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

vulkan: implement GGML_OP_SUB

Browse source
This commit is contained in:
Rémy O 2025-02-08 13:40:29 +01:00
parent deb15e3f53
commit 148f58681b
4 changed files with 68 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

29
ggml/src/ggml-vulkan/vulkan-shaders/sub.comp Normal file
View file

@ -0,0 +1,29 @@
#version 450
#extension GL_EXT_shader_16bit_storage : require
#include "types.comp"
#include "generic_binary_head.comp"
const uint num_threads = 256;
layout(local_size_x = num_threads, local_size_y = 1, local_size_z = 1) in;
void main() {
uint idx = get_idx();
// num_threads * num_iter must equal 512, to match the wg_denoms and get_idx calculation
const uint num_iter = 2;
[[unroll]] for (uint i = 0; i < num_iter; ++i) {
if (idx >= p.ne) {
continue;
}
uint i00, i01, i02, i03;
get_indices(idx, i00, i01, i02, i03);
data_d[get_doffset() + dst_idx(i00, i01, i02, i03)] = D_TYPE(FLOAT_TYPE(data_a[get_aoffset() + src0_idx(i00, i01, i02, i03)]) - FLOAT_TYPE(data_b[get_boffset() + src1_idx(i00, i01, i02, i03)]));
idx += num_threads;
}
}

2
ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp
View file

@ -434,6 +434,8 @@ void process_shaders() {
string_to_spv("add_f32", "add.comp", {{"A_TYPE", "float"}, {"B_TYPE", "float"}, {"D_TYPE", "float"}, {"FLOAT_TYPE", "float"}});
string_to_spv("add_f16_f32_f16", "add.comp", {{"A_TYPE", "float16_t"}, {"B_TYPE", "float"}, {"D_TYPE", "float16_t"}, {"FLOAT_TYPE", "float"}});
string_to_spv("sub_f32", "sub.comp", {{"A_TYPE", "float"}, {"B_TYPE", "float"}, {"D_TYPE", "float"}, {"FLOAT_TYPE", "float"}});
string_to_spv("acc_f32", "acc.comp", {{"A_TYPE", "float"}, {"B_TYPE", "float"}, {"D_TYPE", "float"}, {"FLOAT_TYPE", "float"}});
string_to_spv("split_k_reduce", "mul_mat_split_k_reduce.comp", {});