Inicio Explorar Axuda
Rexistro Iniciar sesión
david
/
flash-attention
réplica de https://github.com/Dao-AILab/flash-attention
1
0
Fork 0
Ficheiros Incidencias 0 Wiki
Árbore: 6807b1ea37
Ramas Etiquetas

Commit History

Autor SHA1 Mensaxe Data
  Tri Dao 2c996ca25f Use SeqlenInfo for bwd and epilogue hai 1 semana
  Tri Dao 82dc825759 Don't use the unsafe convert_type function hai 2 semanas
  Tri Dao e7b93e3902 Clean up mha_combine kernel hai 3 semanas
  Tri Dao 3d0d147940 Early stop on actual num_splits in mha_combine kernel hai 3 semanas
  Tri Dao 9fd6b977bb Precompute the pointers in mha_combine kernel hai 3 semanas
  Tri Dao 01de45b8f8 Add combine kernel hai 1 mes