matt/ollama37
1
0
Fork 0
mirror of https://github.com/dogkeeper886/ollama37.git synced 2025-12-11 16:26:59 +00:00
Code Issues Packages Projects Releases Wiki Activity

Remove CPU build, fixup linux build script

Browse Source
This commit is contained in:
Daniel Hiltgen
2023-12-21 16:54:54 -08:00
parent 325d74985b
commit fa24e73b82
6 changed files with 21 additions and 58 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
llm/llama.cpp/gen_common.sh
View File

@@ -3,7 +3,7 @@
init_vars() {
LLAMACPP_DIR=gguf
PATCHES="0001-Expose-callable-API-for-server.patch"
CMAKE_DEFS="-DLLAMA_ACCELERATE=on"
CMAKE_DEFS="-DLLAMA_ACCELERATE=on -DLLAMA_SERVER_VERBOSE=off"
# TODO - LLAMA_K_QUANTS is stale and needs to be mapped to newer cmake settings
CMAKE_TARGETS="--target ggml --target ggml_static --target llama --target build_info --target common --target ext_server --target llava_static"
if echo "${CGO_CFLAGS}" | grep -- '-g' >/dev/null; then

5
llm/llama.cpp/gen_linux.sh
View File

@@ -22,13 +22,14 @@ if [ -z "${CUDACXX}" -a -x /usr/local/cuda/bin/nvcc ]; then
fi
COMMON_CMAKE_DEFS="-DCMAKE_POSITION_INDEPENDENT_CODE=on -DLLAMA_ACCELERATE=on -DLLAMA_NATIVE=off -DLLAMA_AVX=on -DLLAMA_AVX2=off -DLLAMA_AVX512=off -DLLAMA_FMA=off -DLLAMA_F16C=off"
OLLAMA_DYN_LIB_DIR="gguf/build/lib"
mkdir -p ${OLLAMA_DYN_LIB_DIR}
touch ${OLLAMA_DYN_LIB_DIR}/.generated
source $(dirname $0)/gen_common.sh
init_vars
git_module_setup
apply_patches
mkdir -p ${OLLAMA_DYN_LIB_DIR}
touch ${OLLAMA_DYN_LIB_DIR}/.generated
#
# CPU first for the default library
#

13
llm/llama.cpp/patches/0001-Expose-callable-API-for-server.patch
View File

@@ -1,4 +1,4 @@
From 7184ae16e8fd0e9e91cac4c81daa323057fa992b Mon Sep 17 00:00:00 2001
From 4c72576c5f6c2217b1ecf7fd8523616acc5526ae Mon Sep 17 00:00:00 2001
From: Daniel Hiltgen <daniel@ollama.com>
Date: Mon, 13 Nov 2023 12:25:58 -0800
Subject: [PATCH] Expose callable API for server
@@ -6,10 +6,10 @@ Subject: [PATCH] Expose callable API for server
This adds an extern "C" interface within the example server
---
examples/server/CMakeLists.txt | 24 +++
examples/server/server.cpp | 276 +++++++++++++++++++++++++++++++++
examples/server/server.cpp | 279 +++++++++++++++++++++++++++++++++
examples/server/server.h | 89 +++++++++++
ggml-cuda.cu | 1 +
4 files changed, 390 insertions(+)
4 files changed, 393 insertions(+)
create mode 100644 examples/server/server.h
diff --git a/examples/server/CMakeLists.txt b/examples/server/CMakeLists.txt
@@ -46,7 +46,7 @@ index 859cd12..4ea47a7 100644
+endif()
\ No newline at end of file
diff --git a/examples/server/server.cpp b/examples/server/server.cpp
index 0403853..065420c 100644
index 0403853..5e78e4d 100644
--- a/examples/server/server.cpp
+++ b/examples/server/server.cpp
@@ -5,6 +5,9 @@
@@ -67,7 +67,7 @@ index 0403853..065420c 100644
int main(int argc, char **argv)
{
#if SERVER_VERBOSE != 1
@@ -3123,3 +3127,275 @@ int main(int argc, char **argv)
@@ -3123,3 +3127,278 @@ int main(int argc, char **argv)
llama_backend_free();
return 0;
}
@@ -80,6 +80,9 @@ index 0403853..065420c 100644
+
+void llama_server_init(ext_server_params *sparams, ext_server_resp_t *err)
+{
+#if SERVER_VERBOSE != 1
+ log_disable();
+#endif
+ assert(err != NULL && sparams != NULL);
+ err->id = 0;
+ err->msg[0] = '\0';