vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-13 18:39:48 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

Merge remote-tracking branch 'jeffbolz/scalar_fa_3' into concedo_experimental

Browse source 
# Conflicts:
#	.github/workflows/build.yml
This commit is contained in:
Concedo 2025-05-09 17:19:33 +08:00
commit 0874cd231a
3 changed files with 645 additions and 93 deletions
Download patch file Download diff file Expand all files Collapse all files

12
ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp
View file

@ -435,7 +435,6 @@ void process_shaders() {
#endif
}
#if defined(GGML_VULKAN_COOPMAT2_GLSLC_SUPPORT)
// flash attention
for (const auto& f16acc : {false, true}) {
std::string acctype = f16acc ? "float16_t" : "float";
@ -446,6 +445,7 @@ void process_shaders() {
}
if (tname == "bf16") continue;
#if defined(GGML_VULKAN_COOPMAT2_GLSLC_SUPPORT)
if (tname == "f16") {
string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn_cm2.comp",
merge_maps(base_dict, {{"Q_TYPE", "float"}, {"D_TYPE", "float"}, {"ACC_TYPE", acctype}}), true, false, true, f16acc);
@ -454,9 +454,17 @@ void process_shaders() {
string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn_cm2.comp",
merge_maps(base_dict, {{data_a_key, "1"}, {"Q_TYPE", "float"}, {"D_TYPE", "float"}, {"ACC_TYPE", acctype}, {"DEQUANTFUNC", "dequantFunc"+to_uppercase(tname) }, {"BLOCK_SIZE", "QUANT_K_"+to_uppercase(tname) }}), true, false, true, f16acc);
}
#endif
if (tname == "f16") {
string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn.comp",
merge_maps(base_dict, {{"Q_TYPE", "float"}, {"D_TYPE", "float"}, {"ACC_TYPE", acctype}}), true, false, false, f16acc);
} else if (tname == "q4_0" || tname == "q8_0") {
std::string data_a_key = "DATA_A_" + to_uppercase(tname);
string_to_spv("flash_attn_f32_f16_" + tname, "flash_attn.comp",
merge_maps(base_dict, {{data_a_key, "1"}, {"Q_TYPE", "float"}, {"D_TYPE", "float"}, {"ACC_TYPE", acctype}, {"BLOCK_SIZE", "QUANT_K_"+to_uppercase(tname) }}), true, false, false, f16acc);
}
}
}
#endif
for (const auto& tname : type_names) {
// mul mat vec