Repository navigation

[perf] Add opt-in SM120 sparse attention forward kernel for MiniMax-H3 - #2048

Open

zhongwujie wants to merge 8 commits into

hao-ai-lab:mainfrom

zhongwujie:feat/h3-sm120-sparse-attention

Open

[perf] Add opt-in SM120 sparse attention forward kernel for MiniMax-H3#2048
zhongwujie wants to merge 8 commits into
hao-ai-lab:mainfrom
zhongwujie:feat/h3-sm120-sparse-attention

Commits on Nov 1, 2026
- [perf] Add opt-in SM120 sparse attention for MiniMax-H3
  zhongwujie
  committed
  a991a1e
- [style] Apply pre-commit formatting
  zhongwujie
  committed
  bd12d9d
- change the test script
  zhongwujie
  committed
  9dec654
Commits on Nov 7, 2026
- [refactor] Organize Gluon sparse attention implementation
  zhongwujie
  committed
  77dd1bc
Commits on Nov 8, 2026
- [bugfix] Initialize SM120 state in the H3 compile test
  zhongwujie
  committed
  a6cb034
- Rename the H3 Gluon attention switch
  zhongwujie
  committed
  ee58825
- [misc] Merge upstream main and resolve documentation conflict
  zhongwujie
  committed
  27723cb
Commits on Nov 9, 2026
- [test] Run H3 Gluon configuration checks on non-SM120 devices
  zhongwujie
  committed
  442d75b