Commits

Branch selector

develop

User selector

All users

All time

Commit history

Commits on Dec 25, 2024

• Correct the dtype checking logics (#1775)

  poyenc

  authored
  4c2eff0View commit details

  Copy full SHA for 4c2eff0

  Browse repository at this point

Commits on Dec 23, 2024

• [CK_TILE] optimize moe-sorting kernel (#1771)

  Show description for 3d15f36
  carlushuang

  authored
  3d15f36View commit details

  Copy full SHA for 3d15f36

  Browse repository at this point

Commits on Dec 20, 2024

• fix typo for CK_USE_OCP_FP8 (#1769)

  illsilin

  authored
  07339c7View commit details

  Copy full SHA for 07339c7

  Browse repository at this point

• hot-fix (#1768)

  carlushuang

  authored
  1c45ca3View commit details

  Copy full SHA for 1c45ca3

  Browse repository at this point

• [CK_TILE] Add fmha fwd N-Warp S-Shuffle pipeline (fmha fwd splitkv pipeline variant) (#1705)

  Show description for 37cdbf4
  

  poyenc

  and

  qianfengz

  authored
  37cdbf4View commit details

  Copy full SHA for 37cdbf4

  Browse repository at this point

• fix profiler_grouped_gemm (#1766)

  illsilin

  authored
  2944c50View commit details

  Copy full SHA for 2944c50

  Browse repository at this point

Commits on Dec 19, 2024

• Apply Ck-tile argument parser for vectors [I/O] (#1758)

  Show description for e758d00
  mozga-amd

  authored
  e758d00View commit details

  Copy full SHA for e758d00

  Browse repository at this point

Commits on Dec 18, 2024

• [CK TILE] Refactor GemmKernel to be reused by other GEMM related operators (#1730)

  Show description for 453ca37
  aledudek

  authored
  453ca37View commit details

  Copy full SHA for 453ca37

  Browse repository at this point

• Disambiguate bit_cast (#1749)

  Show description for 1c1b336
  

  xw285cornell

  and

  poyenc

  authored
  1c1b336View commit details

  Copy full SHA for 1c1b336

  Browse repository at this point

• [CK_TILE] Move hipmalloc/memcpy calls out of gpu reference gemm (#1743)

  Show description for f6c4d61
  aledudek

  authored
  f6c4d61View commit details

  Copy full SHA for f6c4d61

  Browse repository at this point

Commits on Dec 17, 2024

• updated fp16 instances to be on parity with universal gemm instances (#1754)

  Show description for d9e37c6
  hsadasiv

  authored
  d9e37c6View commit details

  Copy full SHA for d9e37c6

  Browse repository at this point

• Pass build flags to config.h (#1760)

  Show description for 689a5ae
  illsilin

  authored
  689a5aeView commit details

  Copy full SHA for 689a5ae

  Browse repository at this point

• refactor conditional usage; fix build on rocm6.1 where the reference didn't exist

  tenpercent

  committed
  6ef8d3cView commit details

  Copy full SHA for 6ef8d3c

  Browse repository at this point

• Bump rocm-docs-core from 1.11.0 to 1.12.0 in /docs/sphinx (#1753)

  Show description for 0e54d7a
  dependabot[bot]

  authored
  0e54d7aView commit details

  Copy full SHA for 0e54d7a

  Browse repository at this point

• Added unit tests for CK Tile compute bound gemm pipeline (#1728)

  jakpiase

  authored
  627a27bView commit details

  Copy full SHA for 627a27b

  Browse repository at this point

• Enhance printing functionality (#1751)

  Show description for d46196f
  

  3 peopleauthored
  d46196fView commit details

  Copy full SHA for d46196f

  Browse repository at this point

Commits on Dec 16, 2024

• clarify release notes bullet point

  tenpercent

  committed
  0fd6978View commit details

  Copy full SHA for 0fd6978

  Browse repository at this point

• add contributing placeholder

  tenpercent

  committed
  1b75c77View commit details

  Copy full SHA for 1b75c77

  Browse repository at this point

• add pull request template placeholder

  tenpercent

  committed
  30a37caView commit details

  Copy full SHA for 30a37ca

  Browse repository at this point

• add template placeholders

  tenpercent

  committed
  a8ad7fcView commit details

  Copy full SHA for a8ad7fc

  Browse repository at this point

• upgrade sqlalchemy version (#1748)

  Show description for fdfe210
  illsilin

  authored
  fdfe210View commit details

  Copy full SHA for fdfe210

  Browse repository at this point

Commits on Dec 15, 2024

• added moe interleaving pipeline (#1712)

  Show description for f57d720
  

  shengnxu

  and

  root

  authored
  f57d720View commit details

  Copy full SHA for f57d720

  Browse repository at this point

Commits on Dec 14, 2024

• upgrade pandas package (#1746)

  illsilin

  authored
  d68974aView commit details

  Copy full SHA for d68974a

  Browse repository at this point

• Add zstd lib for building hipTensor. (#1745)

  Show description for 41ebf11
  illsilin

  authored
  41ebf11View commit details

  Copy full SHA for 41ebf11

  Browse repository at this point

Commits on Dec 13, 2024

• Add SplitK support into Batched GEMM V3 (#1729)

  Show description for 4d8fce3
  

  3 peopleauthored
  4d8fce3View commit details

  Copy full SHA for 4d8fce3

  Browse repository at this point

• Ck tile/smoothquant out stride (#1742)

  Show description for 4e73177
  junhaha666

  authored
  4e73177View commit details

  Copy full SHA for 4e73177

  Browse repository at this point

Commits on Dec 12, 2024

• [CK_TILE] naive attn (#1708)

  Show description for 77a38e0
  carlushuang

  authored
  77a38e0View commit details

  Copy full SHA for 77a38e0

  Browse repository at this point

Commits on Dec 10, 2024

• add missing stdexcept header (#1740)

  illsilin

  authored
  357a0b1View commit details

  Copy full SHA for 357a0b1

  Browse repository at this point

• Upgrade to Ubuntu22.04 as default OS. (#1738)

  Show description for 90d8410
  illsilin

  authored
  90d8410View commit details

  Copy full SHA for 90d8410

  Browse repository at this point

• Make sure we call __hneg with half to remove ambigios error (#1736)

  cjatin

  authored
  67497a0View commit details

  Copy full SHA for 67497a0

  Browse repository at this point

• [CK TILE] Use config name instead of data type in FmhaFwdTypeConfig<config> (#1731)

  Show description for 94ae711
  rocking5566

  authored
  94ae711View commit details

  Copy full SHA for 94ae711

  Browse repository at this point

Commits on Dec 9, 2024

• build CI for gfx12 by default (#1734)

  illsilin

  authored
  23cf202View commit details

  Copy full SHA for 23cf202

  Browse repository at this point

• update CI timeout limits (#1733)

  illsilin

  authored
  2f088b8View commit details

  Copy full SHA for 2f088b8

  Browse repository at this point

• remove unnecessary file (#1732)

  illsilin

  authored
  c773cc2View commit details

  Copy full SHA for c773cc2

  Browse repository at this point

Commits on Dec 6, 2024

• Refactor CI performance tests. (#1726)

  Show description for 355893c
  illsilin

  authored
  355893cView commit details

  Copy full SHA for 355893c

  Browse repository at this point