ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-01-26 17:20:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

Set default MLA to 3 also in llama-bench (#949)

Browse Source 
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
This commit is contained in:
Kawrakow
2025-11-13 09:52:06 +02:00
committed by GitHub
parent bbc127d10e
commit 38abd0e289
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
examples/llama-bench/llama-bench.cpp
View File

@@ -289,7 +289,7 @@ static const cmd_params cmd_params_defaults = {
/* main_gpu */ {0},
/* no_kv_offload */ {false},
/* flash_attn */ {true},
/* mla_attn */ {0},
/* mla_attn */ {3},
/* attn_max_batch */ {0},
/* ser */ {{-1,0.0f}},
/* tensor_split */ {std::vector<float>(llama_max_devices(), 0.0f)},