mirror of
https://github.com/tcsenpai/ollama.git
synced 2025-06-18 08:40:19 +00:00
6 lines
200 B
Plaintext
6 lines
200 B
Plaintext
#include "common.cuh"
|
|
|
|
#define CUDA_QUANTIZE_BLOCK_SIZE 256
|
|
|
|
void quantize_row_q8_1_cuda(const float * x, void * vy, const int64_t kx, const int64_t ky, const int64_t kx_padded, cudaStream_t stream);
|