LLM-Inference/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-08-12 22:29:39 +04:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #1977 from ikawrakow/ik/dflash_fix_cpu

Browse Source 
Fix DFlash on the CPU
This commit is contained in:
Kawrakow
2026-06-16 15:26:23 +02:00
committed by GitHub
parent f9078e169b 6f45163a95
commit 8420f91ae3
1 changed files with 7 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/graphs/build_dflash.cpp
+7
View File
@@ -366,6 +366,13 @@ ggml_cgraph * llm_build_context::build_dflash() {
cb(Vcur, "dflash_main_v_pad", il);
}
if (Kcur->type == GGML_TYPE_F32) {
Kcur = ggml_cast(ctx0, Kcur, GGML_TYPE_F16);
}
if (Vcur->type == GGML_TYPE_F32) {
Vcur = ggml_cast(ctx0, Vcur, GGML_TYPE_F16);
}
cb(Qcur, "Qcur", il);
ggml_tensor * q = ggml_permute(ctx0, Qcur, 0, 2, 1, 3);