transformers
[Qwen3Next] preserve linear-attn-mask optimization under torch.compile/export
#46148

Open

Commits

[Qwen3.5] preserve linear-attn-mask optimization under torch.compile/export

yuvrajsharma9981 committed 2 days ago
Apply fix at modular source + regenerate affected modeling files

yuvrajsharma9981 committed 2 days ago

Loading