david/flash-attention: flash-attention from https://github.com/Dao-AILab/flash-attention

Tri Dao 02541ac9e8 [CE] Assert logit_scale > 0		1 月之前
..
triton	02541ac9e8 [CE] Assert logit_scale > 0	1 月之前
__init__.py	ece539abd6 Add __init__.py files to subdirectories for installation	2 年之前
activations.py	7153673c1a Fix swiglu backwards return type (#1337)	3 月之前
fused_dense.py	0cb595ad94 [bugfix] handle_x not define when using checkpoint_lvl = 2 (#502)	1 年之前
layer_norm.py	f1a73d0740 Run isort and black on python files	1 年之前
rms_norm.py	f1a73d0740 Run isort and black on python files	1 年之前