aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
...
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138940265
* Change dimension arguments to be "axis" for NumPy ParityGravatar Andrew Selle2016-11-11
| | | | | | It is a ValueError to specify both. Add keyword argument for backward compatibility. Change: 138939121
* Internal only changesGravatar Gunhan Gulsoy2016-11-11
| | | | Change: 138937852
* Seal image_ops.py's interface.Gravatar Patrick Nguyen2016-11-11
| | | | Change: 138937459
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138937316
* Document caveats about `tf.py_func()`.Gravatar Derek Murray2016-11-11
| | | | | Also fixes a race that could occur if two threads called `tf.py_func()` concurrently. Change: 138937024
* Docstring example and formatting updatesGravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138936680
* Add separate C++ .so and Java android_library targets for tensorflow ↵Gravatar Andrew Harp2016-11-11
| | | | | | inference interface. Change: 138936420
* Re-add Makefile.in for quantization testsGravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138934832
* Extending MonitoredTrainingSession to make model saving and chief operations ↵Gravatar Mustafa Ispir2016-11-11
| | | | | | more flexible. Change: 138934639
* C API: Package rules for a C library and header files.Gravatar Asim Shankar2016-11-11
| | | | | | | The intent is to produce a libtensorflow.tar.gz as part of the release process to facilitate a leaner (no Python dependencies), smoother installation process for TensorFlow usage in other languages. Change: 138934248
* Add python wrapper for StatSummarizer.Gravatar Andrew Harp2016-11-11
| | | | Change: 138933733
* Enable C++ shape fn from python for more quantized ops.Gravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138929922
* Register a C++ test op for testing gradients of ops that have stringGravatar Vijay Vasudevan2016-11-11
| | | | | outputs, replacing the need for a ops.RegisterShape call. Change: 138927915
* Change loss_ops_test to pass in data with batch_size 1.Gravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138925660
* Switch Const to call_cpp_shape_fn, but still do work in python to avoid ↵Gravatar Vijay Vasudevan2016-11-11
| | | | | | | serializing large constants for now. Change: 138922651
* Switch FusedConvPad to use C++ shape function.Gravatar Vijay Vasudevan2016-11-11
| | | | | | Verified against python implementation using debug_python_shape_fn on optimize_for_inference_test. Change: 138921973
* Fixed checkpoint restore conditions to work with both V1 and V2 checkpoints.Gravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138921188
* Update bower dependencies to match internal dependencies.Gravatar Dan Smilkov2016-11-11
| | | | Change: 138919472
* Internal cleanup.Gravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138918418
* Switch PlaceholderWithDefault to C++ shape function.Gravatar Vijay Vasudevan2016-11-11
| | | | Change: 138915194
* Merge the hover and selection contexts into the ProjectorEventContext, migrateGravatar Charles Nicholson2016-11-11
| | | | | | all usage over. The bookmark panel now uses the ProjectorEventContext to deselect the currently-active-bookmark radio button. Change: 138914791
* Update the text of the host your own button to say "publish" and remove all ↵Gravatar Nikhil Thorat2016-11-11
| | | | | | usage of the word "demo". Change: 138914020
* Use `zero_debias=True` for Tensors in ExponentialMovingAverages class.Gravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138910410
* tfdbg: example for debugging tf-learn Estimator.fit() & minor changesGravatar Shanqing Cai2016-11-11
| | | | | | | | | | $ blaze build -c opt third_party/tensorflow/python/debug:debug_tflearn_iris && blaze-bin/third_party/tensorflow/python/debug/debug_tflearn_iris --debug --train_steps=1 Minor changes made in this CL: * Fix a bug in local_cli_wrapper related to computing fetch and feed summary strings. * Minor tweaks in debug/examples/README.md * Curses CLI: Add scroll direction information for UI clarity Change: 138910196
* Switch moreops to using c++ shape functions instead of python.Gravatar Vijay Vasudevan2016-11-11
| | | | Change: 138909864
* Add a functionality to dump verification strings of graph transfererGravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138903870
* Add a functionality to load graph from proto file to transfer graph to SOCGravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138898659
* Fix pca tooltip, which falls off the screen.Gravatar Dan Smilkov2016-11-11
| | | | Change: 138895487
* Fix QuantizedConcat shape function. Enable it from python as well.Gravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138895028
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138893836
* Add a note in queue functions that they create epochs local variable.Gravatar Illia Polosukhin2016-11-11
| | | | Change: 138890126
* Adding some useful dependenciesGravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138887760
* Internal cleanup.Gravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138882350
* tfdbg: put session_debug_test in class hierarchy in preparation for gRPC ↵Gravatar Shanqing Cai2016-11-11
| | | | | | | | | | | debug URLs Refactor session_debug_test to avoid code duplication when gRPC debug URLs are added. session_debug_test.py is moved to session session_debug_testlib.py, which holds a base class for the tests. A new file "session_debug_file_test.py" subclasses this base class and implements file:// URLs. Once gRPC debug URLs are added, session_debug_grpc_test.py subclass it with grpc:// debug URLs and run the same set of tests with grpc:// debug URLs. Change: 138881634
* Internal-only change.Gravatar Andrew Harp2016-11-11
| | | | Change: 138881443
* Fix another TODO: Enable gradient tests of matrix_triangular_solve for ↵Gravatar A. Unique TensorFlower2016-11-11
| | | | | | float32, which now passes (with a slightly larger tolerance) after setting the initial value properly. Change: 138881184
* Add a URL param ?config=http://path/to/projector/config.json in demo mode ↵Gravatar Nikhil Thorat2016-11-11
| | | | | | | | | | | | where users can host their own config. This also adds a UI with a template for building a projector config and instructions on how to host it. Also moves <button> over to <paper-button> in the data panel. Change: 138879601
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-11-11
| | | | Change: 138875050
* Ensuring that each test that accesses a local directory uses a separate ↵Gravatar Nathan Silberman2016-11-11
| | | | | | directory. Change: 138874226
* Increasing timeout threshold in test.Gravatar Nathan Silberman2016-11-11
| | | | Change: 138873970
* Bookmarks save and restore the current data set filter ("Isolate N Points").Gravatar Charles Nicholson2016-11-11
| | | | | | | Inspector panel restores its UI state from bookmark. It also uses hover + selection contexts so the projector doesn't need to forward events to it anymore. Eventually it will be able to forget entirely about the projector, and just use the to-be-built DataModel. Fix unrelated bug in the newly-refactored orbit controls code: restore left-button-click to pan when orthographic projections are enabled. Change: 138872537
* Ensuring that tests do not interact with one another.Gravatar Nathan Silberman2016-11-10
| | | | Change: 138838329
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-11-10
| | | | Change: 138838191
* Feeder: a helper for feeding remote queues.Gravatar A. Unique TensorFlower2016-11-10
| | | | Change: 138835911
* Adding top k evaluation metric for TensorForest.Gravatar A. Unique TensorFlower2016-11-10
| | | | Change: 138835761
* Remove identical code folding from Android linkopts.Gravatar Andrew Harp2016-11-10
| | | | Change: 138833136
* Implenebting basic projetion operation for WALS factorization ops. This allowsGravatar A. Unique TensorFlower2016-11-10
| | | | | | | | the projection from row/column feature vectors to row/column embedding factors from a trained model. In this implementations, the feature weights are always from that of the model, while the projection weights (i.e. example weights) can be specified. Change: 138832522
* Add a transformation function for tf.learn.Gravatar A. Unique TensorFlower2016-11-10
| | | | Change: 138831849
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-11-10
| | | | Change: 138830335
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-03 14:37:46 +0000