提交 · 7127bed70ad6f168e34f372745b0b9edd046d208 · xxadev / tensorflow

18 7月, 2019 31 次提交
- A
  Add basic support for float FusedConvBiasActivation in the tensorrt bridge. · 7127bed7
  由 A. Unique TensorFlower 提交于 7月 17, 2019
```
The side_input and conv_input_scale input tensors are not yet supported.

PiperOrigin-RevId: 258646677
```
  7127bed7
- D
  [Keras] Use math_ops.divide() instead of math_ops.div() in learning_rate_schedule.py. · 7c9e9068
  由 Derek Murray 提交于 7月 17, 2019
```
This prevents deprecation warnings when using the public API.

PiperOrigin-RevId: 258645243
```
  7c9e9068
- S
  Delete broadcastable trait TensorFlow and TFLite tests · a34051c8
  由 Smit Hinsu 提交于 7月 17, 2019
```
The trait is defined the MLIR core repository so tests should be there.

PiperOrigin-RevId: 258645091
```
  a34051c8
- J
  Automated rollback of commit 18b2f268 · 05cd037e
  由 Jeremy Lau 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258640781
```
  05cd037e
- R
  Add mixed precision tests for Sequential/subclassed models. · 054aabf9
  由 Reed Wanderman-Milne 提交于 7月 17, 2019
```
I only added the "run_with_all_model_types" annotation to one test so that the tests don't take too long to run.

PiperOrigin-RevId: 258640310
```
  054aabf9
- D
  Disable the test_annotated_assign unit test because the annotation syntax is... · 3b1f3698
  由 Dong Lin 提交于 7月 17, 2019
```
Disable the test_annotated_assign unit test because the annotation syntax is only recognized by Python 3.6 or later

PiperOrigin-RevId: 258638033
```
  3b1f3698
- D
  Automated rollback of commit 7a2ce469 · ee5bf62b
  由 Dong Lin 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258637728
```
  ee5bf62b
- A
  Add interfaces to modify an HLO module after already being scheduled. · 544ada55
  由 A. Unique TensorFlower 提交于 7月 17, 2019
```
* Replace computation/parameter
* Remove/replace scheduled computation/instruction

PiperOrigin-RevId: 258633717
```
  544ada55
- J
  Tweak build rules in lite/java · 2bf47cbb
  由 Jared Duke 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258630122
```
  2bf47cbb
- J
  Allow specififying the nnapi accelerator in benchmark_model · d88b16b5
  由 Jared Duke 提交于 7月 17, 2019
```
Add a `--nnapi_accelerator_name` flag to benchmark_model, allowing
explicit accelerator control on Android Q+.

PiperOrigin-RevId: 258626781
```
  d88b16b5
- T
  Have `model_to_estimator` make a V2 Estimator if TF2 is enabled. · 75e3f928
  由 Thomas O'Malley 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258620905
```
  75e3f928
- A
  Allow building of MLIR converter in OSS. · 18b2f268
  由 Andrew Selle 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258618714
```
  18b2f268
- Y
  Remove no_windows tab for fixed flaky tests. · e81526ae
  由 Yuefeng Zhou 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258612058
```
  e81526ae
- A
  Improve micro_vision README to include detailed instructions for running on the · 208ff887
  由 A. Unique TensorFlower 提交于 7月 17, 2019
```
Sparkfun Edge board with a camera.  Add raw2bmp script to convert images from
log file to viewable bitmaps.

PiperOrigin-RevId: 258608334
```
  208ff887
- D
  Avoid copying a FunctionDef in CapturedFunction. · 4499d732
  由 Derek Murray 提交于 7月 17, 2019
```
The recently added `FunctionLibraryDefinition::CopyFunctionDefFrom()` method performs a more efficient shallow copy.

PiperOrigin-RevId: 258604819
```
  4499d732
- T
  Merge pull request #30053 from yongtang:30040-BinaryCrossEntropy · 11b33344
  由 TensorFlower Gardener 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258598105
```
  11b33344
- P
  Enable roundtrip of all TF tests through MLIR using the with_mlir_support flag. · ca146efb
  由 Prakalp Srivastava 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258597426
```
  ca146efb
- R
  Update all the TPUStrategy examples to use the new v2 APIs, i.e. · eba2610f
  由 Ruoxin Sang 提交于 7月 17, 2019
```
make_dataset_iterator -> experimental_distribute_dataset,
make_input_fn_iterator -> experimental_distribute_datasets_from_function,
unwrap -> experimental_local_results,
experimental_run -> experimental_run_v2

PiperOrigin-RevId: 258597234
```
  eba2610f
- D
  Disable test_nonlocal_symbol unit test · 75a0cbbd
  由 Dong Lin 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258596408
```
  75a0cbbd
- C
  Fix control dependencies in Keras distributed TPU loops. · 30c1d66f
  由 Chris Jones 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258596337
```
  30c1d66f
- A
  Move IsLocal() into Device and return false only for RemoteDevice. · e1f6d9e4
  由 A. Unique TensorFlower 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258591071
```
  e1f6d9e4
- E
  Add tests for TensorSpec.name · 109bad93
  由 Edward Loper 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258583855
```
  109bad93
- T
  Merge pull request #30760 from yongtang:use_resource · fc47b582
  由 TensorFlower Gardener 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258582774
```
  fc47b582
- E
  Register GPU kernel for TensorListElementShape · 7eb4f192
  由 Eugene Zhulenev 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258582149
```
  7eb4f192
- P
  Export LICENSE.TXT file. Update BUILD file. · 5361b623
  由 Prakalp Srivastava 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258582046
```
  5361b623
- T
  Update keras v2 optimizers to reuse coefficients which are shared across all... · 498df5d8
  由 Taylor Robie 提交于 7月 17, 2019
```
Update keras v2 optimizers to reuse coefficients which are shared across all updates, which reduces the total number of ops created by between 5% (for simple optimizers such as SGD and Adagrad) and 25% (for complicated optimizers such as Adam and NAdam). Separate copies are made for each device and dtype.

