Datasets:

Modalities:
Text
Formats:
text
Size:
< 1K
ArXiv:
Libraries:
Datasets
llama.cpp / ggml /src /ggml-virtgpu /virtgpu-apir.h
dlxj
todo: 基于 CUDA 13.0 编译
2517be1
#include "backend/shared/apir_backend.h"
#include "ggml-alloc.h"
#include "ggml-impl.h"
#include "ggml.h"
#include "virtgpu-shm.h"
#include "virtgpu-utils.h"
struct apir_buffer_context_t {
apir_buffer_host_handle_t host_handle;
struct virtgpu_shmem shmem;
apir_buffer_type_host_handle_t buft_host_handle;
};
#include "virtgpu-forward.gen.h"