david/flash-attention: flash-attention from https://github.com/Dao-AILab/flash-attention @ b5a89310427f0f08f95caa7c333c16991bd40cc4

огледало от https://github.com/Dao-AILab/flash-attention

Chirag Jain 50896ec574 Make nvcc threads configurable via environment variable (#885)		преди 11 месеца
..
rotary.cpp	e45a46a5b7 [Rotary] Implement GPT-J style (interleaved) rotary	преди 1 година
rotary_cuda.cu	e45a46a5b7 [Rotary] Implement GPT-J style (interleaved) rotary	преди 1 година
setup.py	50896ec574 Make nvcc threads configurable via environment variable (#885)	преди 11 месеца