ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-04-26 17:39:37 +00:00
Code Issues Packages Projects Releases Wiki Activity

Add command line option to merge experts up/gate

Browse Source
This commit is contained in:
Kawrakow
2026-01-12 12:49:18 +02:00
parent 80f2b090d5
commit 7671335ac9
8 changed files with 20 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
include/llama.h
View File

@@ -392,6 +392,7 @@ extern "C" {
bool use_thp; // use transparent huge pages (linux only)
bool validate_quants; // if true, check for NaNs while loading the model
bool merge_qkv; // if true, merge separate Q, K, V tensors into a single, contiguous tensor
bool merge_up_gate_exps; // if true, merge ffn_up_exps and ffn_gate_exps tensors into a single, contiguous tensor
};
// NOTE: changing the default values of parameters marked as [EXPERIMENTAL] may cause crashes or incorrect results in certain configurations