sci-ml/flash-attn
Flash Attention: Fast and Memory-Efficient Exact Attention (Python component).
Screenshots
-
flash-attn-2.8.3~amd64cuda rocm python_single_target_python3_11 python_single_target_python3_12 python_single_target_python3_13 debug
View 
Download 
Browse License: BSD Overlay: tatsh-overlay