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). @ sampler_order_v2

AlpinDale e31c6f0b45 feat: refactor modeling logic and support more models (#274)		11 mesiacov pred
..
all_reduce	6305e6f3f2 fix: no repeated IPC registration (#227)	1 rok pred
attention	8e1cd54497 fix: do not include fp8 for rocm (#271)	11 mesiacov pred
hadamard	5d288aa76c feat: add fast hadamard transformation kernels (#232)	11 mesiacov pred
moe	7d6ba53602 feat: fused top-k kernels for MoE (#273)	11 mesiacov pred
punica	c0aac15421 feat: S-LoRA support (#222)	1 rok pred
quantization	8e1cd54497 fix: do not include fp8 for rocm (#271)	11 mesiacov pred
activation_kernels.cu	e31c6f0b45 feat: refactor modeling logic and support more models (#274)	11 mesiacov pred
cache.h	31c95011a6 feat: FP8 E5M2 KV Cache (#226)	1 rok pred
cache_kernels.cu	8e1cd54497 fix: do not include fp8 for rocm (#271)	11 mesiacov pred
cuda_compat.h	8fa608aeb7 feat: replace Ray with NCCL for control plane comms (#221)	1 rok pred
cuda_utils.h	31c95011a6 feat: FP8 E5M2 KV Cache (#226)	1 rok pred
cuda_utils_kernels.cu	31c95011a6 feat: FP8 E5M2 KV Cache (#226)	1 rok pred
dispatch_utils.h	d9b65e6c5f feat: DeepSeek MoE support (#237)	11 mesiacov pred
layernorm_kernels.cu	8fa608aeb7 feat: replace Ray with NCCL for control plane comms (#221)	1 rok pred
ops.h	e31c6f0b45 feat: refactor modeling logic and support more models (#274)	11 mesiacov pred
pos_encoding_kernels.cu	8fa608aeb7 feat: replace Ray with NCCL for control plane comms (#221)	1 rok pred
pybind.cpp	e31c6f0b45 feat: refactor modeling logic and support more models (#274)	11 mesiacov pred
reduction.cuh	2755a48d51 merge dev branch into main (#153)	1 rok pred