The effect of this change on run time is fairly minimal since Grappler is expected to consolidate most of these ops; however it does improve graph construction time.

PiperOrigin-RevId: 258581998
```
  498df5d8
- G
  [XLA GPU] Factor out helper utils for FileCheck testing into gpu_codegen_test · a9611e1c
  由 George Karpenkov 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258581066
```
  a9611e1c
- P
  Checks shape compatibility between `initial_value` and `shape` in ResourceVariable's constructor. · cbb4561c
  由 Peng Wang 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258580730
```
  cbb4561c
- T
  Merge pull request #29939 from terryky:fix_tflite_detect_postprocess · f34d5713
  由 TensorFlower Gardener 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258579478
```
  f34d5713
- J
  Update GPU delegate documentation · 5841e357
  由 Jared Duke 提交于 7月 17, 2019
```
The delegate is no longer experimental.

PiperOrigin-RevId: 258578637
```
  5841e357
- J
  Add documentation for building/using the TensorFlow Lite AAR · fce56a77
  由 Jared Duke 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258578415
```
  fce56a77
17 7月, 2019 9 次提交
- P
  Workaround for CI build problems with Arduino nightly · 1c50e736
  由 Pete Warden 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258577336
```
  1c50e736
- D
  Avoid converting the old-style print statement, because it would otherwise generate a syntax error. · 617f1011
  由 Dan Moldovan 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258575670
```
  617f1011
- I
  Make LinearOperatorDiag tape safe. Also add a standard "check_tape_safe"... · 9eabc5c6
  由 Ian Langmore 提交于 7月 17, 2019
```
Make LinearOperatorDiag tape safe.  Also add a standard "check_tape_safe" method in the base test class.

PiperOrigin-RevId: 258570322
```
  9eabc5c6
- J
  Add int16 support to Quant. · f0d6424d
  由 Jian Li 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258563058
```
  f0d6424d
- T
  Merge pull request #30739 from feihugis:Refactor_DatasetOps_9 · 4fd66235
  由 TensorFlower Gardener 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258562390
```
  4fd66235
- C
  Fix a compilation error when defining "TFLITE_CUSTOM_OPS_HEADER" macro. · c6948c48
  由 Chao Mei 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258548017
```
  c6948c48
- A
  Add tests for transpose convolution with constant inputs · 66e9e93f
  由 A. Unique TensorFlower 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258547874
```
  66e9e93f
- T
  Merge pull request #29224 from houtoms:pr_fast_3dy_reduce · c6352706
  由 TensorFlower Gardener 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258539769
```
  c6352706
- C
  Update MLIR revision. · 229bc846
  由 Christian Sigg 提交于 7月 17, 2019
```
PiperOrigin-RevId: 258533317
```
  229bc846

xxadev / tensorflow 与 Fork 源项目一致

xxadev / tensorflow
与 Fork 源项目一致