aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
* Add warning note to Variable.initialized_value documentation.Gravatar A. Unique TensorFlower2016-11-28
| | | | Change: 140374197
* Adds the following ops to android_core_ops:Gravatar A. Unique TensorFlower2016-11-28
| | | | | | | | "non_max_suppression_op.cc", "non_max_suppression_op.h", "one_hot_op.cc", "one_hot_op.h" Change: 140364162
* Factor out parent-weight code in TensorForest to use elsewhere.Gravatar A. Unique TensorFlower2016-11-28
| | | | Change: 140362909
* Remove stillborn 'params' argument of dynamic_rnn_estimator factory functions:Gravatar A. Unique TensorFlower2016-11-28
| | | | | | Any value other than None fails, because dynamic_rnn_estimator's model_fn does not accept a params argument. Change: 140351422
* Honor mode in _dynamic_rnn_model_fn(): no training_op if not TRAIN,Gravatar A. Unique TensorFlower2016-11-28
| | | | | no loss and eval_metrics if INFER. Without this, .export() crashes. Change: 140337700
* This introduces the RelaxedBernoulli, RelaxedOneHotCategorical, and ↵Gravatar A. Unique TensorFlower2016-11-28
| | | | | | OneHotCategorical distribution classes (along with some useful variants) to tf.contrib.distributions. These distributions were concurrently introduced by https://arxiv.org/abs/1611.00712 and https://arxiv.org/abs/1611.01144. Change: 140336235
* Fix comments.Gravatar A. Unique TensorFlower2016-11-28
| | | | Change: 140334954
* Seal rnn_cell's interface.Gravatar Patrick Nguyen2016-11-27
| | | | Change: 140308941
* Seal tf.app's interface.Gravatar Patrick Nguyen2016-11-27
| | | | Change: 140308750
* Fixing scatter_nd validation code to handle indices tensors with extra outer ↵Gravatar A. Unique TensorFlower2016-11-25
| | | | | | dimensions and adding a test. Change: 140222390
* Update ops-related pbtxt files.Gravatar A. Unique TensorFlower2016-11-25
| | | | Change: 140215986
* Support multiple delimiters for splitting strings.Gravatar A. Unique TensorFlower2016-11-25
| | | | Change: 140215361
* Documentation: defining how existing directories are handled by ↵Gravatar A. Unique TensorFlower2016-11-24
| | | | | | recursive_create_dir. Change: 140127945
* Changed StepCounterHook to use [operation name]/sec instead of hard-coded ↵Gravatar A. Unique TensorFlower2016-11-24
| | | | | | global_step/sec. Change: 140114701
* Internal-only changeGravatar Peter Hawkins2016-11-23
| | | | Change: 140098006
* Split _BinaryLogisticHead from _MultiClassHead.Gravatar A. Unique TensorFlower2016-11-23
| | | | | Combine train and eval loss fn (since they're always the same). Change: 140088698
* Internal-only change.Gravatar Peter Hawkins2016-11-23
| | | | Change: 140088388
* Fix a bug in gradient computation that got the control flow context wrong in ↵Gravatar Yuan Yu2016-11-23
| | | | | | a corner case of nested cond and while loop. Change: 140083287
* Add genrules for generating Saved Model test data for half plus two.Gravatar Li Lao2016-11-23
| | | | Change: 140081034
* Eliminate dupe loss and predictions ops.Gravatar A. Unique TensorFlower2016-11-23
| | | | | Add name to prediction ops. Change: 140080563
* Add missing dependencies and resolve circular dependencies.Gravatar Patrick Nguyen2016-11-23
| | | | Change: 140079490
* Move the debug_io_utils dependency out of the ARRAY_DEPS list,Gravatar A. Unique TensorFlower2016-11-23
| | | | | | where it included in a huge ton of targets, and into a new target just for "debug_ops" where it is needed. Change: 140079200
* Fixes bug with recent change on ResourceVariables.Gravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140073799
* Clearer error for unsupported types.Gravatar A. Unique TensorFlower2016-11-23
| | | | | Fixes #5738. Change: 140072838
* Enable cygwin in build_helper.subr, which would populate to make -j optionGravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140072478
* Add an option to pass additional copts to the kernel, e.g. ↵Gravatar A. Unique TensorFlower2016-11-23
| | | | | | tf_opts_nortti_if_android(). Change: 140071090
* Automated rollback of change 139971064Gravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140070326
* Lazy device setting for TensorArrays.Gravatar Eugene Brevdo2016-11-23
| | | | | | | | | | | | Prior to this change, TensorArrays were always created on the device set by the device scope (if any); which is not necessarily the device on which the Tensors written to the given TensorArray sit. Since TensorArrays have strong colocation requirements, this often meant expensive round-trips to write and read Tensors. With this change, TensorArray ops are created with no device set; and the first call to write/unpack/split to a TensorArray with a Tensor bound to a particular device will set the TensorArray's device to match. Change: 140067532
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140067123
* Adding -j${JOBS_COUNT} for make in build_all_${target_os_type}.sh to speed ↵Gravatar A. Unique TensorFlower2016-11-23
| | | | | | up build Change: 140066519
* Made scattered_embedding_column use SparseFeatureCrossOpV2Gravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140065785
