Commits

Branch selector

 master

User selector

liufengdb

Datepicker

All time

Commit History

Commits on Dec 24, 2020

• Fix the op quant trait of tfl.rsqrt and the nudging for the calibration data

  Show description for 05303d0
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  05303d0

  Copy full SHA for 05303d0

• Fix the op quant trait of tfl.prelu

  Show description for f02d09c
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  f02d09c

  Copy full SHA for f02d09c

Commits on Dec 23, 2020

• Add a pattern to convert the shape operand of the tf.reshape op to be a vector

  Show description for 6b119d5
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  6b119d5

  Copy full SHA for 6b119d5

Commits on Dec 22, 2020

• Skip the dequantize op check when the tensor type is not float

  Show description for 3376c11
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  3376c11

  Copy full SHA for 3376c11

Commits on Dec 21, 2020

• Fix the same element type trait of tfl.greater_equal op

  Show description for f697293
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  f697293

  Copy full SHA for f697293

Commits on Dec 17, 2020

• Rerun the fully_quantize zip tests with the new quantizer

  Show description for 483d6fe
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  483d6fe

  Copy full SHA for 483d6fe

• Support dtypes.DType in the composition

  Show description for 84d2f52
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  84d2f52

  Copy full SHA for 84d2f52

Commits on Dec 16, 2020

• Support high dimension list in the tfr dialect

  Show description for f912703
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  f912703

  Copy full SHA for f912703

• disable the new quantizer for the 16*8 mode for now

  Show description for 062d6bd
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  062d6bd

  Copy full SHA for 062d6bd

Commits on Dec 15, 2020

• Disable tf_stack_test on windows

  Show description for 73f94a9
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  73f94a9

  Copy full SHA for 73f94a9

Commits on Dec 11, 2020

• disable //tensorflow/python:tf_stack_test on windows

  Show description for 75643bd
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  75643bd

  Copy full SHA for 75643bd

Commits on Dec 10, 2020

• disable all tap tests for gradient_checker_test

  Show description for 6bdd30e
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  6bdd30e

  Copy full SHA for 6bdd30e

Commits on Dec 7, 2020

• Skip the input tensor order check if the tensor list size can be determined by the attributes

  Show description for eaa8513
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  eaa8513

  Copy full SHA for eaa8513

• Add the support of getting the length of tensor list

  Show description for e5ea20d
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  e5ea20d

  Copy full SHA for e5ea20d

• Propagate the allowed unregistered attributes to the expanded operations

  Show description for 1c552a6
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  1c552a6

  Copy full SHA for 1c552a6

Commits on Dec 1, 2020

• Enable nested composition functions

  Show description for f432994
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  f432994

  Copy full SHA for f432994

• Visit and generate functions in the order of function definition

  Show description for 6b9e35f
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  6b9e35f

  Copy full SHA for 6b9e35f

Commits on Nov 25, 2020

• Update the documentation of the composite tensorflow project

  Show description for b3e7785
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  b3e7785

  Copy full SHA for b3e7785

Commits on Nov 24, 2020

• Emit a warning if the tensor range is too wide

  Show description for c6aaecf
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  c6aaecf

  Copy full SHA for c6aaecf

Commits on Nov 23, 2020

• Infer quantization parameters from the constant and activation ops at the same

  Show description for 47469e1
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  47469e1

  Copy full SHA for 47469e1

Commits on Nov 20, 2020

• Fixed the constant duplication bug in the quantization pass

  Show description for e8ca927
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  e8ca927

  Copy full SHA for e8ca927

Commits on Nov 14, 2020

• Disable the xla compiling in the mnist example

  Show description for 988c4be
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  988c4be

  Copy full SHA for 988c4be

Commits on Nov 10, 2020

• Added a flag to the quantization config to disable enforced fixed output range

  Show description for b2ff807
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  b2ff807

  Copy full SHA for b2ff807

Commits on Nov 5, 2020

• Fix the mnist_train example

  Show description for 5a9b3ae
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  5a9b3ae

  Copy full SHA for 5a9b3ae

Commits on Oct 29, 2020

• Add a README file for the composable TensorFlow project

  Show description for ede9f23
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  ede9f23

  Copy full SHA for ede9f23

Commits on Oct 20, 2020

• Add metrics for the usage of the op composition feature

  Show description for 2406f50
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  2406f50

  Copy full SHA for 2406f50

Commits on Oct 19, 2020

• Do not run op_reg_gen_test with pip

  Show description for 2d6bdab
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  2d6bdab

  Copy full SHA for 2d6bdab

Commits on Oct 18, 2020

• Use the public tensorflow dependence to fix the nightly build

  Show description for a038cf2
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  a038cf2

  Copy full SHA for a038cf2

Commits on Oct 16, 2020

• Add end2end examples for defining and running composite ops

  Show description for cae7bb0
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  cae7bb0

  Copy full SHA for cae7bb0

• Fixed the nightly cpu build

  Show description for 03bf890
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  03bf890

  Copy full SHA for 03bf890

Commits on Oct 15, 2020

• Clean up the TFR Decompose context implementation

  Show description for 5952da0
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  5952da0

  Copy full SHA for 5952da0

Commits on Oct 13, 2020

• Disable the node_expansion_test and graph_decompose_test on mac and windows

  Show description for ed26ad2
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  ed26ad2

  Copy full SHA for ed26ad2

• Add graph decompose pass to expand composite op if it was illegal in the tf

  Show description for 716e817
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  716e817

  Copy full SHA for 716e817

• Add node expansion pass to expand composite op if its kernel was not found

  Show description for 5c0ffc8
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  5c0ffc8

  Copy full SHA for 5c0ffc8

• Add the op expansion graph optimization pass to tensorflow

  Show description for a05f8bd
  

  liufengdb

  authored and

  tensorflower-gardener

  committed
  a05f8bd

  Copy full SHA for a05f8bd

Pagination

Previous

Next