Commits

Branch selector

main

User selector

All users

All time

Commit history

Commits on Jun 2, 2026

• [float8] fix `==` typo in `float8_transpose` axiswise dim remapping (#4038)

  Show description for cc49e86
  GotFusion

  authored
  cc49e86View commit details

  Copy full SHA for cc49e86

  Browse repository at this point

Commits on Jun 1, 2026

• [CI] Switch release_model H100 job to linux_job_v3 (OSDC/ARC) (#4457)

  Show description for 8c244cb
  huydhn

  authored
  8c244cbView commit details

  Copy full SHA for 8c244cb

  Browse repository at this point

• [CI] Move H100/A100 benchmark & tutorial jobs to OSDC (ARC) (#4458)

  Show description for 18278f9
  huydhn

  authored
  18278f9View commit details

  Copy full SHA for 18278f9

  Browse repository at this point

• [CI] Switch H100 jobs to linux_job_v3 (OSDC/ARC) (#4456)

  Show description for 211cf47
  huydhn

  authored
  211cf47View commit details

  Copy full SHA for 211cf47

  Browse repository at this point

Commits on May 29, 2026

• Add "32x32 transposed" variant to MXFP8 3D quantization kernel (#4447)

  alexsamardzic

  authored
  9c010aeView commit details

  Copy full SHA for 9c010ae

  Browse repository at this point

• clean up tutorials (#4449)

  Show description for 8826885
  vkuzo

  authored
  8826885View commit details

  Copy full SHA for 8826885

  Browse repository at this point

Commits on May 28, 2026

• [mxfp8 training] don't use offs as key in cutedsl compile cache (#4442)

  danielvegamyhre

  authored
  49783ecView commit details

  Copy full SHA for 49783ec

  Browse repository at this point

• [mxfp8 moe training] add support for bias parm in _to_mxfp8_then_scaled_grouped_mm (#4386)

  danielvegamyhre

  authored
  620db65View commit details

  Copy full SHA for 620db65

  Browse repository at this point

• [float8] fix `cc1` typo in `Float8LinearConfig.__post_init__` operand-precision assertion (#4115)

  Show description for 2fc4b1a
  

  GotFusion

  and

  wangzhenwu

  authored
  2fc4b1aView commit details

  Copy full SHA for 2fc4b1a

  Browse repository at this point

• Add _grouped_mm dispatch handler to Float8Tensor (#4390)

  Show description for 329d8b6
  xiaowangintel

  authored
  329d8b6View commit details

  Copy full SHA for 329d8b6

  Browse repository at this point

• Back out D105996100 — breaks `fbpkg build light_cli` (S668656)

  Show description for 038b088
  jerryzh168

  authored
  038b088View commit details

  Copy full SHA for 038b088

  Browse repository at this point

Commits on May 27, 2026

• Add `RopeSDPAFusionPass(CustomGraphPass)` with `uuid()` (#4423)

  Show description for 3179676
  frgossen

  authored
  3179676View commit details

  Copy full SHA for 3179676

  Browse repository at this point

• Convert `quant_lift_up` to `CustomGraphPass` with `uuid()` (#4424)

  Show description for 0133053
  frgossen

  authored
  0133053View commit details

  Copy full SHA for 0133053

  Browse repository at this point

• Enable building under a free-threaded Python interpreter (#4428)

  Show description for 3ddc90c
  rgommers

  authored
  3ddc90cView commit details

  Copy full SHA for 3ddc90c

  Browse repository at this point

• create a new CI job with clang (#4440)

  Show description for d74c75f
  vkuzo

  authored
  d74c75fView commit details

  Copy full SHA for d74c75f

  Browse repository at this point

Commits on May 26, 2026

• Use transpose_result=True for fp8_sparse_mm (#4433) (#4433)

  Show description for b4805e2
  gyllstromk

  authored
  b4805e2View commit details

  Copy full SHA for b4805e2

  Browse repository at this point

• [nvfp4_training] Support NVFP4 linear training on Blackwell (#4369)

