Commits

Branch selector

 main

User selector

kimishpatel

Datepicker

All time

Commit History

Commits on Nov 1, 2023

• [Quantization] Add "quantization_tag" as metadata to fx proxy (pytorch#108764)

  Show description for 9e2af97
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  9e2af97

  Copy full SHA for 9e2af97

Commits on Oct 24, 2023

• [Quantization] Add a test for QAT + PTQ selective quantization in (pytorch#111689)

  Show description for a8760f1
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  a8760f1

  Copy full SHA for a8760f1

Commits on Sep 18, 2023

• [Pytorch][quant] Move xnnpack quantizer to use aten.linear (pytorch#109254)

  Show description for 73ac814
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  73ac814

  Copy full SHA for 73ac814

Commits on Sep 7, 2023

• [Quant] Move to BFS instead of DFS to check for connectedness (pytorch#108572)

  Show description for c1877e9
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  c1877e9

  Copy full SHA for c1877e9

Commits on Sep 6, 2023

• [sdpa decomp] change sdpa decomp to be consistent with flash attention (pytorch#108608)

  Show description for ebed490
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  ebed490

  Copy full SHA for ebed490

Commits on Sep 3, 2023

• [aten decomp] Update sdpa decom (pytorch#108371)

  Show description for cc50e65
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  cc50e65

  Copy full SHA for cc50e65

Commits on Sep 2, 2023

• [Quantization] Add metadata porting for nodes added by quantization (pytorch#107107)

  Show description for ffc0c46
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  ffc0c46

  Copy full SHA for ffc0c46

• [Quant] Add DQ duplication pass (pytorch#107900)

  Show description for eb67c45
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  eb67c45

  Copy full SHA for eb67c45

• [Quant] Bug fix (pytorch#107899)

  Show description for f8d1ca9
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  f8d1ca9

  Copy full SHA for f8d1ca9

• [Quantization] Make annotation util functions return annotated nodes (pytorch#107106)

  Show description for 37b0d76
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  37b0d76

  Copy full SHA for 37b0d76

• [Quant] Use input_qspec_map for weight quantization of linear (pytorch#107105)

  Show description for 99168c1
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  99168c1

  Copy full SHA for 99168c1

Commits on Aug 24, 2023

• [pytorch][Quant] Fix bias quant bug (pytorch#107810)

  Show description for 2fbe6ef
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  2fbe6ef

  Copy full SHA for 2fbe6ef

Commits on Jul 1, 2023

• [PT2][Quant] Enable symbolic shape in linear quantization (pytorch#104473)

  Show description for bd0f0f4
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  bd0f0f4

  Copy full SHA for bd0f0f4

Commits on Jun 15, 2023

• [PT2][Quant] Update op names for decomposed quantized lib (pytorch#103251)

  Show description for 90ee6a7
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  90ee6a7

  Copy full SHA for 90ee6a7

Commits on Jun 14, 2023

• [PT2][Quant] Change quat conv bn fusion code (pytorch#103556)

  Show description for 49dcf48
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  49dcf48

  Copy full SHA for 49dcf48

Commits on Jun 8, 2023

• [PT2][Quant] In linear partition include functional.linear (pytorch#103186)

  Show description for a49aefd
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  a49aefd

  Copy full SHA for a49aefd

Commits on Jun 7, 2023

• [PT2][Quant] Use composble quantizer for embedding + static conv + dynamic (pytorch#103116)

  Show description for 471407c
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  471407c

  Copy full SHA for 471407c

Commits on Jun 6, 2023

• [PT2][Quant] Move embedding quantization to osss (pytorch#103088)

  Show description for 8e0837c
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  8e0837c

  Copy full SHA for 8e0837c

• [PT2][Quant] Introduce composable quantizer (pytorch#102846)

  Show description for 8824101
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  8824101

  Copy full SHA for 8824101

Commits on Jun 2, 2023

• [PT2][Quant][BE] Test refactor to be organize them better (pytorch#102704)

  Show description for 2296ee0
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  2296ee0

  Copy full SHA for 2296ee0

• [PT2][Quant] Enable dynamic quantization (pytorch#102703)

  Show description for a53acaf
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  a53acaf

  Copy full SHA for a53acaf

• [PT2][Quant] Update quconfig to contain input/qoutput activation qspec (pytorch#102702)

  Show description for 2301b62
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  2301b62

  Copy full SHA for 2301b62

• [PT2][Quant][BE] Refactor qnnpack_quantizer.py (pytorch#102701)

  Show description for 6492b7d
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  6492b7d

  Copy full SHA for 6492b7d

Commits on May 30, 2023

• [PT2][Quant][BE] refactor tets cose to reduce duplication and standardize (pytorch#102497)

  Show description for 4f46864
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  4f46864

  Copy full SHA for 4f46864

Commits on May 28, 2023

• [PT2][Quant] Enable test_qnnpack_quantizer_conv_linear test (pytorch#102399)

  Show description for af70fe9
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  af70fe9

  Copy full SHA for af70fe9

• [PT2][Quant] Move observer sharing ops to use module partitions (pytorch#102398)

  Show description for 0d876f7
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  0d876f7

  Copy full SHA for 0d876f7

• [PT2][Quant] Move add/add relu pattern via module partitioner (pytorch#102397)

  Show description for 9fac5af
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  9fac5af

  Copy full SHA for 9fac5af

• [PT2][Quant] Move maxpool_2d quant to use module partitioners (pytorch#102396)

  Show description for 3d8f405
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  3d8f405

  Copy full SHA for 3d8f405

• [PT2][Quant] Use module partitions for conv2d and conv2d + relu (pytorch#102395)

  Show description for d997e3a
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  d997e3a

  Copy full SHA for d997e3a

• [PT2][Quant] Use module partition for fused patterns (pytorch#102394)

  Show description for 4cb6add
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  4cb6add

  Copy full SHA for 4cb6add

Commits on May 26, 2023

• [PT2][Quant][BE] Apply formatting to test_quantize_pt2e (pytorch#102275)

  Show description for 9b5e4c3
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  9b5e4c3

  Copy full SHA for 9b5e4c3

Commits on May 17, 2023

• [PT2][Quant] Move to module partitioner for linear pattern quantization (pytorch#101122)

  Show description for 07e759e
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  07e759e

  Copy full SHA for 07e759e

• [PT] Update module partitioner to return parameter node (pytorch#101121)

  Show description for bec655f
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  bec655f

  Copy full SHA for bec655f

• [PT2][Quant] Remove None annotations (pytorch#101120)

  Show description for 2c807a4
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  2c807a4

  Copy full SHA for 2c807a4

Commits on May 4, 2023

• [PT2E][Quant] Use subgraph matcher annotate linear pattern (pytorch#100566)

  Show description for 24e9b8f
  

  kimishpatel

  authored and

  pytorchmergebot

  committed
  24e9b8f

  Copy full SHA for 24e9b8f

Pagination

Previous

Next