* Introduce the FullyConnected layer class and functional wrapper.Gravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140064894
* Introduce MonitoredSession into the examples. cl #1Gravatar Mustafa Ispir2016-11-23
| | | | Change: 140063357
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140062662
* Cleanup and consistency for variable handling in RNNCells.Gravatar Eugene Brevdo2016-11-23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In the run-up to TF 1.0, we are making RNNCells' variable names compatible with those of tf layers. This is a breaking change for those who wish to reload their old RNN model checkpoints in newly created graphs. After this change is in, variables created with RNNCells will have slightly different names than before; loading old checkpoints to run with newly created graphs requires renaming at load time. Loading and executing old graphs with old checkpoints will continue to work without any problems. Creating and loading new checkpoints with graphs after this change is in will work without any problems. The only people affected by this change are those who want to load old RNN model checkpoints graphs created after this change is in. Renaming on checkpoint load can be performed with tf.contrib.framework.variables.assign_from_checkpoint. Example usage is available here[1] if you use Saver and/or Supervisor, and [2] if you are using the newer tf.learn classes. Examples of renamed parameters: LSTMCell without sharding: my_scope/LSTMCell/W_0 -> my_scope/lstm_cell/weights my_scope/LSTMCell/W_F_diag -> my_scope/lstm_cell/w_f_diag my_scope/LSTMCell/B -> my_scope/lstm_cell/biases LSTMCell with sharding: my_scope/LSTMCell/W_0 -> my_scope/lstm_cell/weights/part_0 my_scope/LSTMCell/W_1 -> my_scope/lstm_cell/weights/part_1 my_scope/LSTMCell/W_2 -> my_scope/lstm_cell/weights/part_2 my_scope/LSTMCell/W_F_diag -> my_scope/lstm_cell/w_f_diag my_scope/LSTMCell/B -> my_scope/lstm_cell/biases BasicLSTMCell: my_scope/BasicLSTMCell/Linear/Matrix -> my_scope/basic_lstm_cell/weights my_scope/BasicLSTMCell/Linear/Bias -> my_scope/basic_lstm_cell/biases MultiRNNCell: my_scope/MultiRNNCell/Cell0/LSTMCell/W_0 -> my_scope/multi_rnn_cell/cell_0/lstm_cell/weights my_scope/MultiRNNCell/Cell0/LSTMCell/W_F_diag -> my_scope/multi_rnn_cell/cell_0/lstm_cell/w_f_diag my_scope/MultiRNNCell/Cell0/LSTMCell/B -> my_scope/multi_rnn_cell/cell_0/lstm_cell/biases 1. https://github.com/tensorflow/tensorflow/blob/master/tensorflow/contrib/slim/README.md 2. https://github.com/tensorflow/tensorflow/blob/86f5ab7474825da756838b34e1b4eac93f5fc68a/tensorflow/contrib/framework/python/ops/variables_test.py#L810 Change: 140060366
* Make third_party/tensorflow/core/framework/allocator_test.cc less brittle inGravatar A. Unique TensorFlower2016-11-23
| | | | | | | | the face of changes to malloc(). This change both adjusts the numbers and adds 5kBytes of slop on either side of the numbers in the hope of avoiding trivial breakage. Change: 140060311
* BigQuery table partition accessor to be used in BigQuery Reader.Gravatar A. Unique TensorFlower2016-11-23
| | | | | | | | This code implements a component to read table partitions from BigQuery using the HTTP API for BigQuery. It specifically uses tabledata.list mehtod. This implementation uses Application Default Credentials (similar to GCS Filesystem). The current implementation does not support nested types. Change: 140058933
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140058389
* Add an `elem_shape` argument to TensorArray to allow users to specify the ↵Gravatar Yuan Yu2016-11-23
| | | | | | shape of the elements. If `elem_shape` is not None, we will check for shape equality for all writes and return elem_shape for all reads. Change: 140056114
* Makes resource variables saveable/restorable.Gravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140055398
* Comment fix.Gravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140054781
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140053628
* Internal change:Gravatar Eugene Brevdo2016-11-23
| | | | | Allow resetting colocation groups via colocate_with(None, ignore...=True) Change: 140052949
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140048087
* Add binary classification confusion matrix metrics.Gravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140046076
* Consolidate compile protobuf script to use the helper function, which bases ↵Gravatar A. Unique TensorFlower2016-11-23
| | | | | | on numbers of logical CPUs on the host system Change: 140045246
* Don't re-install dependencies if they exist on build machines.Gravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140044561
* Relax Pool3D shape function when some dimensions are unknown.Gravatar Vijay Vasudevan2016-11-23
| | | | | | | Added test. Fixes #5807. Change: 140042698
* Fix RemoveNewDefaultAttrsFromGraphDef() for graphs with functions.Gravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140034297
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2016-11-23
| | | | Change: 140030944
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-05 23:44:21 +0000