Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Aug 27, 2024

• Support Smooth Softmax in GroupQueryAttention (#21867)

  Show description for 6e57576
  tianleiwu

  authored
  6e57576

  Copy full SHA for 6e57576

• [js] add big data file to formatter ignore list (#21767)

  Show description for 99bc45d
  fs-eire

  authored
  99bc45d

  Copy full SHA for 99bc45d

• [VitisAI] add OpSchema, VitisAI use IKernelLookup to check supported ops, VitisAI def_builder adds TypeConstraint related processing (#21688)

  Show description for 422e6e6
  

  zz002

  and

  Zhenze Wang

  authored
  422e6e6

  Copy full SHA for 422e6e6

Commits on Aug 26, 2024

• [JS/WebGPU] Add GatherBlockQuantized op support (#21734)

  Show description for af18824
  satyajandhyala

  authored
  af18824

  Copy full SHA for af18824

• [CUDA] enable causal in MultiHeadAttention (#21852)

  Show description for ad38212
  tianleiwu

  authored
  ad38212

  Copy full SHA for ad38212

• [js/webgpu] Enable pad f16 uniform (#21691)

  Show description for d9c57ac
  

  axinging

  and

  fs-eire

  authored
  d9c57ac

  Copy full SHA for d9c57ac

• sign native dll with correct cert (#21854)

  Show description for 2877de7
  mszhanyi

  authored
  2877de7

  Copy full SHA for 2877de7

• Fix typo for react native pipeline (#21845)

  Show description for 983c4d5
  carzh

  authored
  983c4d5

  Copy full SHA for 983c4d5

Commits on Aug 25, 2024

• [MIGraphX EP Support]Remove default noopt for Migraphx EP in Benchmark.py (#21843)

  Show description for 9a70475
  

  TedThemistokleous

  and

  Ted Themistokleous

  authored
  9a70475

  Copy full SHA for 9a70475

Commits on Aug 23, 2024

• [js/webgpu] optimize MatmulNBits (#21747)

  Show description for 87165b9
  qjia7

  authored
  87165b9

  Copy full SHA for 87165b9

• Refine `op_types_to_quantize` argument handling in matmul_4bits_quantizer.py (#21815)

  Show description for 4af6291
  duanshengliu

  authored
  4af6291

  Copy full SHA for 4af6291

• Replace "DML CPU" Allocator with onnxruntime::CpuAllocator (#21818)

  Show description for 44dcc3a
  

  smk2007

  and

  Sheil Kumar

  authored
  44dcc3a

  Copy full SHA for 44dcc3a

• [CoreML EP] Fix ArgMaxOpBuilder::AddToModelBuilderImpl() nullptr Node access. (#21797)

  edgchen1

  authored
  5726318

  Copy full SHA for 5726318

• enable large initializer offset align for save external data in ORT (#21604)

  Show description for 4c4ae1e
  frank-dong-ms

  authored
  4c4ae1e

  Copy full SHA for 4c4ae1e

• [js/webgpu] Optimize conv1d by conv2d (#19388)

  Show description for 27a6890
  

  qjia7

  and

  fs-eire

  authored
  27a6890

  Copy full SHA for 27a6890

• Ovep 1.19 bug fix 2 (#21829)

  Show description for 0368dd4
  

  preetha-intel

  and

  jatinwadhwa921

  authored
  0368dd4

  Copy full SHA for 0368dd4

Commits on Aug 22, 2024

• [VitisAI] optimize model clone (#21706)

  Show description for 37a7dd7
  

  3 peopleauthored
  37a7dd7

  Copy full SHA for 37a7dd7

• Revert "Upgrade emsdk from 3.1.59 to 3.1.62" (#21817)

  Show description for ba7baae
  guschmue

  authored
  ba7baae

  Copy full SHA for ba7baae

• [QNN EP] Apply workaround for Conv validation bug when bias input is implicit (#21764)

  Show description for 514b469
  adrianlizarraga

  authored
  514b469

  Copy full SHA for 514b469

• Do not allow clearing Android logs if the emulator is not running (#21578)

  Show description for 6c1a3f8
  jchen351

  authored
  6c1a3f8

  Copy full SHA for 6c1a3f8

• [VSINPU]Update vsinpu patches (#21402)

  Show description for ff3e8b0
  chenfeiyue-cfy

  authored
  ff3e8b0

  Copy full SHA for ff3e8b0

• [VitisAI] remove wrong error msg, required by Microsoft (#21715)

  Show description for 3ff8ca2
  

  BoarQing

  and

  Yueqing Zhang

  authored
  3ff8ca2

  Copy full SHA for 3ff8ca2

• [CUDA] Update benchmark_mha.py to capture debug info to identify sdpa kernel (#21804)

  Show description for 25d7a4f
  tianleiwu

  authored
  25d7a4f

  Copy full SHA for 25d7a4f

• run sparse attention test sequentially (#21808)

  Show description for 44a3923
  tianleiwu

  authored
  44a3923

  Copy full SHA for 44a3923

Commits on Aug 21, 2024

• Fix warnings (#21809)

  Show description for c0b68e7
  Jamather

  authored
  c0b68e7

  Copy full SHA for c0b68e7

• Add K=0 check to MatMul<float>::Compute() specialization. (#21803)

  Show description for fb9ce18
  edgchen1

  authored
  fb9ce18

  Copy full SHA for fb9ce18

• [MIGraphX EP] Add support for MIGraphX Exhaustive tune flag (#46) (#21599)

  Show description for 0e827c2
  

  3 peopleauthored
  0e827c2

  Copy full SHA for 0e827c2

• [MIGraphX EP Support] Update migx scripts (#21806)

  Show description for 26a4993
  TedThemistokleous

  authored
  26a4993

  Copy full SHA for 26a4993

• [MIGraphX EP] Ensure we support all inputs for MatMulInteger and ConvInteger. (#21680)

  Show description for ed155ad
  

  TedThemistokleous

  and

  Ted Themistokleous

  authored
  ed155ad

  Copy full SHA for ed155ad

• Fix Orttraining Linux Lazy Tensor CI Pipeline (#21652)

  Show description for 009209e
  mindest

  authored
  009209e

  Copy full SHA for 009209e

• [DML] Add int4 QDQ (#21592)

  PatriceVignola

  authored
  de6ebcb

  Copy full SHA for de6ebcb

• update size limit check of training GPU wheel (#21762)

  Show description for 12f426c
  mszhanyi

  authored
  12f426c

  Copy full SHA for 12f426c

Commits on Aug 20, 2024

• [TransposeOptimizer] Fix axis for QuantizeLinear inserted after DQ (per-channel) -> Unsqueeze (#21793)

  Show description for 6fbb0ae
  adrianlizarraga

  authored
  6fbb0ae

  Copy full SHA for 6fbb0ae

• [QNN EP] Fix compile error for QNN EP on Windows x64 due to missing /bigobj flag (#21795)

  Show description for 28c252c
  adrianlizarraga

  authored
  28c252c

  Copy full SHA for 28c252c

• [CUDA] cuDNN Flash Attention (#21629)

  Show description for fbc3927
  tianleiwu

  authored
  fbc3927

  Copy full SHA for fbc3927

Pagination

Previous

Next