Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Mar 11, 2021

• [HLO:MLIR] Make binary op type reification emit shape_of instead of tensor ops

  Show description for be34b71
  

  d0k

  authored and

  tensorflower-gardener

  committed
  be34b71

  Copy full SHA for be34b71

• compat: Update forward compatibility horizon to 2021-03-11

  Show description for dbbc241
  tensorflower-gardener

  committed
  dbbc241

  Copy full SHA for dbbc241

• Update GraphDef version to 702.

  Show description for eeff430
  tensorflower-gardener

  committed
  eeff430

  Copy full SHA for eeff430

• Fix compiling cuda_blas.h with cuda11.

  Show description for c3f1039
  tensorflower-gardener

  committed
  c3f1039

  Copy full SHA for c3f1039

• Add FTRL support in the tpu embedding v2 optimizers API.

  Show description for 6b108b7
  tensorflower-gardener

  committed
  6b108b7

  Copy full SHA for 6b108b7

• Fix tf.data build on Windows

  Show description for 74d3466
  tensorflower-gardener

  committed
  74d3466

  Copy full SHA for 74d3466

• Support 3 different TFLite op resolvers in TfLite Python Interpreter APIs:

  Show description for 6886228
  

  multiverse-tf

  authored and

  tensorflower-gardener

  committed
  6886228

  Copy full SHA for 6886228

• Remove duplicated quantization property.

  Show description for 915ae15
  

  jianlijianli

  authored and

  tensorflower-gardener

  committed
  915ae15

  Copy full SHA for 915ae15

• [tf.data] Support bounding FunctionHandleCache capacity and extend it with LRU policy. The new functionality is used in ToSingleElementOp to enable re-use of function instantiations across the op e…

  Show description for 7244c34
  tensorflower-gardener

  committed
  7244c34

  Copy full SHA for 7244c34

• Add GibibytesPerSecond in math_utils.

  Show description for 8ac5262
  

  Jiho Choi

  authored and

  tensorflower-gardener

  committed
  8ac5262

  Copy full SHA for 8ac5262

• [XLA/GPU] Avoid using cuLink* APIs when the compilation parallelism is limited to 1.

  Show description for 483e9bc
  

  timshen91

  authored and

  tensorflower-gardener

  committed
  483e9bc

  Copy full SHA for 483e9bc

• Ensure that trivial types are passed by value rather than reference

  Show description for ae2ee47
  tensorflower-gardener

  committed
  ae2ee47

  Copy full SHA for ae2ee47

• Add iOS platform profiler.

  Show description for eaf15c4
  

  mihaimaruseac

  authored and

  tensorflower-gardener

  committed
  eaf15c4

  Copy full SHA for eaf15c4

• Fix tf.data build on Windows

  Show description for 566876a
  

  mihaimaruseac

  authored and

  tensorflower-gardener

  committed
  566876a

  Copy full SHA for 566876a

• [tf.data] Support bounding FunctionHandleCache capacity and extend it with LRU policy. The new functionality is used in ToSingleElementOp to enable re-use of function instantiations across the op e…

  Show description for 29ad14b
  

  jsimsa

  authored and

  tensorflower-gardener

  committed
  29ad14b

  Copy full SHA for 29ad14b

• Remove no-op pylint disable comments.

  Show description for ca9ce4a
  

  yilei

  authored and

  tensorflower-gardener

  committed
  ca9ce4a

  Copy full SHA for ca9ce4a

• Introduce hash table legalization pass to implement flex fallback

  Show description for f77a48e
  

  abattery

  authored and

  tensorflower-gardener

  committed
  f77a48e

  Copy full SHA for f77a48e

• Remove extra quotes which break the build

  Show description for 62cf804
  

  mihaimaruseac

  authored and

  tensorflower-gardener

  committed
  62cf804

  Copy full SHA for 62cf804

• Set TextVectorization.pad_to_max_tokens default to False

  Show description for 1e27427
  

  mattdangerw

  authored and

  tensorflower-gardener

  committed
  1e27427

  Copy full SHA for 1e27427

• Enable new quantizer by default in `tflite_convert`.

  Show description for af389cd
  

  teijeong

  authored and

  tensorflower-gardener

  committed
  af389cd

  Copy full SHA for af389cd

• [XLA] NFC: wrap templatized class around the functions used by HloEvaluator::HandleFft to easily switch the type used by the internal FFT computations (complex128 or complex64).

  Show description for cfacec1
  tensorflower-gardener

  committed
  cfacec1

  Copy full SHA for cfacec1

• # [TF:TRT] Add GPU memory limitation in model tests.

  Show description for 58768cb
  tensorflower-gardener

  committed
  58768cb

  Copy full SHA for 58768cb

• Update Python flatbuffers dependency

  Show description for cf0634e
  

  mihaimaruseac

  authored and

  tensorflower-gardener

  committed
  cf0634e

  Copy full SHA for cf0634e

• [tf.data] Code clean up to match other benchmark format.

  Show description for b916ff7
  

  jayxiaojieshi

  authored and

  tensorflower-gardener

  committed
  b916ff7

  Copy full SHA for b916ff7

• Expose `select-user-tf-ops` flag in tf_tfl_translate tool so that it's convenient to specify list of user provided tf ops.

  Show description for ebbd378
  

  haozha111

  authored and

  tensorflower-gardener

  committed
  ebbd378

  Copy full SHA for ebbd378

• Go: Update generated wrapper functions for TensorFlow ops.

  Show description for 4dcc714
  tensorflower-gardener

  committed
  4dcc714

  Copy full SHA for 4dcc714

• [XLA] Fix output sharding of passthrough scatter partitioning.

  Show description for 8ae3f78
  

  Marcello Maggioni

  authored and

  tensorflower-gardener

  committed
  8ae3f78

  Copy full SHA for 8ae3f78

• Add iOS platform profiler.

  Show description for 3a358d4
  

  Juho Ha

  authored and

  tensorflower-gardener

  committed
  3a358d4

  Copy full SHA for 3a358d4

• Fix OSS Keras API export issue for v1 endpoints.

  Show description for 64442a5
  

  qlzh727

  authored and

  tensorflower-gardener

  committed
  64442a5

  Copy full SHA for 64442a5

• Fixed profiler.start_server() documentation where port must be passed as int not string.

  Show description for 388ad7e
  

  yisitu

  authored and

  tensorflower-gardener

  committed
  388ad7e

  Copy full SHA for 388ad7e

• Fix macos py39 build

  Show description for 4e134b5
  

  mihaimaruseac

  authored and

  tensorflower-gardener

  committed
  4e134b5

  Copy full SHA for 4e134b5

• [XLA:GPU] Introduce NCCL based collective permute.

  Show description for b503269
  

  jurahul

  authored and

  tensorflower-gardener

  committed
  b503269

  Copy full SHA for b503269

• Split bins into num_bins and bin_boundaries arguments for discretization

  Show description for d2ba4e9
  

  mattdangerw

  authored and

  tensorflower-gardener

  committed
  d2ba4e9

  Copy full SHA for d2ba4e9

• Update ops-related pbtxt files.

  Show description for 888eae6
  tensorflower-gardener

  committed
  888eae6

  Copy full SHA for 888eae6

• Add ODS for tf.ReduceDataset.

  Show description for f11fdbb
  

  haozha111

  authored and

  tensorflower-gardener

  committed
  f11fdbb

  Copy full SHA for f11fdbb

Pagination

Previous

Next