Gitiles
Code Review
Sign In
gerrit-public.fairphone.software
/
platform
/
external
/
tensorflow
/
3f8506e6057f189d977697f8d0a152d6afbee7c7
3f8506e
Delete custom.md
by Deepak B
· 6 years ago
1750163
Added new issue templates
by Deepak B
· 6 years ago
0e19321
Delete bug_template.md
by Deepak B
· 6 years ago
3930c0c
Create bug_template.md
by Deepak B
· 6 years ago
a45b907
Create ISSUES.md
by Deepak B
· 6 years ago
6f0357c
Allow the "del" statement in static analysis.
by Dan Moldovan
· 6 years ago
a1e4360
Merge pull request #21559 from accraze:patch-2
by TensorFlower Gardener
· 6 years ago
15a8a88
Rely on call op placement for the default device inside a graph function
by Allen Lavoie
· 6 years ago
5d5b026
Merge pull request #22839 from georgepaw:configure_xla_macos
by TensorFlower Gardener
· 6 years ago
eddd87f
Move macros to cc from headers. This will allow us to undef the macros.
by A. Unique TensorFlower
· 6 years ago
2b97643
Add support for tracking deleted variables in activity analysis. Refactor the field names to be more consistent. Remove the unused `returned` field.
by Dan Moldovan
· 6 years ago
e4211ac
Fix Keras support in Python 3.
by Russell Power
· 6 years ago
e48365c
Update tf.keras version number.
by Francois Chollet
· 6 years ago
bbc543d
[tf.data: NUMA-map-and-batch]: Synchronize access to internal state.
by Brennan Saeta
· 6 years ago
e70b97f
Add utility function to remove FunctionDef.
by Tong Shen
· 6 years ago
11e9f65
Don't use hardcoded device string when merging device specs.
by Akshay Modi
· 6 years ago
3aa5ab7
Disable tests that have data dependencies on windows.
by Gunhan Gulsoy
· 6 years ago
d354efc
Fix use-of-uninitialized-value bug for bfloat16 in MatrixBandPart.
by Peter Hawkins
· 6 years ago
a56e036
[tf.data vectorization] Add vectorizer for "Reshape"
by Rachel Lim
· 6 years ago
df9a5f9
Remove GlobalKernelRegistry (returning void*) from the op kernel API.
by Gunhan Gulsoy
· 6 years ago
6437cee
Calculate batch size correctly when create a Dataset from numpy inputs. Also add a warning message when we don't process all the input samples when we create a Dataset object from input numpy arrays.
by Anjali Sridhar
· 6 years ago
3f75030
Remove 'not trainable' int types from NCCL gradient tests.
by A. Unique TensorFlower
· 6 years ago
6a81e44
Make predicted and output types match, so equal op can run.
by Akshay Modi
· 6 years ago
397c2c5
[XLA] Expose multi-value sort through the XlaBuilder interface, and add Evaluator support.
by Michael Kuperstein
· 6 years ago
6a1f334
Fix typo
by Richard Wei
· 6 years ago
5fede0d
Make sure the ops are not created before TF fully initializes.
by Gunhan Gulsoy
· 6 years ago
fa343ef
[tf.data] Performance modeling fixes.
by Jiri Simsa
· 6 years ago
79bc195
Merge pull request #22848 from gelembungsahabat:master
by TensorFlower Gardener
· 6 years ago
fc2b559
In control flow functionalization, mark generated If/While with outside compilation attribute if necessary.
by Tong Shen
· 6 years ago
43b0431
Move ParseHostComputeCore to side_effect_util.cc. No functionality change.
by Tong Shen
· 6 years ago
3fd902b
Validate the control flow information after subgraph rewrite, since the rewrite may remove some special nodes (e.g. TPUReplicateMetadata) which break the rule.
by A. Unique TensorFlower
· 6 years ago
2f509fe
Convert TensorFlow's highwayhash dependency to new third party import method.
by A. Unique TensorFlower
· 6 years ago
2b279ed
[XLA] More module proto verification added.
by A. Unique TensorFlower
· 6 years ago
55cf8c0
Don't segfault if unable to convert non-rectangular inputs to tensor
by Akshay Modi
· 6 years ago
028410c
[Resubmit][XLA] Introduce input/output alias config.
