david/flash-attention

miroir de https://github.com/Dao-AILab/flash-attention

Auteur	SHA1 Message	Date
Tri Dao	88173a1aaf [FusedDense] Support relu, rename FusedDenseGeluDense -> FusedMLP	il y a 1 an
Tri Dao	43798966cf [Docs] Fix formatting	il y a 1 an
Tri Dao	3c7cbfc195 [Docs] Mention that dropout_layer_norm supports all dims up to 6k	il y a 1 an
Tri Dao	4a6eaa9f27 Update configs, add results	il y a 2 ans
Tri Dao	0bf5e50038 Release training code	il y a 2 ans