove unnecessary fill in clip_by_value
Change 109591880
Remove Android demo's libpthread.so dummy file (required by protobuf) from repo and generate it at compile-time.
This makes the Android demo more portable, as the generated file will now always be the correct archictecture for linking.
Change 109589028
Isolating out the RTTI part of TensorFlow and add non-RTTI backups for Android.
This saves about 400KB of the compiled library, when compiling the Android
tensorflow target with -fno-rtti.
Change 109589018
Internal reworking of LSTMCell.
Change 109588229
Allow bool-valued tensors to be persisted.
Change 109577175
TensorBoard host defaults to 0.0.0.0
Change 109551438
TensorFlow: move mnist g3doc tutorials into tensorflow/examples.
Update examples to point to the correct location.
Adds tests to make sure they don't regress, do some lint cleanup.
Base CL: 109630240
Very minor fix to clean up comment, erased blank line (line 209 basic_usage.md)
Added a session close the Interactive Session in the 'Getting Started' tutorial
-so that no exceptions are thrown when a user tests later examples in the same python kernel.
Highlighted 't[i,j,k]' for read consistency in tutorial (line 21, resources/dim_types.md)
Change-Id: I9c0d8a6f6651bbb6a00bddecc0977b512e394669
Changes:
* Add port mapping options to `docker run` commands in
docker_run_gpu.sh and README.md. Docker on Mac needs an extra option
`-p` to expose 8888 for Jupyter Notebook. The added option makes the
behavior of containers consistent on both Ubuntu and Mac.
Change-Id: I29002329f08d7dc05415925e9b2aedbd3f112813
Change 109537918
TensorFlow pip setup: wheel >= 0.26 for python3 pip install
Change 109505848
Fix distortion default value to 1.0 in fixed_unigram_candidate_sampler. This means we default to the actual provided unigram distribution, instead of to the uniform (as it is currently).
Change 109470494
Bugfix in gradients calculation when the ys rely on each other.
Change 109467619
Fix CIFAR-10 model to train on all the training data instead of just 80% of it. Fixes#396.
Change 109467557
Replaced checkpoint file with binary GraphDef.
Change 109467433
Updates to C++ tutorial section.
Change 109465269
TensorFlow: update documentation for tutorials to not assume use of bazel
(when possible).
Change 109462916
A tutorial for image recognition to coincide with the release of the latest Inception image classification model.
Change 109462342
Clear control dependencies in variable_scope.get_variable() when creating
ops for the initializer.
Add tests of various error conditions.
Change 109461981
Various performance improvements in low-level node execution code paths.
Speeds up ptb_word_lm on my desktop with a Titan X from
3638 words per second to 3751 words per second (3.1% speedup).
Changes include:
o Avoided many strcmp operations per node execution and extra touches
of cache lines in executor.cc, by making all the various IsMerge,
IsSwitch, IsSend, etc. operations instead be based on an internal enum
value that is pre-computed at Node construction time, rather than doing
string comparisons against node->type_string(). We were doing about
6 such comparisons per executed node.
o Removed mutex_lock in executor.cc in ExecutorState::Process. The
lock was not needed and the comment about the iterations array being
potentially resized is not true (the iterations arrays are created
with a fixed size). Checked with yuanbyu to confirm this.
o Added new two-argument port::Tracing::ScopedAnnotation constructor
that takes two StringPiece arguments, and only concatenates them
lazily if tracing is enabled. Also changed the code in
platform/tracing.{h,cc} so that the ScopedAnnotation constructor and
the TraceMe constructor can be inlined.
o In BaseGPUDevice::Compute, used the two-argument ScopedAnnotation
constructor to avoid doing StrCat(opkernel->name(), ":",
op_kernel->type_string()) on every node execution on a GPU.
o Introduced a new TensorReference class that just holds a reference to an
underlying TensorBuffer, and requires an explicit Unref().
o Changed the EventMgr interface to take a vector of TensorReference objects
for EventMgr::ThenDeleteTensors, rather than a vector of Tensor objects.
o Used TensorReference in a few places in gpu_util.cc
o Minor: switched to using InlinedVectors in a few places to get better
cache locality.
Change 109456692
Updated the label_image example to use the latest Inception model
Change 109456545
Provides classify_image which performs image recognition on a 1000 object label set.
$ ./classify_image
giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca (score = 0.88493)
indri, indris, Indri indri, Indri brevicaudatus (score = 0.00878)
lesser panda, red panda, panda, bear cat, cat bear, Ailurus fulgens (score = 0.00317)
custard apple (score = 0.00149)
earthstar (score = 0.00127)
Change 109455002
TensorFlow: make the helper libraries for various models available
in the pip package so that when users type:
python translate.py ...
the absolute import works.
This change is supposed to help make our tutorials run without the
*need* to use bazel.
Change 109450041
TensorFlow: remove cifar and convolutional binary copies from pip install.
Adds embedding and some other models to the list.
Change 109448520
Move the description of a failing invariant from a comment into the dcheck-fail message text.
Change 109447577
TensorBoard has release tagging (tensorboard/TAG)
Also track TensorBoard changes (tensorboard/CHANGES)
Change 109444161
Added ParseSingleSequenceExample + python wrappers + unit tests.
Change 109440864
Update all the TensorFlow Dockerfiles, and simplify GPU containers.
This change updates all four of our Dockerfiles to match the targets discussed
in https://github.com/tensorflow/tensorflow/issues/149. The most notable
change here is moving the GPU images to use the NVidia containers which
include cudnn and other build-time dependencies, dramatically simplifying both
the build and run steps.
A description of which tags exist and get pushed where will be in a follow-up.
Change 109432591
Some pylint and pydoc changes in saver.
Change 109430127
Remove unused hydrogen components
Change 109419354
The RNN api, although moved into python/ops/, remains undocumented.
It may still change at any time.
Base CL: 109538006
