david/flash-attention

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

Auteur	SHA1 Message	Date
Tri Dao	393882bc08 [LayerNorm] Implement LN with parallel residual, support dim 8k	il y a 1 an
Tri Dao	6738d9477d [LayerNorm] Implement RMS Norm	il y a 1 an
Tri Dao	ae137ed17a [LayerNorm] Fuse LayerScale	il y a 2 ans
Tri Dao	8c6609ae1a [LayerNorm] Support all dimensions up to 6k (if divisible by 8)	il y a 2 ans
Tri Dao	fa6d1ce44f Add fused_dense and dropout_add_layernorm CUDA extensions	il y a 2 ans