Commits

Branch selector

 master

User selector

All users

All time

Commit History

Commits on Jul 23, 2020

• Support conversion of tf.text ops to Flex ops

  Show description for 537f3ec
  

  thaink

  authored and

  tensorflower-gardener

  committed
  537f3ec

  Copy full SHA for 537f3ec

• Add module comments for tf.experimental.numpy. Point Github README.md to

  Show description for e543b68
  tensorflower-gardener

  committed
  e543b68

  Copy full SHA for e543b68

• [XLA:SPMD] Improve resharding

  Show description for f59ff5d
  

  Yuanzhong Xu

  authored and

  tensorflower-gardener

  committed
  f59ff5d

  Copy full SHA for f59ff5d

• Wrap/rewrap ndarrays in tf.vectorized_map

  Show description for 488448c
  

  Akshay Modi

  authored and

  tensorflower-gardener

  committed
  488448c

  Copy full SHA for 488448c

• Integrate LLVM at llvm/llvm-project@99ad956fdaee

  Show description for ee74b70
  tensorflower-gardener

  committed
  ee74b70

  Copy full SHA for ee74b70

• Remove two unused LOG(FATAL) methods from TpuNodeContext.

  Show description for f942f4a
  

  jiawenhao

  authored and

  tensorflower-gardener

  committed
  f942f4a

  Copy full SHA for f942f4a

• Small LSTM test cleanup: Move setting lstm_golden_output_ to SetUp() for a few tests, consistent with other tests. Drop the input/output parameters of VerifyGoldens, it already sees the member vari…

  Show description for 663cd75
  

  lrdxgm

  authored and

  tensorflower-gardener

  committed
  663cd75

  Copy full SHA for 663cd75

• Merge pull request #41429 from dnguyen28061:scalar_shape

  Show description for 0b2019b
  tensorflower-gardener

  committed
  0b2019b

  Copy full SHA for 0b2019b

• Include Ops that are used via PartitionedCalls to MetaGraphDef.MetaInfoDef.stripped_op_list

  Show description for 0554618
  tensorflower-gardener

  committed
  0554618

  Copy full SHA for 0554618

• Fix documentation for int16x8 quantization

  Show description for fffa1e6
  

  MeghnaNatraj

  authored and

  tensorflower-gardener

  committed
  fffa1e6

  Copy full SHA for fffa1e6

• Disable collective ops abort tests on windows since they seem to be flaky.

  Show description for f112d4b
  

  annarev

  authored and

  tensorflower-gardener

  committed
  f112d4b

  Copy full SHA for f112d4b

• Port the circular buffer kernel to the new TfLiteEvalTensor API.

  Show description for b7ac6f4
  

  Nick Kreeger

  authored and

  tensorflower-gardener

  committed
  b7ac6f4

  Copy full SHA for b7ac6f4

• Internal change

  Show description for 4fde36c
  tensorflower-gardener

  committed
  4fde36c

  Copy full SHA for 4fde36c

• Use standard absl::BitGen instead of function unavailable in open-source absl

  Show description for aedf43f
  

  Frank Chen

  authored and

  tensorflower-gardener

  committed
  aedf43f

  Copy full SHA for aedf43f

• Fix GCC build error with tpu_on_demand_compiler.cc and erroneous inclusion

  Show description for 82e7ce3
  

  Frank Chen

  authored and

  tensorflower-gardener

  committed
  82e7ce3

  Copy full SHA for 82e7ce3

• Port the ceil kernel to the new TfLiteEvalTensor API.

  Show description for dfa281c
  

  Nick Kreeger

  authored and

  tensorflower-gardener

  committed
  dfa281c

  Copy full SHA for dfa281c

• Change Tensorboard callback to directly write out the train_function graphs at the end of the first training batch, rather than trying to write the potentially-incorrect Keras backend graph when th…

  Show description for c72c903
  

  Tomer Kaftan

  authored and

  tensorflower-gardener

  committed
  c72c903

  Copy full SHA for c72c903

• Implement a function cache that caches the lowered BEF for tf.function (graph execution). Observed ~22x performance improvement in microbenchmark compared to without cache.

