aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
* [TF:XLA] Add a placeholder implementation of Log1p (via log(1+x), which is ↵Gravatar Peter Hawkins2017-02-01
| | | | | | | not numerically accurate for x near 0). Make some cleanups to unary_ops_test.py. Change: 146282294
* Sparse repeated indices optimizer fix for resource variables.Gravatar A. Unique TensorFlower2017-02-01
| | | | | Resource variables are quite new, so this change should not affect any preexisting models. Follows up on the fix for Optimizers dealing with sparse gradients which have repeated indices. Change: 146279378
* Decrement initial reference to XlaDeviceContext object to fix memory leak.Gravatar Peter Hawkins2017-02-01
| | | | Change: 146278252
* C API: Slight tweaks to binary distribution scriptsGravatar Asim Shankar2017-02-01
| | | | | Helps with #6267 Change: 146277178
* [XLA:GPU] Make WhileTransformer match computations post fusion and copy ↵Gravatar A. Unique TensorFlower2017-02-01
| | | | | | | | | | | | | insertion HLO passes. Switches implementation to use simple expression tree matching, and reduces the number of patterns to match to those that we care about in production (i.e. post fusion and copy insertion). // ExecuteThunks runtime FRNN forward pass (sequence_length=30) WhileThunk: 112ms // ExecuteThunks runtime FRNN forward pass (sequence_length=30) ForThunk: 6ms // ExecuteThunks runtime FRNN forward pass (sequence_length=100) WhileThunk: 585ms // ExecuteThunks runtime FRNN forward pass (sequence_length=100) ForThunk: 191ms Change: 146273048
* Script for converting @{symbol} in Python guides into links.Gravatar A. Unique TensorFlower2017-02-01
| | | | Change: 146272519
* Made the event accumulator accumulate health pill valuesGravatar A. Unique TensorFlower2017-02-01
| | | | | | We create a reservoir that organizes health pill values by node name. Also added a method onto the event multiplexor to fetch the health pills per run. Change: 146272468
* Remove the "Experimental" markers of RunOptions and RunMetadata protos.Gravatar Zongheng Yang2017-02-01
| | | | | This seems like a reasonable thing to do, as (1) they have been in existence for ~1 year, (2) people have added new fields to them since their inception, and (3) the C API supports them. Change: 146271432
* Multi-thread TensorForest's inference op.Gravatar A. Unique TensorFlower2017-02-01
| | | | Change: 146268982
* Android: fix issue where mismatch between Inception output size and ↵Gravatar Andrew Harp2017-02-01
| | | | | | inception num classes could cause demo app to crash (due to recent CHECK added in inference interface). Change: 146268764
* Fix the code that looks up the hardware node stats. If an op is implemented byGravatar Yao Zhang2017-02-01
| | | | | multiple GPU kernels, aggregate the time for all kernels for that op. Change: 146268331
* Update ops-related pbtxt files.Gravatar A. Unique TensorFlower2017-02-01
| | | | Change: 146267759
* Add a mechanism for hiding, skipping, and modifying the generated opGravatar A. Unique TensorFlower2017-02-01
| | | | | | | | | | | | | | | | functions for C++. A souped-up version of the hidden_ops mechanism in Python, the intent is to use this for most or all of the client languages, with a common list of changes to make in a common file and per-language overrides. Also: * include the documentation for outputs in the generated comments * several updates to C++ API to match Python * fix C++ shape function for ConcatV2 now that we use it by default * split op_gen_lib out of core:framework, since it is only used by the op generators, and I don't want to add another proto to mobile builds Change: 146267344
* Fix another bug in computing shape for SDCA fake bias columnGravatar David Soergel2017-02-01
| | | | Change: 146266321
* Doc generator fixes:Gravatar Martin Wicke2017-02-01
| | | | | | | | | - enable contrib - show __init__ - remove core.protobuf module - fix bug in _get_arg_spec - hide contrib.learn.head Change: 146265365
* tfdbg CLI: fix a bug RichTextLines.prepend that caused font_attr_segs to be NoneGravatar Shanqing Cai2017-02-01
| | | | Change: 146260296
* Allow exporting and loading graphs with no variables.Gravatar David Soergel2017-02-01
| | | | Change: 146257820
* Fix lookup ops documentation to use correct function names for ↵Gravatar A. Unique TensorFlower2017-02-01
| | | | | | index_to_string_table_from_tensor and index_to_string_table_from_file. Change: 146255499
* Fixes bug in sparse resource variable optimization.Gravatar Alexandre Passos2017-02-01
| | | | Change: 146241206
* Fix bug in computing shape for SDCA fake bias columnGravatar David Soergel2017-02-01
| | | | Change: 146238867
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2017-02-01
| | | | Change: 146231606
* graph_editor: easier subgraph remapping of inputs/outputs.Gravatar A. Unique TensorFlower2017-02-01
| | | | Change: 146230363
* Make timestamp-based test less flaky by increasing delayGravatar David Soergel2017-01-31
| | | | Change: 146204375
* Allow passing arguments through saved_model loader to saver.import_meta_graph.Gravatar David Soergel2017-01-31
| | | | | These **kwargs are ultimately passed to meta_graph.import_scoped_meta_graph. This CL allows setting useful loading options such as import_scope and input_map. Change: 146201221
* SDCA: Stop relying on the existence of labels to get the fake bias column's ↵Gravatar A. Unique TensorFlower2017-01-31
| | | | | | desired shape. Change: 146200993
