This is a Triton implementation of the Flash Attention v2 algorithm from Tri Dao (https://tridao.me/publications/flash2/flash2.pdf) ...
SRC: .safetensors (lazy) or torch pickle .pth/.pt/.ckpt (safe weights_only load, held in RAM).
Some results have been hidden because they may be inaccessible to you
Show inaccessible results