chore: ⬆️ Update ggml-org/llama.cpp to 6bf28f0111ff9f21b3c1b1eace20c590281e7ba6 (#5127)

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

Makefile

@@ -6,7 +6,7 @@ BINARY_NAME=local-ai
 DETECT_LIBS?=true
 
 # llama.cpp versions
-CPPLLAMA_VERSION?=3e1d29348b5d77269f6931500dd1c1a729d429c8
+CPPLLAMA_VERSION?=6bf28f0111ff9f21b3c1b1eace20c590281e7ba6
 
 # whisper.cpp version
 WHISPER_REPO?=https://github.com/ggerganov/whisper.cpp

backend/cpp/llama/CMakeLists.txt

@@ -2,7 +2,7 @@
 ## XXX: In some versions of CMake clip wasn't being built before llama.
 ## This is an hack for now, but it should be fixed in the future.
 set(TARGET myclip)
-add_library(${TARGET} clip.cpp clip.h llava.cpp llava.h)
+add_library(${TARGET} clip.cpp clip.h clip-impl.h llava.cpp llava.h)
 install(TARGETS ${TARGET} LIBRARY)
 target_include_directories(myclip PUBLIC .)
 target_include_directories(myclip PUBLIC ../..)

backend/cpp/llama/grpc-server.cpp

@@ -514,7 +514,7 @@ struct llama_server_context
             LOG_INFO("Multi Modal Mode Enabled", {});
             clp_ctx = clip_init(params.mmproj.path.c_str(), clip_context_params {
                 /* use_gpu */ has_gpu,
-                /*verbosity=*/ 1,
+                /*verbosity=*/ GGML_LOG_LEVEL_INFO,
             });
             if(clp_ctx == nullptr) {
                 LOG_ERR("unable to load clip model: %s", params.mmproj.path.c_str());

backend/cpp/llama/prepare.sh

@@ -21,6 +21,7 @@ fi
 ## XXX: In some versions of CMake clip wasn't being built before llama.
 ## This is an hack for now, but it should be fixed in the future.
 cp -rfv llama.cpp/examples/llava/clip.h llama.cpp/examples/grpc-server/clip.h
+cp -rfv llama.cpp/examples/llava/clip-impl.h llama.cpp/examples/grpc-server/clip-impl.h
 cp -rfv llama.cpp/examples/llava/llava.cpp llama.cpp/examples/grpc-server/llava.cpp
 echo '#include "llama.h"' > llama.cpp/examples/grpc-server/llava.h
 cat llama.cpp/examples/llava/llava.h >> llama.cpp/examples/grpc-server/llava.h