* Move `pop("self")` logic into `Distribution` baseclass.Gravatar Joshua V. Dillon2017-01-31
| | | | Change: 146198059
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2017-01-31
| | | | Change: 146197150
* BREAKING CHANGE: Rename bijector shape getters, e.g., ↵Gravatar Joshua V. Dillon2017-01-31
| | | | | | "get_forward_event_shape()" to "forward_event_shape()", "forward_event_shape()" to "forward_event_shape_tensor(), and same for "inverse" counterparts. Change: 146196054
* Updated run_and_gather_logs.py to use argparse.Gravatar A. Unique TensorFlower2017-01-31
| | | | Change: 146194766
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2017-01-31
| | | | Change: 146191509
* Don't try to constant-fold operators that do not have a CPU kernel.Gravatar Peter Hawkins2017-01-31
| | | | | Fixes spurious "Executor failed to create kernel. Not found: No registered '...' OpKernel for CPU devices" errors from the constant folder. Change: 146188668
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2017-01-31
| | | | Change: 146187659
* BREAKING CHANGE: Standardize "concentration", "rate", "total_count" ↵Gravatar Joshua V. Dillon2017-01-31
| | | | | | | | distribution arguments. BUGFIX: Correct undefined mode in dirichlet.mode. BUGFIX: Correct broadcasting in dirichletmultinomial.mean. Change: 146187147
* Update ops-related pbtxt files.Gravatar A. Unique TensorFlower2017-01-31
| | | | Change: 146186989
* Add string support to tf.zeros.Gravatar A. Unique TensorFlower2017-01-31
| | | | Change: 146185845
* Update MultiLabelHead to support sparse target labels.Gravatar A. Unique TensorFlower2017-01-31
| | | | Change: 146184013
* [XLA] Move fast-math flags into HLO module config.Gravatar Justin Lebar2017-01-31
| | | | | | | | | | | | | | | | | | | | | | Previously, XLA controlled the presence/absence of fast-math flags (FMF) via a command-line flag. This patch changes things so we use a new CompileOptions proto instead. This proto lives in HloModuleConfig, and is passed to the service via ExecuteRequest. This change lets us entirely remove llvm_backend_flags.{h,cc}. In addition, this change takes us from two to one fast-math flags. Previously we tried to control "unsafe FP transformations" separately from "full fast math". It turns out that LLVM is misleadingly inconsistent in how it handles these. In the backend, they are indeed two separate options that can be enabled/disabled independently. In the frontend, however, unsafe-fp-math implies all the other FMFs. As a result, it doesn't really make sense for XLA to attempt to split out these two flags, at least not until LLVM changes how it handles them. Change: 146183994
* Automated rollback of change 145619741Gravatar Dan Ringwalt2017-01-31
| | | | Change: 146183030
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2017-01-31
| | | | Change: 146182979
* TensorBoard: Don't render smoothed lines towards endGravatar Justine Tunney2017-01-31
| | | | Change: 146182705
* BREAKING CHANGE: Rename shape getters, e.g., "get_event_shape()" to ↵Gravatar Joshua V. Dillon2017-01-31
| | | | | | | "event_shape()", "event_shape()" to "event_shape_tensor(), and same for "batch_shape". BUGFIX: *onehot_categorical.py returns vector not scalar. Change: 146182622
* Docs: Use tf.stack instead of the deprecated tf.packGravatar Asim Shankar2017-01-31
| | | | Change: 146182278
* Changed the default size guidance values for image and audio summaries to ↵Gravatar A. Unique TensorFlower2017-01-31
| | | | | | 10. It used to be 4. We upped this number because the image dashboard now has a slider that lets the user scroll through steps. Previously, image summaries at those past steps were inaccessible. Change: 146180737
* Added a spinner that indicates when an image summary at a certain step is ↵Gravatar A. Unique TensorFlower2017-01-31
| | | | | | | loading. This avoids flickering when the user scrolls through steps of an image summary, allowing the user to readily compare images across steps. Change: 146180073
* Add a Saver test to ensure it continues working on URIs.Gravatar Jonathan Hseu2017-01-31
| | | | Change: 146179848
* Make a better doc generator.Gravatar Martin Wicke2017-01-31
| | | | Change: 146177492
* Improve performance of ResizeArea op:Gravatar A. Unique TensorFlower2017-01-31
| | | | | | | | | | | - cache computed values, hoist computations out of loops - avoid bounds checks in many cases. - access input data with pointer offsets instead of through 4-d eigen tensor (which requires more complicated index computations). - add custom accumulation fn for 3-channel images Added tests to resize_area_op_test.cc, and benchmarks to image_ops_test.py. Change: 146177262
* Update generated Python Op docs.Gravatar A. Unique TensorFlower2017-01-31
| | | | Change: 146176865
* Seal tf.contrib.framework.Gravatar Martin Wicke2017-01-31
| | | | Change: 146174108
* Make the request manager request health pills for a set of node namesGravatar A. Unique TensorFlower2017-01-31
| | | | | | | | | | | Specifically, the request manager sends a request to the /health_pills endpoint with the list of node names stored in POST data. Health pills are currently a feature internal to Google, but this feature may soon be open-sourced. Health pills are summaries of tensor element values, ie the counts of tensor element values that are -Inf, negative, 0, positive, +Inf, NaN, etc. Change: 146160274
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-10 01:16:18 +0000