Merge pull request #3925 from dhiltgen/bump

Bump llama.cpp to b2737

Merge pull request #3925 from dhiltgen/bump
Bump llama.cpp to b2737
36666c21 · Daniel Hiltgen · GitHub · d459dc4a · 85801317 · 46e12c46
Unverified Commit 36666c21 authored Apr 26, 2024 by Daniel Hiltgen Committed by GitHub Apr 26, 2024
Hide whitespace changes
Inline Side-by-side

Showing with 13 additions and 1 deletion

llm/llama.cpp llm/llama.cpp +1 -1

llm/patches/02-clip-log.diff llm/patches/02-clip-log.diff +12 -0

No files found.
--- a/llama.cpp @ 46e12c46
+++ b/llama.cpp @ 46e12c46
-Subproject commit 7593639ce335e8d7f89aa9a54d616951f273af60
+Subproject commit 46e12c4692a37bdd31a0432fc5153d7d22bc7f72
--- a/llm/patches/02-clip-log.diff
+++ b/llm/patches/02-clip-log.diff
+diff --git a/examples/llava/clip.cpp b/examples/llava/clip.cpp
+index e431c7f7..f077e688 100644
+--- a/examples/llava/clip.cpp
+++ b/examples/llava/clip.cpp
+@@ -3,6 +3,7 @@
+ // I'll gradually clean and extend it
+ // Note: Even when using identical normalized image inputs (see normalize_image_u8_to_f32()) we have a significant difference in resulting embeddings compared to pytorch
+ #include "clip.h"
+#include "common.h"
+ #include "log.h"
+ #include "ggml.h"
+ #include "ggml-alloc.h"