Repository navigation

[shardformer]: fix qwen2/qwen3 causal mask with pipeline + sequence parallelism and no flash attention - #6451

Open

LiRunGuo wants to merge 2 commits into

hpcaitech:mainfrom

LiRunGuo:fix/qwen-pp-sp-attn-mask

Open

[shardformer]: fix qwen2/qwen3 causal mask with pipeline + sequence parallelism and no flash attention#6451
LiRunGuo wants to merge 2 commits into
hpcaitech:mainfrom
LiRunGuo:fix/qwen-pp-sp-attn-mask

Commits on Oct 13, 2026
- [shardformer] fix qwen2/qwen3 causal mask with pipeline + sequence parallelism
  LiRunGuo
  committed
  6ee408e
Commits on Oct 14, 2026
- [pre-commit.ci] auto fixes from pre-commit.com hooks
  pre-commit-ci[bot]
  committed
  1593654