Datasets:

echodict
/

llama.cpp

Modalities:

Formats:

Size:

Libraries:

llama.cpp / ggml /src /ggml-cuda /template-instances /fattn-tile-instance-dkq72-dv72.cu

dlxj

todo: 基于 CUDA 13.0 编译

2517be1 about 1 month ago

144 Bytes

	// This file has been autogenerated by generate_cu_files.py, do not edit manually.

	#include "../fattn-tile.cuh"

	DECL_FATTN_TILE_CASE(72, 72);