[Kernel] Less compute for Self-Attention (Q * K) - Githubissues

intel / xFasterTransformer

Apache License 2.0

270 stars 53 forks source link

[Kernel] Less compute for Self-Attention (Q * K) #420

Closed pujiang2018 closed 1 month ago