[Experimental] Scaled Dot Product Attention FlashAttention Algorithm Conversion #147

Open

Open

[Experimental] Scaled Dot Product Attention FlashAttention Algorithm Conversion#147

Assignees

Milestone

0.1 CPU - General

Utilize flash attention algorithm to optimize the performance of attention

Metadata

Assignees

yifeizh2

Labels

No labels

No labels

Type

No type

Projects

No projects

Milestone

0.1 CPU - GeneralNo due date

Relationships

None yet

Development

No branches or pull requests