  Show description for 6a98324
  

  chuanhaozhuge

  authored and

  tensorflower-gardener

  committed
  6a98324

  Copy full SHA for 6a98324

• Go: Update generated wrapper functions for TensorFlow ops.

  Show description for aa30205
  tensorflower-gardener

  committed
  aa30205

  Copy full SHA for aa30205

• Port the arg min/max kernel to the new TfLiteEvalTensor API.

  Show description for d7f7908
  

  Nick Kreeger

  authored and

  tensorflower-gardener

  committed
  d7f7908

  Copy full SHA for d7f7908

• [tf.data] Have auto_shard use regex match instead of exact match to be more robust to different versions of an op. Also adds test for MatchesAnyVersionRE utility function.

  Show description for 2f32eed
  

  Rachel Lim

  authored and

  tensorflower-gardener

  committed
  2f32eed

  Copy full SHA for 2f32eed

• fix the failure in notebook for model maker.

  Show description for c8bfbd3
  

  ziyeqinghan

  authored and

  tensorflower-gardener

  committed
  c8bfbd3

  Copy full SHA for c8bfbd3

• Remove unused functions.

  Show description for 30ff851
  

  advaitjain

  authored and

  tensorflower-gardener

  committed
  30ff851

  Copy full SHA for 30ff851

• Update ops-related pbtxt files.

  Show description for 4b5edac
  tensorflower-gardener

  committed
  4b5edac

  Copy full SHA for 4b5edac

• sanity -> consistency / smoke.

  Show description for 983d000
  

  advaitjain

  authored and

  tensorflower-gardener

  committed
  983d000

  Copy full SHA for 983d000

• Port the add kernel to the new TfLiteEvalTensor API.

  Show description for 9c433fe
  

  Nick Kreeger

  authored and

  tensorflower-gardener

  committed
  9c433fe

  Copy full SHA for 9c433fe

• [tf.data] Calculate the average input time for the root node of the data input pipeline. Also add a unit test for the `SelfProcessingTime` function.

  Show description for 969fdb0
  

  jayxiaojieshi

  authored and

  tensorflower-gardener

  committed
  969fdb0

  Copy full SHA for 969fdb0

Commits on Jul 22, 2020

• Add APIs for using different processes to register and read from tf.data service.

  Show description for 0576752
  

  aaudiber

  authored and

  tensorflower-gardener

  committed
  0576752

  Copy full SHA for 0576752

• [tf.data] Add RebatchDatasetV2. Also updates RebatchDataset(V1) tests so that they're more easily readable.

  Show description for d2ab722
  

  Rachel Lim

  authored and

  tensorflower-gardener

  committed
  d2ab722

  Copy full SHA for d2ab722

• Support integer input and output type for Quantize-Aware Trained models

  Show description for d75d7d1
  

  annarev

  authored and

  tensorflower-gardener

  committed
  d75d7d1

  Copy full SHA for d75d7d1

• Make AbstractTensorHandle RefCounted so that we can use it for refcounting tensors under a GradientTape.

  Show description for ac9c8e2
  

  saxenasaurabh

  authored and

  tensorflower-gardener

  committed
  ac9c8e2

  Copy full SHA for ac9c8e2

• By default, hide all compiler warnings.

  Show description for f9dd454
  

  gunan

  authored and

  tensorflower-gardener

  committed
  f9dd454

  Copy full SHA for f9dd454

• Test experimental_io_device option.

  Show description for 957502c
  

  kenfranko

  authored and

  tensorflower-gardener

  committed
  957502c

  Copy full SHA for 957502c

• Add MWMS test to ctl_correctness_test. Fixed an issue with reducing int values on GPUs.

  Show description for e7710bf
  

  Anjali Sridhar

  authored and

  tensorflower-gardener

  committed
  e7710bf

  Copy full SHA for e7710bf

• Merge pull request #40901 from rahul-kamat:tf-function-tracing-with-annotations

  Show description for 24ba299
  tensorflower-gardener

  committed
  24ba299

  Copy full SHA for 24ba299

Pagination

Previous

Next