Lumpiasty/llama.cpp
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

ggml : update ggml_prec specification (#26675)

Browse Source 
* ggml : update ggml_prec specification

[no ci]

* cont : add GGML_PREC_BF16

* cont : rework API

* cont : use new API

* cont : swap arg order

* cont : support for MUL_MAT_ID

* cont : fix accidental remove of "break;"

* cont : return bools, add doc TAG_GGML_PREC, clean-up

* cont : add search tag

* cont : ws
This commit is contained in:
Georgi Gerganov
2026-09-08 09:06:24 +03:00
committed by GitHub
parent 9dcf84e5ae
commit 5a6caa05fc
10 changed files with 134 additions and 22 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tools/tuning/fa-vec.cpp
+1 -1
View File
@@ -56,7 +56,7 @@ static ggml_tensor * fa_build_graph(ggml_context * ctx, const fa_shape & s) {
ggml_set_name(m, "m");
ggml_tensor * out = ggml_flash_attn_ext(ctx, q, k, v, m, 1.0f / sqrtf((float) s.dk), 0.0f, 0.0f);
ggml_flash_attn_ext_set_prec(out, GGML_PREC_F32);
ggml_prec_set_acc(out, GGML_PREC_F32);
ggml_set_name(out, "out");
return out;