[BugFix] Unwrap AttentionMask wrapper for transformers 5.9 in sequence parallel#206

Merged

tastelikefeet merged 4 commits into

modelscope:mainfrom

May 28, 2026

Commits on May 26, 2026

[Feature] Enhance NPU kernel support for Qwen3.5 MoE and FLA
ys2025-AI
committed
[Fix] Improve NPU expert weight caching and FLA model resolution
ys2025-AI
committed

Commits on May 28, 2026