Files
llama.cpp/ggml-cuda/getrows.cuh
T

6 lines
141 B
Plaintext
Raw Normal View History

2024-03-25 13:50:23 +01:00
#include "common.cuh"
#define CUDA_GET_ROWS_BLOCK_SIZE 256
void ggml_cuda_op_get_rows(ggml_backend_cuda_context & ctx, ggml_tensor * dst);