Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Aug 24, 2022

• Update Pytorch GPTQ docs

  

  eladco

  committed
  be2c8d0

  Copy full SHA for be2c8d0

Commits on Aug 23, 2022

• Merge pull request #352 from ofirgo/ofirgo/min-kpi-duplications

  Show description for 1091b48
  reuvenperetz

  authored
  1091b48

  Copy full SHA for 1091b48

• Support pytorch GPTQ: STE algorithm (#351)

  Show description for a0c8413
  eladc-git

  authored
  a0c8413

  Copy full SHA for a0c8413

• Skip computation of KPI when calculating KPI matrix for min configuration (for all KPI targets).

  Show description for a97303e
  

  Ofir Gordon

  authored and

  Ofir Gordon

  committed
  a97303e

  Copy full SHA for a97303e

• BOPs KPI integration into MP search (#335)

  Show description for 8505858
  

  ofirgo

  and

  Ofir Gordon

  authored
  8505858

  Copy full SHA for 8505858

Commits on Aug 22, 2022

• Fix optimizer import from Keras to support 2.9 (#348)

  Show description for 76cf896
  

  reuvenperetz

  and

  Reuven Peretz

  authored
  76cf896

  Copy full SHA for 76cf896

• Merge pull request #350 from reuvenperetz/remove-old-readme-img

  Show description for b61a477
  lior-dikstein

  authored
  b61a477

  Copy full SHA for b61a477

• Remove old block diagaram image

  

  Reuven Peretz

  committed
  57c2784

  Copy full SHA for 57c2784

• Fix fully quantized sanity test (#349)

  Show description for b384be1
  

  reuvenperetz

  and

  Reuven Peretz

  authored
  b384be1

  Copy full SHA for b384be1

Commits on Aug 21, 2022

• Change defulat GPTQ config values (#347)

  Show description for 0e54053
  

  reuvenperetz

  and

  Reuven Peretz

  authored
  0e54053

  Copy full SHA for 0e54053

• Merge pull request #346 from sony/fix-github-actions

  Show description for 8d0d762
  reuvenperetz

  authored
  8d0d762

  Copy full SHA for 8d0d762

• Remove old Keras models runner

  Show description for caecc0a
  

  Reuven Peretz

  committed
  caecc0a

  Copy full SHA for caecc0a

• Fix fully quantized weights quantization (#344)

  Show description for 5f0aa92
  

  reuvenperetz

  and

  Reuven Peretz

  authored
  5f0aa92

  Copy full SHA for 5f0aa92

• Add UniformQuantizer to QAT (#340)

  Show description for a05ea8e
  

  elad-c

  and

  elad cohen

  authored
  a05ea8e

  Copy full SHA for a05ea8e

Commits on Aug 19, 2022

• Fix gptq experimental facade (#345)

  Show description for 276f08b
  

  reuvenperetz

  and

  Reuven Peretz

  authored
  276f08b

  Copy full SHA for 276f08b

Commits on Aug 18, 2022

• Remove old TF github actions (#343)

  Show description for b6d84d2
  

  reuvenperetz

  and

  Reuven Peretz

  authored
  b6d84d2

  Copy full SHA for b6d84d2

• Fix README (#341)

  Show description for 852cc84
  

  reuvenperetz

  and

  Reuven Peretz

  authored
  852cc84

  Copy full SHA for 852cc84

• Fix import in SNC to support TF2.6 (#342)

  Show description for 976df39
  

  reuvenperetz

  and

  Reuven Peretz

  authored
  976df39

  Copy full SHA for 976df39

Commits on Aug 17, 2022

• Update README and workflows to run tests by framework version (#330)

  Show description for 54559d4
  

  reuvenperetz

  and

  Reuven Peretz

  authored
  54559d4

  Copy full SHA for 54559d4

• Move some GPTQ functions to common (#339)

  eladc-git

  authored
  452e362

  Copy full SHA for 452e362

• Pytorch buffers (#337)

  Show description for 9d97afb
  lapid92

  authored
  9d97afb

  Copy full SHA for 9d97afb

Commits on Aug 16, 2022

• Quantize keras weights fully for pot quantizers (#327)

  Show description for 7b50825
  

  reuvenperetz

  and

  Reuven Peretz

  authored
  7b50825

  Copy full SHA for 7b50825

• BOPs KPI computation (#334)

  Show description for 113f493
  

  ofirgo

  and

  Ofir Gordon

  authored
  113f493

  Copy full SHA for 113f493

Commits on Aug 14, 2022

• Graph substitutions for weight and activation composition (BOPs KPI preparations) (#333)

  Show description for 80bc647
  

  ofirgo

  and

  Ofir Gordon

  authored
  80bc647

  Copy full SHA for 80bc647

• Increase number of filters in Conv layer in gptq test to increase the chance of gptq effect on the weights (#336)

  Show description for 58dbd02
  

  ofirgo

  and

  Ofir Gordon

  authored
  58dbd02

  Copy full SHA for 58dbd02

Commits on Aug 10, 2022

• Add keras QAT tutorial (#320)

  Show description for 4d59769
  

  elad-c

  and

  elad cohen

  authored
  4d59769

  Copy full SHA for 4d59769

• fix license (#332)

  Show description for b93da02
  

  elad-c

  and

  elad cohen

  authored
  b93da02

  Copy full SHA for b93da02

Commits on Aug 9, 2022

• Qat layers test (#331)

  Show description for 814d070
  

  elad-c

  and

  elad cohen

  authored
  814d070

  Copy full SHA for 814d070

• Add layer replacement to network editor (#323)

  Show description for 036d207
  

  IdanBenAmi

  and

  Idan-BenAmi

  authored
  036d207

  Copy full SHA for 036d207

• Minor fix to Keras model gradients call with norm_weights argument (#329)

  Show description for e4ec557
  

  ofirgo

  and

  Ofir Gordon

  authored
  e4ec557

  Copy full SHA for e4ec557

• Replace TensorFlowOpLayer with TFOpLambda in SNC (#328)

  Show description for 3a7cda7
  

  reuvenperetz

  and

  Reuven Peretz

  authored
  3a7cda7

  Copy full SHA for 3a7cda7

Commits on Aug 7, 2022

• Bug fix in GPTQ config test and symmetric qunatizer (#325)

  Show description for 5ba2040
  haihabi

  authored
  5ba2040

  Copy full SHA for 5ba2040

• Support DW-Conv2D in QAT when quantizing per weights per channel (#319)

  Show description for f6ca728
  

  elad-c

  and

  elad cohen

  authored
  f6ca728

  Copy full SHA for f6ca728

• fix weight quantization (#322)

  Show description for c5c8345
  

  elad-c

  and

  elad cohen

  authored
  c5c8345

  Copy full SHA for c5c8345

• Gumbel Rounding API and Testing Update (#321)

  Show description for 7f7403f
  haihabi

  authored
  7f7403f

  Copy full SHA for 7f7403f

Pagination

Previous

Next