Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Oct 16, 2019

• Added LSTM test

  Show description for 1f84bd0
  tensorflower-gardener

  committed
  1f84bd0

  Copy full SHA for 1f84bd0

• Add printing errors.

  Show description for 96c8f3a
  tensorflower-gardener

  committed
  96c8f3a

  Copy full SHA for 96c8f3a

• conv_test.cc code cleanup

  Show description for 0164fb1
  

  teijeong

  authored and

  tensorflower-gardener

  committed
  0164fb1

  Copy full SHA for 0164fb1

• Implement simple loop-invariant-code-motion based on dialect interfaces.

  Show description for 78a29c1
  

  Stephan Herhut

  authored and

  tensorflower-gardener

  committed
  78a29c1

  Copy full SHA for 78a29c1

• Script moved to root OpenCL directory.

  Show description for b60615f
  tensorflower-gardener

  committed
  b60615f

  Copy full SHA for b60615f

• Implement subgraph aware profiling tool

  Show description for b366763
  

  abattery

  authored and

  tensorflower-gardener

  committed
  b366763

  Copy full SHA for b366763

• compat: Update forward compatibility horizon to 2019-10-16

  Show description for 818704e
  tensorflower-gardener

  committed
  818704e

  Copy full SHA for 818704e

• Fix a pytype error in system_info_lib.

  Show description for c3d86dc
  

  akuegel

  authored and

  tensorflower-gardener

  committed
  c3d86dc

  Copy full SHA for c3d86dc

• Add a PrecisionAtRecall metric.

  Show description for 95bb950
  tensorflower-gardener

  committed
  95bb950

  Copy full SHA for 95bb950

• In `multi_process_runner`, in the case where subprocess raises an error, re-raise the exception in addition to reporting it to the parent process, so that even if `--test_timeout` flag is set and t…

  Show description for 1ecb21d
  

  rchao

  authored and

  tensorflower-gardener

  committed
  1ecb21d

  Copy full SHA for 1ecb21d

• BatchMatMul conversion bug and variable naming convention fixed.

  Show description for 8ab8e09
  tensorflower-gardener

  committed
  8ab8e09

  Copy full SHA for 8ab8e09

• Fix bug where `vectorized_map` failed on inputs with unknown static shape.

  Show description for 8694260
  tensorflower-gardener

  committed
  8694260

  Copy full SHA for 8694260

• Add a flag to the QuantSpec to indicate post-training quantization workflow

  Show description for dac8ede
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  dac8ede

  Copy full SHA for dac8ede

• Evaluation tool internal refactors

  Show description for 98a48ac
  

  thaink

  authored and

  tensorflower-gardener

  committed
  98a48ac

  Copy full SHA for 98a48ac

• Add unit tests for int8 quantized model for the benchmark tool in addition to the existing fp32-model one.

  Show description for 682d02b
  

  multiverse-tf

  authored and

  tensorflower-gardener

  committed
  682d02b

  Copy full SHA for 682d02b

• softmax multi-threading for the float path.

  Show description for 9fe3c54
  

  renjie-liu

  authored and

  tensorflower-gardener

  committed
  9fe3c54

  Copy full SHA for 9fe3c54

• Add back missing check in ConstantPushDown.

  Show description for 38c43d3
  tensorflower-gardener

  committed
  38c43d3

  Copy full SHA for 38c43d3

• Optimize scalar broadcast mul for float path.

  Show description for 3157ab6
  

  renjie-liu

  authored and

  tensorflower-gardener

  committed
  3157ab6

  Copy full SHA for 3157ab6

• Build file fix.

  Show description for 0a07914
  

  lrdxgm

  authored and

  tensorflower-gardener

  committed
  0a07914

  Copy full SHA for 0a07914

• Add support for OpBitwiseOr, OpBitwiseXor, and OpBitwiseAnd in SPIR-V dialect.

  Show description for 8cbe7bb
  tensorflower-gardener

  committed
  8cbe7bb

  Copy full SHA for 8cbe7bb

• Change ShapeN to support dynamic dimension.

  Show description for 94526a9
  

  yunxing

  authored and

  tensorflower-gardener

  committed
  94526a9

  Copy full SHA for 94526a9

• [Grappler] Extend ConstantPushDown optimization to also handle BiasAdd when it is safe to do so.

  Show description for 12a6de4
  tensorflower-gardener

  committed
  12a6de4

  Copy full SHA for 12a6de4

• Wiring tensorflow/core/platform:grpc_services into tensorflow/core/BUILD.

  Show description for d3f98ab
  

  bmzhao

  authored and

  tensorflower-gardener

  committed
  d3f98ab

  Copy full SHA for d3f98ab

• Merge pull request #33386 from feihugis:DS_Test_Format_4

  Show description for 417aa4a
  tensorflower-gardener

  committed
  417aa4a

  Copy full SHA for 417aa4a

• [tf.data] Fix an internal test.

  Show description for 733c42e
  

  jsimsa

  authored and

  tensorflower-gardener

  committed
  733c42e

  Copy full SHA for 733c42e

• Wiring the following build targets into tensorflow/core/BUILD

  Show description for c702052
  

  bmzhao

  authored and

  tensorflower-gardener

  committed
  c702052

  Copy full SHA for c702052

• Added multilabel handling to AUC keras metric. Computes separate AUCs for each label, and returns their average. Disabled by default unless `multi_label` is set to True upon initialization of the A…

  Show description for 1c8530a
  tensorflower-gardener

  committed
  1c8530a

  Copy full SHA for 1c8530a

• Do not capture constants from forward pass in cond_v2.

  Show description for 21f3a69
  

  saxenasaurabh

  authored and

  tensorflower-gardener

  committed
  21f3a69

  Copy full SHA for 21f3a69

• Add XLAControlFlowContext when experimental_compile=True.

  Show description for f2733d6
  

  zhangyujing

  authored and

  tensorflower-gardener

  committed
  f2733d6

  Copy full SHA for f2733d6

• Add fuzzer for DecodeCSV Op.

  Show description for 1745367
  

  mihaimaruseac

  authored and

  tensorflower-gardener

  committed
  1745367

  Copy full SHA for 1745367

• Making sure that when we're running MultiDeviceFunctions that span processes, we have the right rendezvous available at run time.

  Show description for 6141e99
  

  rohan100jain

  authored and

  tensorflower-gardener

  committed
  6141e99

  Copy full SHA for 6141e99

Commits on Oct 15, 2019

• Add a boolean flag `keep_original_batch_size` in saving_utils.model_input_signature function, to allow preserving the batch dim from the model's input signature.

  Show description for ccd5f2d
  

  haozha111

  authored and

  tensorflower-gardener

  committed
  ccd5f2d

  Copy full SHA for ccd5f2d

• [TF] Allow batch size 0 images (relax some checks)

  Show description for f5b3eed
  

  ebrevdo

  authored and

  tensorflower-gardener

  committed
  f5b3eed

  Copy full SHA for f5b3eed

• Inline ADDITIONAL_CORE_PROTO_SRCS into the one place it is used.

  Show description for 17fc974
  

  nairb774

  authored and

  tensorflower-gardener

  committed
  17fc974

  Copy full SHA for 17fc974

• Add quantization support for StridedSlice.

  Show description for d08bf91
  tensorflower-gardener

  committed
  d08bf91

  Copy full SHA for d08bf91

Pagination

Previous

Next