david/aphrodite-engine: PygmalionAI's large-scale inference engine pygmalion.chat It is designed to serve as the inference endpoint for the PygmalionAI website, and to allow serving the Pygmalion models to a large number of users with blazing fast speeds (thanks to vLLM's Paged Attention). @ 884470ec36c0f1822dbecbdbbe7ff9f3b392384a

AlpinDale 39beed0b87 Revert "Refactor AWQ support."		hai 1 ano
..
attention	45f6d9f923 initial refactor commit	hai 1 ano
activation.cpp	32844c1522 add GELU kernels and remove compile bloat	hai 1 ano
activation_kernels.cu	32844c1522 add GELU kernels and remove compile bloat	hai 1 ano
attention.cpp	24c78e7306 optimization: multi-query attention kernel	hai 1 ano
cache.cpp	081545bde6 fix: various CUDA kernel tweaks	hai 1 ano
cache_kernels.cu	32844c1522 add GELU kernels and remove compile bloat	hai 1 ano
dispatch_utils.h	32844c1522 add GELU kernels and remove compile bloat	hai 1 ano
layernorm.cpp	081545bde6 fix: various CUDA kernel tweaks	hai 1 ano
layernorm_kernels.cu	32844c1522 add GELU kernels and remove compile bloat	hai 1 ano
pos_encoding.cpp	45f6d9f923 initial refactor commit	hai 1 ano
pos_encoding_kernels.cu	45f6d9f923 initial refactor commit	hai 1 ano
reduction.cuh	081545bde6 fix: various CUDA kernel tweaks	hai 1 ano