llama_cpp_sys 0.3.2

Automatically-generated bindings to llama.cpp's C API
Documentation
1
2
3
4
5
#include "common.cuh"

#define CUDA_QUANTIZE_BLOCK_SIZE 256

void quantize_row_q8_1_cuda(const float * x, void * vy, const int kx, const int ky, const int kx_padded, cudaStream_t stream);