From b8feff411fc170de48c592cced8660f9e692db2b Mon Sep 17 00:00:00 2001
From: Pierrick Hymbert <pierrick.hymbert@gmail.com>
Date: Thu, 21 Mar 2024 04:36:06 +0100
Subject: [PATCH] Avoir copying the entire vector

Co-authored-by: slaren <slarengh@gmail.com>
---
 llama.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/llama.cpp b/llama.cpp
index cd7a7b8d6..a8e6b4208 100644
--- a/llama.cpp
+++ b/llama.cpp
@@ -5251,7 +5251,7 @@ static bool llm_load_tensors(
     // load tensor data
     for (auto & it : ctx_bufs) {
         ggml_context * ctx = it.first;
-        std::vector<ggml_backend_buffer_t> bufs = it.second;
+        auto & bufs = it.second;
         if (!ml.load_all_data(ctx, progress_callback, progress_callback_user_data, bufs, use_mlock ? &model.mlock_mmaps : NULL)) {
             return false;
         }