  Show description for c81eb79
  

  rdspring1

  and

  tbqh

  authored
  c81eb79View commit details

  Copy full SHA for c81eb79

  Browse repository at this point

Commits on May 25, 2026

• [CPU] skip registration of PT2E inductor passes if mkldnn not available (#4435)

  Xia-Weiwen

  authored
  8bbe1a5View commit details

  Copy full SHA for 8bbe1a5

  Browse repository at this point

Commits on May 23, 2026

• fix rocm CI (#4437)

  Show description for e1a6499
  vkuzo

  authored
  e1a6499View commit details

  Copy full SHA for e1a6499

  Browse repository at this point

Commits on May 22, 2026

• fix ruff on main (#4436)

  Show description for fae917b
  vkuzo

  authored
  fae917bView commit details

  Copy full SHA for fae917b

  Browse repository at this point

• Add algorithm search to FP8 sparse linear benchmark (#4432) (#4432)

  Show description for 15a9736
  gyllstromk

  authored
  15a9736View commit details

  Copy full SHA for 15a9736

  Browse repository at this point

• switch CI to PyTorch 2.12 (#4402)

  Show description for 246348d
  vkuzo

  authored
  246348dView commit details

  Copy full SHA for 246348d

  Browse repository at this point

• drop PyTorch <2.11 support (#4401)

  Show description for e545e35
  vkuzo

  authored
  e545e35View commit details

  Copy full SHA for e545e35

  Browse repository at this point

• drop PyTorch <2.10 support (#4400)

  Show description for 8e40aef
  vkuzo

  authored
  8e40aefView commit details

  Copy full SHA for 8e40aef

  Browse repository at this point

• drop PyTorch<2.9 support (#4399)

  Show description for b1ebe79
  vkuzo

  authored
  b1ebe79View commit details

  Copy full SHA for b1ebe79

  Browse repository at this point

• add int8 qkv-fusion pass for x86 (#3986)

  yuchengliu1

  authored
  f9fc641View commit details

  Copy full SHA for f9fc641

  Browse repository at this point

Commits on May 21, 2026

• Move collect_producer_nodes to graph_utils.py (#4294)

  Show description for 6181601
  tom-arm

  authored
  6181601View commit details

  Copy full SHA for 6181601

  Browse repository at this point

• fix(setup.py): exclude top-level test/ from find_packages (#4361)

  Show description for 918b32d
  Anai-Guo

  authored
  918b32dView commit details

  Copy full SHA for 918b32d

  Browse repository at this point

• [float8] read `dim` from kwargs in `Float8Tensor` aten.split handler (#4429)

  Show description for cd9f0e8
  GotFusion

  authored
  cd9f0e8View commit details

  Copy full SHA for cd9f0e8

  Browse repository at this point

• Fix Python 3.14 compatibility of quantization/pt2e (#4427)

  Show description for 5e92da4
  rgommers

  authored
  5e92da4View commit details

  Copy full SHA for 5e92da4

  Browse repository at this point

• Fix build failure with GCC 14 in `rowwise_scaled_linear_sparse_cutlass_f8f8.cu` (#4426)

  Show description for 55c9d5e
  

  rgommers

  and

  claude

  authored
  55c9d5eView commit details

  Copy full SHA for 55c9d5e

  Browse repository at this point

• [test][xpu] Fix RuntimeError for nf4 cases after the running of test_float8_tensor.py (#4298)

  Show description for a04d2ac
  zxd1997066

  authored
  a04d2acView commit details

  Copy full SHA for a04d2ac

  Browse repository at this point

Commits on May 19, 2026

• [X86] multi-ISA portable kernel compilation and runtime dispatch (#4159)

  Show description for b506da5
  

  3 peopleauthored
  b506da5View commit details

  Copy full SHA for b506da5

  Browse repository at this point

• [CPU] disable a few test cases due to an issue in torch nightly (#4425)

  Xia-Weiwen

  authored
  bc73441View commit details

  Copy full SHA for bc73441

  Browse repository at this point

Commits on May 14, 2026

• drop PyTorch 2.8 support from the code (#4398)

  Show description for 13cd013
  vkuzo

  authored
  13cd013View commit details

  Copy full SHA for 13cd013

  Browse repository at this point