Commits

Breadcrumbs

History for

onmaster

User selector

All users

Datepicker

All time

Commit History

Commits on Sep 5, 2024

• Op_builder->is_compatible quite warning (microsoft#6093)

  Show description for 4f80385
  

  terry-for-github

  and

  loadams

  authored
  4f80385

Commits on Sep 4, 2024

• Handle an edge case where CUDA_HOME is not defined on ROCm systems (microsoft#6488)

  Show description for 10ba3dd
  amorehead

  authored
  10ba3dd

• Safe usage of popen (microsoft#6490)

  Show description for 662a421
  tjruwase

  authored
  662a421

Commits on Aug 29, 2024

• Avoid gds build errors on ROCm (microsoft#6456)

  Show description for a7ffe54
  

  rraminen

  and

  loadams

  authored
  a7ffe54

Commits on Aug 28, 2024

• Allow triton==3.0.x for fp_quantizer (microsoft#6447)

  Show description for 4864991
  siddartha-RE

  authored
  4864991

Commits on Aug 27, 2024

• Revert "Revert "Fix torch check (microsoft#6402)""

  Show description for eb37cac
  loadams

  committed
  eb37cac

• Revert "Fix torch check (microsoft#6402)"

  Show description for 77f61e6
  loadams

  committed
  77f61e6

Commits on Aug 22, 2024

• Fix torch check (microsoft#6402)

  loadams

  authored
  55b4cae

• [XPU] API align with new intel pytorch extension release (microsoft#6395)

  Show description for b81b197
  

  YizhouZ

  and

  loadams

  authored
  b81b197

• add pip install cutlass version check (microsoft#6393)

  Show description for 51da191
  GuanhuaWang

  authored
  51da191

Commits on Aug 19, 2024

• DeepNVMe GDS (microsoft#5852)

  Show description for 5f0d177
  

  4 peopleauthored
  5f0d177

Commits on Aug 14, 2024

• Log operator warnings only in verbose mode (microsoft#5917)

  tjruwase

  authored
  0f2d485

Commits on Aug 1, 2024

• [ROCm] Get rocm version from /opt/rocm/.info/version (microsoft#5815)

  Show description for f82d088
  

  rraminen

  and

  loadams

  authored
  f82d088

Commits on Jul 31, 2024

• Fix CPU Adam JIT compilation (microsoft#5780)

  Show description for 681be6f
  

  lekurile

  and

  loadams

  authored
  681be6f

• Find ROCm on Fedora (microsoft#5705)

  Show description for 550f9c7
  

  trixirt

  and

  loadams

  authored
  550f9c7

• Fix op_builder for CUDA 12.5 (microsoft#5806)

  Show description for 08598db
  

  3 peopleauthored
  08598db

Commits on Jul 29, 2024

• Add fp8-fused gemm kernel (microsoft#5764)

  Show description for 4f95067
  

  sfc-gh-reyazda

  and

  jeffra

  authored
  4f95067

Commits on Jul 23, 2024

• Fix accuracy error of NPUFusedAdam (microsoft#5777)

  Show description for 5a100f6
  

  3 peopleauthored
  5a100f6

Commits on Jun 28, 2024

• Change source of CPUAdam for xpu accelerator (microsoft#5703)

  Show description for 4b8a4a0
  Liangliang-Ma

  authored
  4b8a4a0

Commits on Jun 24, 2024

• Fixed Windows inference build. (microsoft#5609)

  Show description for b3767d0
  

  4 peopleauthored
  b3767d0

Commits on Jun 20, 2024

• [XPU] support op builder from intel_extension_for_pytorch kernel path (microsoft#5425)

  Show description for d254d75
  

  3 peopleauthored
  d254d75

Commits on Jun 5, 2024

• Add Compressedbackend for Onebit optimizers (microsoft#5473)

  Show description for 11a62a0
  

  Liangliang-Ma

  and

  tjruwase

  authored
  11a62a0

Commits on May 31, 2024

• Update minor CUDA version compatibility (microsoft#5591)

  Show description for 1baf688
  adk9

  authored
  1baf688

Commits on May 28, 2024

• Rename files in fp_quantize op from quantize.* to fp_quantize.* (microsoft#5577)

  Show description for bf66acd
  loadams

  authored
  bf66acd

Commits on May 20, 2024

• CPUAdam fp16 and bf16 support (microsoft#5409)

  Show description for 69af361
  

  BacharL

  and

  tjruwase

  authored
  69af361

Commits on May 17, 2024

• Rocm warp size fix (microsoft#5402)

  Show description for 76c9c69
  

  rraminen

  and

  loadams

  authored
  76c9c69

Commits on May 10, 2024

• Fused adam for HPU (microsoft#5500)

  Show description for df4ef0a
  

  BacharL

  and

  tjruwase

  authored
  df4ef0a

Commits on Apr 17, 2024

• [CPU] Support SHM based inference_all_reduce in TorchBackend (microsoft#5391)

  Show description for b22706a
  

  3 peopleauthored
  b22706a

Commits on Apr 15, 2024

• Fix the FP6 kernels compilation problem on non-Ampere GPUs. (microsoft#5333)

  Show description for e3d873a
  

  3 peopleauthored
  e3d873a

Commits on Apr 4, 2024

• CPU SHM based inference_all_reduce improve (microsoft#5320)

  Show description for 731fd68
  

  3 peopleauthored
  731fd68

• FP [6,8,12] quantizer op (microsoft#5336)

  Show description for 3fbd01c
  

  jeffra

  and

  sfc-gh-reyazda

  authored
  3fbd01c

Commits on Mar 12, 2024

• Make op builder detection adapt to accelerator change (microsoft#5206)

  Show description for c08e69f
  

  3 peopleauthored
  c08e69f

Commits on Mar 8, 2024

• FP6 quantization end-to-end. (microsoft#5234)

  Show description for ccfdb84
  

  6 peopleauthored
  ccfdb84

Commits on Feb 21, 2024

• Use ninja to speed up build (microsoft#5088)

  Show description for b00533e
  

  4 peopleauthored
  b00533e

Commits on Feb 1, 2024

• [NPU] Add NPU to support hybrid engine (microsoft#4831)

  Show description for 88cca60
  

  6 peopleauthored
  88cca60

Pagination

PreviousNext