by Yunxing Dai
· 6 years ago
c304bd9
Add Unimplemented visitor for token element types to HLO evaluator.
by Mark Heffernan
· 6 years ago
5c286cf
Register int64 version of AddN on GPU.
by A. Unique TensorFlower
· 6 years ago
e0cda45
Merge pull request #22858 from hsgkim:volume_patches
by TensorFlower Gardener
· 6 years ago
779746b
Symmetric quantization support for tf.contrib.quantize.
by A. Unique TensorFlower
· 6 years ago
d4f6eb4
Merge pull request #22822 from Intel-tensorflow:slice_bug_fix
by TensorFlower Gardener
· 6 years ago
20f0338
Automated rollback of commit 0d054f20851f6156b1af26c35e68f6083bca8e13
by Gunhan Gulsoy
· 6 years ago
4c5a810
[Java]: Release 1.12.0-rc0
by Asim Shankar
· 6 years ago
7f8a7f2
Modify converted_call to accept a separate argument when the callable is an attribute, representing the attribute's owner.
by Dan Moldovan
· 6 years ago
53181bd
Make sure that all operands and outputs of Sort have the same layout.
by Adrian Kuegel
· 6 years ago
2e1f603
Internal change.
by Tong Shen
· 6 years ago
2502702
repo.bzl: Increase timeout of repo_ctx.execute
by A. Unique TensorFlower
· 6 years ago
41e37ae
Use the functional operator pattern for logical expressions. The current implementation relied on the TF import, which is now gone.
by Dan Moldovan
· 6 years ago
7b313d3
Make sure no objects are used before TF module gets initialized.
by Gunhan Gulsoy
· 6 years ago
c84976b
Java support for initializing tensorflow libraries.
by A. Unique TensorFlower
· 6 years ago
78d2d95
Update all the RNN tests to also run in eager context.
by Scott Zhu
· 6 years ago
0d054f2
Have TensorFlow opensource code use tensorflow_estimator module if available.
by Michael Case
· 6 years ago
ac54c45
Account for the possibility of nested lambda functions.
by Dan Moldovan
· 6 years ago
7ac4860
Raise the VLOG level for CalculateTensorSize information.
by A. Unique TensorFlower
· 6 years ago
0b1ce4d
Automated rollback of commit 9b558126e31d25ec4e82cb4f50033d6eca44349a
by A. Unique TensorFlower
· 6 years ago
085c3e4
Update downloadable clang to r343880
by Ilya Biryukov
· 6 years ago
4c08012
Automated rollback of commit e5537748753491f803fbddebdcb1cdb710631db9
by A. Unique TensorFlower
· 6 years ago
5eb3d92
Disallow PRED as element type for arithmetic operations in HloVerifier
by A. Unique TensorFlower
· 6 years ago
b2ab2da
Support arbitrary many values in KeyValueSort on GPU backend.
by Adrian Kuegel
· 6 years ago
b55ff87
Internal change.
by A. Unique TensorFlower
· 6 years ago
fe18d06
Updated model and test data for microcontroller speech example
by Pete Warden
· 6 years ago
c77588f
Don't use xla::Add on PRED types in tf-xla bridge for TileOp
by A. Unique TensorFlower
· 6 years ago
6280f41
Only file changed was the tensorcore outfeed manager test, which was passing wrong shape.
by A. Unique TensorFlower
· 6 years ago
100f9de
compat: Update forward compatibility horizon to 2018-10-11
by A. Unique TensorFlower
· 6 years ago
2e75c63
Query on Windows as well
by Grzegorz Pawelczak
· 6 years ago
96a6333
Part 3/4 of the update of tf.keras to the 2.2.4 API.
by Francois Chollet
· 6 years ago
2b8f592
Add Range to schema.
by A. Unique TensorFlower
· 6 years ago
0f79b5f
Update momentum.py
by Andy Craze
· 6 years ago
905deea
Automated rollback of commit 3abfe2cd9befa263de57edfae7d4c0d29c9c9182
by A. Unique TensorFlower
· 6 years ago
9153b89
Fp16 LSTMBlocKCell and LSTMBlockFusedCell
by James Qin
· 6 years ago
9ef04f5
Disable signed-compare warnings in TensorFlow default build.
