Commits

Breadcrumbs

History for

oncdb9120

User selector

All users

Datepicker

All time

Commit History

Commits on Dec 3, 2020

• longformer onnx conversion and benchmark tools (microsoft#6007)

  Show description for cdb9120
  tianleiwu

  authored
  cdb9120

  Copy full SHA for cdb9120

• Support Fusion for 1 and 2 Inputs Bert Models Converted From tf (microsoft#5993)

  Show description for 3b198c9
  liuziyue

  authored
  3b198c9

  Copy full SHA for 3b198c9

• Fix build.py bug which prevents running some unit tests (microsoft#5990)

  Show description for 2b35f7d
  baijumeswani

  authored
  2b35f7d

  Copy full SHA for 2b35f7d

• Make operator TreeEnsemble 5x faster for batches of size 100.000 (microsoft#5965)

  Show description for 0acc383
  xadupre

  authored
  0acc383

  Copy full SHA for 0acc383

• Initialize a structure in operator ReduceSum (microsoft#6005)

  Show description for 524b9fa
  xadupre

  authored
  524b9fa

  Copy full SHA for 524b9fa

• Fix bugs for 1: Calibrator should check model inputs; 2: (microsoft#6017)

  Show description for 648c9c7
  zhanghuanrong

  authored
  648c9c7

  Copy full SHA for 648c9c7

Commits on Dec 2, 2020

• Fix PR microsoft#5550 reverted in microsoft#5911 (performance improvment for operator Transpose) (microsoft#5916)

  Show description for bdd06f6
  xadupre

  authored
  bdd06f6

  Copy full SHA for bdd06f6

• Fuse MatMulIntegerToFloat only when scales are scalar (microsoft#6008)

  Show description for f2dcba7
  yufenglee

  authored
  f2dcba7

  Copy full SHA for f2dcba7

• Add int32_t support for DeQuantizeLinear (microsoft#5994)

  Show description for 4fdfbfd
  yufenglee

  authored
  4fdfbfd

  Copy full SHA for 4fdfbfd

• include gemm_helper.h (microsoft#5988)

  oliviajain

  authored
  c727a28

  Copy full SHA for c727a28

• [NNAPI] Support non-1d tensor for C of Gemm op (microsoft#5982)

  Show description for cdacee6
  guoyu-wang

  authored
  cdacee6

  Copy full SHA for cdacee6

Commits on Dec 1, 2020

• optimize a bert model converted using tf2onnx (microsoft#5492)

  Show description for 5f51689
  wangyems

  authored
  5f51689

  Copy full SHA for 5f51689

Commits on Nov 30, 2020

• Add python 3.9 support (microsoft#5874)

  Show description for 2d9dcc4
  snnn

  authored
  2d9dcc4

  Copy full SHA for 2d9dcc4

• Use CUDA's IsAllFinite kernel for ROCm

  jessebenson

  committed
  bd96f60

  Copy full SHA for bd96f60

Commits on Nov 27, 2020

• [OpenVINO]Fix memory leak upon exception throwing (microsoft#5954)

  Show description for 5f5d4a1
  mosdav

  authored
  5f5d4a1

  Copy full SHA for 5f5d4a1

Commits on Nov 26, 2020

• Add support for Python 3.8+ on Windows when CUDA is enabled (microsoft#5956)

  ivanst0

  authored
  015fbb3

  Copy full SHA for 015fbb3

• [OpenVINO]Fix memory leak in IsDebugEnabled() under Windows (microsoft#5948)

  Show description for e207589
  

  mosdav

  and

  modav

  authored
  e207589

  Copy full SHA for e207589

• [NNAPI EP] Update squeeze ops (microsoft#5946)

  Show description for 4afdced
  guoyu-wang

  authored
  4afdced

  Copy full SHA for 4afdced

• Enable scalar input/output for NNAPI EP (microsoft#5922)

  Show description for d52b9ac
  guoyu-wang

  authored
  d52b9ac

  Copy full SHA for d52b9ac

Commits on Nov 25, 2020

• Add Longformer Attention Cuda Op(microsoft#5932)

  Show description for 31a6be3
  tianleiwu

  authored
  31a6be3

  Copy full SHA for 31a6be3

• Bug fix for MaskRCNN and FasterRCNN (microsoft#5935)

  Show description for e39e82b
  MaajidKhan

  authored
  e39e82b

  Copy full SHA for e39e82b

• Make NNAPI EP reject nodes with no-shape inputs (microsoft#5927)

  guoyu-wang

  authored
  8736865

  Copy full SHA for 8736865

Commits on Nov 24, 2020

• [Java] Initial Apple Silicon support (microsoft#5891)

  Show description for 8b83c51
  Craigacp

  authored
  8b83c51

  Copy full SHA for 8b83c51

• Symbolic shape inference: fix rank for ConstantOfShape (microsoft#5912)

  

  KeDengMS

  authored
  ee908eb

  Copy full SHA for ee908eb

• Reenable skip flatten/reshape if it's Gemm's input (microsoft#5904)

  guoyu-wang

  authored
  c49d5f1

  Copy full SHA for c49d5f1

• Revert "Improves performance of operator Transpose (microsoft#5550)" (microsoft#5911)

  Show description for 7823033
  snnn

  authored
  7823033

  Copy full SHA for 7823033

• Implement QLinear GlobalAveragePool with sse2/neon. (microsoft#5838)

  Show description for 9992f0f
  

  zhanghuanrong

  and

  tracysh

  authored
  9992f0f

  Copy full SHA for 9992f0f

• Add info level logging for NNAPI EP partitions and number of supported nodes (microsoft#5903)

  guoyu-wang

  authored
  d0a007e

  Copy full SHA for d0a007e

Commits on Nov 23, 2020

• Fix for hetero multi python binding with new shared library (microsoft#5895)

  Show description for 9164101
  

  sfatimar

  and

  sfatimar

  authored
  9164101

  Copy full SHA for 9164101

• remove use_cdn when loading pretrained model (microsoft#5900)

  wangyems

  authored
  3d5b48a

  Copy full SHA for 3d5b48a

• Fuzz testing misc (microsoft#5862)

  Show description for 353e071
  satyajandhyala

  authored
  353e071

  Copy full SHA for 353e071

• Add QLinearMatMul(u8s8) (microsoft#5899)

  tracysh

  authored
  f473dd2

  Copy full SHA for f473dd2

• Implement missing pieces for ARM QLinearConv support (microsoft#5894)

  tracysh

  authored
  57c9206

  Copy full SHA for 57c9206

Commits on Nov 21, 2020

• Expose knobs to create and share (CPU) allocators across sessions in C# and Python (microsoft#5634)

  hariharans29

  authored
  d46dbea

  Copy full SHA for d46dbea

• Temporary fix for Android CI failure (microsoft#5889)

  Show description for 26e6ced
  guoyu-wang

  authored
  26e6ced

  Copy full SHA for 26e6ced

Pagination

Previous

Next