vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Add some comments to satisfy PR reviewer

Browse source
This commit is contained in:
Iwan Kawrakow 2023-07-23 23:26:03 +03:00
parent 6baa4ead58
commit f3a92117a7
1 changed files with 6 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

6
ggml-cuda.cu
View file

@ -2349,6 +2349,9 @@ static void mul_mat_vec_q4_K_q8_1_cuda(const void * vx, const void * vy, float *
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(1, block_num_y, 1); const dim3 block_nums(1, block_num_y, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
// Note: we use QI4_K/2 instead of QI4_K to make the dot product template require 4 groups of quants to be processed per
// kernel call instead of 2. This results in a better perfmance because the cost of computing the k-quant scales
// is better amortized.
mul_mat_vec_q<QK_K, QI4_K/2, block_q4_K, vec_dot_q4_K_q8_1> mul_mat_vec_q<QK_K, QI4_K/2, block_q4_K, vec_dot_q4_K_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows);
} }
@ -2358,6 +2361,9 @@ static void mul_mat_vec_q5_K_q8_1_cuda(const void * vx, const void * vy, float *
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(1, block_num_y, 1); const dim3 block_nums(1, block_num_y, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
// Note: we use QI5_K/2 instead of QI5_K to make the dot product template require 4 groups of quants to be processed per
// kernel call instead of 2. This results in a better perfmance because the cost of computing the k-quant scales
// is better amortized.
mul_mat_vec_q<QK_K, QI5_K/2, block_q5_K, vec_dot_q5_K_q8_1> mul_mat_vec_q<QK_K, QI5_K/2, block_q5_K, vec_dot_q5_K_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows);
} }