by Justin Lebar
· 6 years ago
073c727
[XLA:GPU] Use HloVerifiedModule in cudnn_conv_pad_for_tensor_cores_test.
by Justin Lebar
· 6 years ago
d8ebd78
Merge pull request #22851 from scttl:mkl-bazelrc-build-fixes
by TensorFlower Gardener
· 6 years ago
ba2d5c3
Automated rollback of commit 9bad98c61f27b60152119bb1c2cfd402c3bf7f3d
by Jiri Simsa
· 6 years ago
4e97b77
[XLA::GPU] Explicitly use {lhs,rhs}_contracting_dimension from DotDimensionNumbers as the reduction dimension for Dot.
by Kay Zhu
· 6 years ago
f043aab
Fixes in CompareConstantArrays:
by A. Unique TensorFlower
· 6 years ago
d45c30f
[XLA:GPU] Allow input fusion into scatter
by Benjamin Kramer
· 6 years ago
787f16c
Getting rid of MOVING_AVERAGE_VARIABLES collection usage in quantize_ops.
by Rohan Jain
· 6 years ago
ed68320
[XLA:GPU] Rework CudnnConvPadForSpeed.
by Justin Lebar
· 6 years ago
e553774
Batched per_image_standardization
by A. Unique TensorFlower
· 6 years ago
2b010f2
Adds a simple util to build a GrapplerItem from a MetaGraphDef stored in a file.
by A. Unique TensorFlower
· 6 years ago
331683c
Include <cmath> in cost_estimator.h
by Jared Duke
· 6 years ago
3abfe2c
Allow the XRTCompile op to return the ProgramShape resulted form the XLA compilation.
by A. Unique TensorFlower
· 6 years ago
0be7b32
[XLA:GPU] Rename cudnn convolution passes.
by Justin Lebar
· 6 years ago
1289033
Turn on worker watchdog in TPUEstimator.
by Russell Power
· 6 years ago
9bad98c
[tf.data] Changing the background performance modeling to be on by default.
by Jiri Simsa
· 6 years ago
84f81b9
Add HloModule::Clone overload which takes an HloModuleConfig.
by Mark Heffernan
· 6 years ago
09e098e
Automated rollback of commit d6a3d6a8295359364c86aecc479e6392bcde0ce4
by Rachel Lim
· 6 years ago
8830837
[XLA] Don't overestimate the amount of bytes a gather reads
by Benjamin Kramer
· 6 years ago
dd03b7d
Don't crash an XRT server if a client leaks a compilation reference.
by A. Unique TensorFlower
· 6 years ago
a67fda8
Only acquire the flex delegate when flex ops present
by Jared Duke
· 6 years ago
0c28404
[TF:XLA] Don't aggregate writes to TensorArrays that aren't gradient TensorArrays. In the non-XLA case it is an error to write the same entry multiple times.
by Peter Hawkins
· 6 years ago
20c47de
Internal change
by Jared Duke
· 6 years ago
ea38b38
First draft of tf.compat.v1.wrap_function.
by Alexandre Passos
· 6 years ago
6905ffd
[XLA] Update TODO references related to all-reduce.
by A. Unique TensorFlower
· 6 years ago
4fe9033
Bump open source abseil-cpp revision to f340f773edab951656b19b6f1a77c964a78ec4c2
by A. Unique TensorFlower
· 6 years ago
a584dc9
Automated rollback of commit 6aebb0866718cae2c921e875f3fd74573ee9acc8
by Jonathan Hseu
· 6 years ago
5275bd0
create a reusable function for initializing tensorflow from tf lite.
by A. Unique TensorFlower
· 6 years ago
8e468c3
[TF CTC] Add ctc_beam_search_decoder_v2, which disables merge_repeated.
by Eugene Brevdo
· 6 years ago
ef64a86
Update to new toolchain with Clang 8.0.0 r340178.
by Yifei Feng
· 6 years ago
3fb4383
Don't force random ops (that generate data) to be on the CPU.
by Akshay Modi
· 6 years ago
Next »