sleepy/llama.cpp
1
0
Fork 0
Code Issues 6 Pull Requests 2 Actions 210 Packages Projects Releases Wiki Activity

vulkan: Implement mmvq for iq1_s/iq1_m (#18450)

Browse Source
This commit is contained in:
Jeff Bolz
2026-01-02 13:19:04 -06:00
committed by GitHub
parent 5755e52d15
commit 706e3f93a6
5 changed files with 433 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp
+2 -2
View File
@@ -685,7 +685,7 @@ void process_shaders() {
// mul mat vec with integer dot product
#if defined(GGML_VULKAN_INTEGER_DOT_GLSLC_SUPPORT)
if (is_legacy_quant(tname) || tname == "mxfp4" || is_k_quant(tname)) {
if (is_legacy_quant(tname) || tname == "mxfp4" || is_k_quant(tname) || tname == "iq1_s" || tname == "iq1_m") {
string_to_spv("mul_mat_vec_" + tname + "_q8_1_f32", "mul_mat_vecq.comp", merge_maps(base_dict, {{data_a_key, "1"}, {"D_TYPE", "float"}, {"FLOAT_TYPE", "float"}, {"FLOAT_TYPE_VEC2", "vec2"}, {"ACC_TYPE", "float"}}));
string_to_spv("mul_mat_vec_" + tname + "_q8_1_f32_subgroup", "mul_mat_vecq.comp", merge_maps(base_dict, {{data_a_key, "1"}, {"D_TYPE", "float"}, {"FLOAT_TYPE", "float"}, {"FLOAT_TYPE_VEC2", "vec2"}, {"ACC_TYPE", "float"}, {"USE_SUBGROUP_ADD", "1"}}));
string_to_spv("mul_mat_vec_" + tname + "_q8_1_f32_subgroup_no_shmem", "mul_mat_vecq.comp", merge_maps(base_dict, {{data_a_key, "1"}, {"D_TYPE", "float"}, {"FLOAT_TYPE", "float"}, {"FLOAT_TYPE_VEC2", "vec2"}, {"ACC_TYPE", "float"}, {"USE_SUBGROUP_ADD_NO_SHMEM", "1"}}));
@@ -1123,7 +1123,7 @@ void write_output_files() {
for (const std::string& btype : btypes) {
for (const auto& tname : type_names) {
if (btype == "q8_1" && !is_legacy_quant(tname) && tname != "mxfp4" && !is_k_quant(tname)) {
if (btype == "q8_1" && !is_legacy_quant(tname) && tname != "mxfp4" && !is_k_quant(tname) && tname != "iq1_s" && tname != "iq1_m") {
continue;
}
hdr << "extern const void * arr_dmmv_" << tname << "_" << btype << "_f32_data[3];\n";