β νΈλ λ©
Tangerine Indexλ ChatGPT μ΄ν λμ¨ νλ‘λνΈμ μ¬λλ€μ λ°μμ μ£μ΅λλ€.
flash-linear-attention
github.com
π Efficient implementations for emerging model architectures
μ€νμμ€
github.com λ°©λ¬ΈμΈλΆ μ¬μ΄νΈGitHub6,027 μ€νμΈλΆ μ¬μ΄νΈ
λΉμ·ν νλ‘λνΈ
μ€λͺ μ΄ λΉμ·ν μ
FlashQLA
high-performance linear attention kernel library built on TileLang
flash-sparse-attention
Trainable fast and memory-efficient sparse attention
ring-flash-attention
Ring attention implementation with flash attention
cuLA
CUDA kernels for linear attention variants, written in CuTe DSL and CUTLASS C++.
mav
Model Activity Visualiser
