Python

flash-attention

Fast and memory-efficient exact attention

PythonEmerging

GitHub

Stars

—

Forks

—

Contributors

Last push

4mo ago

Recent commits

Latest commits.

basics working (#2070)
ac9b5f1Driss Guessous6mo ago
[FIRST] Fix softcap scoremod kwargs typo. (#2072)
0a5339fLeo Dong6mo ago
[CUTE] Seeing if tvvm reduces cpu overhead (#2042)
179f793Driss Guessous6mo ago
[Cute,Fwd] Extend score_mod to variable sequence length (#2043)
fd8d5ebReuben Stern6mo ago
[CUTE] Allow grads to be preallocated (#2065)
e240e0fDriss Guessous6mo ago

Fix use-after-free in FA3 deterministic mode. The pytorch caching allocator actually saves us here, but if you turn it off, then compute-sanitizer will detect this. (#2063)

Recent commits

Top contributors