Change flash attention to be on by default

2026-04-25 17:09:22 +00:00 · 2025-10-25 09:32:01 +03:00
parent 9dc0c89bc9
commit 6d05977940
4 changed files with 18 additions and 18 deletions
--- a/examples/llama-bench/llama-bench.cpp
+++ b/examples/llama-bench/llama-bench.cpp
@@ -285,7 +285,7 @@ static const cmd_params cmd_params_defaults = {
    /* split_mode           */ {LLAMA_SPLIT_MODE_LAYER},
    /* main_gpu             */ {0},
    /* no_kv_offload        */ {false},
-    /* flash_attn           */ {false},
+    /* flash_attn           */ {true},
    /* mla_attn             */ {0},
    /* attn_max_batch       */ {0},
    /* ser                  */ {{-1,0.0f}},