Commits

Branch selector

 6f4d653

User selector

All users

Datepicker

All time

Commit History

Commits on Aug 25, 2023

• reland remove conv weight scale reciprocal (#226)

  leslie-fang-intel

  authored
  6f4d653

  Copy full SHA for 6f4d653

Commits on Aug 24, 2023

• fix oneDNN 3.1 issue (#225)

  leslie-fang-intel

  authored
  fccbd67

  Copy full SHA for fccbd67

• Revert remove conv weight scale reciprocal

  yanbing-j

  committed
  a09457d

  Copy full SHA for a09457d

Commits on Aug 23, 2023

• Fixes oneDNN 3.1.1 build on AArch64 (#223)

  Show description for 7dfb139
  

  nSircombe

  and

  milpuz01

  authored
  7dfb139

  Copy full SHA for 7dfb139

Commits on Aug 19, 2023

• remove conv weight scale reciprocal (#222)

  leslie-fang-intel

  authored
  c90e52c

  Copy full SHA for c90e52c

Commits on Jun 21, 2023

• fix the issue of qsum calculation of qconv2d sum (#211)

  leslie-fang-intel

  authored
  3c90e36

  Copy full SHA for 3c90e36

Commits on Jun 9, 2023

• Update oneDNN to v3.1.1 (aligned with ideep_dev) (#209)

  Xia-Weiwen

  authored
  3801422

  Copy full SHA for 3801422

Commits on Mar 22, 2023

• make primitive creation aware of reorder_src/reorder_weight setting (#177)

  XiaobingSuper

  authored
  fe83782

  Copy full SHA for fe83782

• add is_channels_last for deconv to align channels last format with pytorch (#165)

  CaoE

  authored
  899f80e

  Copy full SHA for 899f80e

Commits on Feb 13, 2023

• PyTorch Primitive caching (#162)

  Show description for 7bc3e12
  snadampal

  authored
  7bc3e12

  Copy full SHA for 7bc3e12

Commits on Jan 16, 2023

• Switch mkl-dnn branch to dev-graph-beta-3-no-submod and Upgrade oneDNN to v2.7.3

  yanbing-j

  committed
  e7925bc

  Copy full SHA for e7925bc

Commits on Jan 4, 2023

• add zp into pd cache key (#152)

  leslie-fang-intel

  authored
  7201315

  Copy full SHA for 7201315

Commits on Dec 15, 2022

• Enable set post op sum zero point of conv_add and conv_add_relu (#148)

  Show description for ff2e09f
  leslie-fang-intel

  authored
  ff2e09f

  Copy full SHA for ff2e09f

Commits on Dec 2, 2022

• switch mkl-dnn to public graph-v0.7.2

  yanbing-j

  committed
  e533c77

  Copy full SHA for e533c77

Commits on Nov 16, 2022

• Add a parameter to control expected deconv weight format (#146)

  Xia-Weiwen

  authored
  5ddc65e

  Copy full SHA for 5ddc65e

Commits on Oct 28, 2022

• switch mkl-dnn to public graph-v0.7

  yanbing-j

  committed
  d275877

  Copy full SHA for d275877

• Add a validator of guesses of conv input when prepacking weight (#139)

  

  Xia-Weiwen

  authored and

  yanbing-j

  committed
  e88877a

  Copy full SHA for e88877a

• Fix a typo in conv.hpp (#138)

  Xia-Weiwen

  authored
  6947606

  Copy full SHA for 6947606

• add fuse_swish_sum (#140) (#141)

  jiayisunx

  authored
  cf07725

  Copy full SHA for cf07725

Commits on Sep 22, 2022

• Update LLGA commit to alpha patch release (0.5.2) (#125)

  sanchitintel

  authored
  e4383b8

  Copy full SHA for e4383b8

• Add deprecated api back for compatibility (#122)

  Xia-Weiwen

  authored
  177b04f

  Copy full SHA for 177b04f

Commits on Sep 16, 2022

• Bug fix found by Caffe2 UT (#119)

  Show description for 8dfb364
  

  Xia-Weiwen

  authored and

  XiaobingSuper

  committed
  8dfb364

  Copy full SHA for 8dfb364

Commits on Sep 15, 2022

• Fix typo in comments

  

  Xia-Weiwen

  authored and

  yanbing-j

  committed
  3e7b510

  Copy full SHA for 3e7b510

• Fix bug in conv when dst tensor does not have the expected shape; use int data type to represent scales vector size to avoid narrowing conversion

  

  Xia-Weiwen

  authored and

  yanbing-j

  committed
  0eebb69

  Copy full SHA for 0eebb69

Commits on Sep 9, 2022

• downgrade llga and related fusion for latest stock pytorch

  Show description for 0646992
  

  Guobing-Chen

  authored and

  XiaobingSuper

  committed
  0646992

  Copy full SHA for 0646992

• Fix convolution perf regression and IPEX UT failure (#12)

  Show description for bc6a300
  

  Xia-Weiwen

  authored and

  XiaobingSuper

  committed
  bc6a300

  Copy full SHA for bc6a300

• Bug fix: bias tensor is out of its scope (#11)

  

  Xia-Weiwen

  authored and

  XiaobingSuper

  committed
  b1f96ad

  Copy full SHA for b1f96ad

• Weiwen/sync xiaobings commits (#10)

  Show description for 944a47f
  

  Xia-Weiwen

  authored and

  XiaobingSuper

  committed
  944a47f

  Copy full SHA for 944a47f

• update llga to bf7b1ba, oneDNN to 9b2cefc

  

  Guobing-Chen

  authored and

  XiaobingSuper

  committed
  fb323b3

  Copy full SHA for fb323b3

• Sync with recent changes in IPEX's ideep (#9)

  

  Xia-Weiwen

  authored and

  XiaobingSuper

  committed
  e4b3987

  Copy full SHA for e4b3987

• Add hardsigmoid fusion together with oneDNN update (#8)

  

  Guobing-Chen

  authored and

  XiaobingSuper

  committed
  73936bb

  Copy full SHA for 73936bb

• Bug fixes about conv and matmul (#7)

  

  Xia-Weiwen

  authored and

  XiaobingSuper

  committed
  438c02a

  Copy full SHA for 438c02a

• Ideep converge (#6)

  Show description for 76eee04
  

  Guobing-Chen

  authored and

  XiaobingSuper

  committed
  76eee04

  Copy full SHA for 76eee04

• Merge conv/deconv/inner_product/matmul (#5)

  Show description for 226df3a
  

  Xia-Weiwen

  authored and

  XiaobingSuper

  committed
  226df3a

  Copy full SHA for 226df3a

• Refactor code format

  Show description for afb360b
  

  Guobing-Chen

  authored and

  XiaobingSuper

  committed
  afb360b

  Copy full SHA for afb360b

Pagination

Previous

Next