LLM inference in C/C++
C++ ggml
cmake : introduce semantic versioning (#26839)
Daniel Bevenius680a9ae63d60
diff --git a/.github/workflows/build-cmake-pkg.yml b/.github/workflows/build-cmake-pkg.ymlindex 5becff09c1..a958870de3 100644--- a/.github/workflows/build-cmake-pkg.yml+++ b/.github/workflows/build-cmake-pkg.yml@@ -21,6 +21,7 @@ jobs:-DLLAMA_BUILD_TOOLS=OFF \-DLLAMA_BUILD_EXAMPLES=OFF \-DLLAMA_BUILD_APP=OFF \+ -DLLAMA_BUILD_IS_DEV=OFF \-DCMAKE_BUILD_TYPE=Releasecmake --build build --config Releasecmake --install build --prefix "$PREFIX" --config Release@@ -29,7 +30,12 @@ jobs:tclsh <<'EOF'set build(commit) [string trim [exec git rev-parse --short HEAD]]set build(number) [string trim [exec git rev-list --count HEAD]]- set build(version) "0.0.$build(number)"++ set cmakelists [read [open "CMakeLists.txt" r]]+ regexp {set\(LLAMA_VERSION_MAJOR\s+(\d+)\)} $cmakelists -> major+ regexp {set\(LLAMA_VERSION_MINOR\s+(\d+)\)} $cmakelists -> minor+ regexp {set\(LLAMA_VERSION_PATCH\s+(\d+)\)} $cmakelists -> patch+ set build(version) "$major.$minor.$patch"set llamaconfig [read [open "$env(LLAMA_CONFIG)" r]]set checks [list "set\\(LLAMA_VERSION \\s+$build(version)\\)" \diff --git a/.github/workflows/build-cpu.yml b/.github/workflows/build-cpu.ymlindex 30b07ce788..57622e1d45 100644--- a/.github/workflows/build-cpu.yml+++ b/.github/workflows/build-cpu.yml@@ -95,7 +95,8 @@ jobs:run: |cmake -B build \-DLLAMA_FATAL_WARNINGS=ON \- -DGGML_RPC=ON+ -DGGML_RPC=ON \+ -DGGML_NATIVE=OFFtime cmake --build build --config Release -j $(nproc)- name: Testdiff --git a/.github/workflows/make-release.yml b/.github/workflows/make-release.ymlnew file mode 100644index 0000000000..fed9c877c7--- /dev/null+++ b/.github/workflows/make-release.yml@@ -0,0 +1,46 @@+name: Make Release++on:+ workflow_dispatch:+ inputs:+ dry_run:+ description: 'Dry run - validate without creating the tag'+ required: true+ type: boolean+ default: true++env:+ GH_TOKEN: ${{ github.token }}++permissions:+ contents: write++jobs:+ make-release:+ runs-on: ubuntu-latest++ steps:+ - name: Checkout+ uses: actions/checkout@v6++ - name: Run release checks+ id: checks+ run: bash scripts/make-release-checks.sh ${{ github.event.inputs.dry_run == 'true' && '--dry-run' || '' }}+ env:+ GITHUB_REPOSITORY: ${{ github.repository }}++ - name: Create release tag+ if: ${{ github.event.inputs.dry_run == 'false' }}+ run: |+ VERSION="${{ steps.checks.outputs.version }}"+ git config user.name "github-actions[bot]"+ git config user.email "github-actions[bot]@users.noreply.github.com"+ git tag -a "${VERSION}" -m "Release ${VERSION}"+ git push origin "${VERSION}"+ echo "Created and pushed tag ${VERSION}"++ - name: Dry run summary+ if: ${{ github.event.inputs.dry_run == 'true' }}+ run: |+ echo "Dry run complete - all checks passed."+ echo "Would have created tag: ${{ steps.checks.outputs.version }}"diff --git a/.github/workflows/winget.yml b/.github/workflows/winget.ymlindex 69e24f9400..c0a814f3ad 100644--- a/.github/workflows/winget.yml+++ b/.github/workflows/winget.yml@@ -19,6 +19,8 @@ jobs:run: |cargo binstall [email protected] -y+ # TODO: This should later be updated to publish releases instead of+ # development release builds.- name: Find latest releaseid: find_latest_releaseuses: actions/github-script@v8diff --git a/CMakeLists.txt b/CMakeLists.txtindex 3df1d82dbe..b2092d12dd 100644--- a/CMakeLists.txt+++ b/CMakeLists.txt@@ -2,6 +2,26 @@ cmake_minimum_required(VERSION 3.14...3.28) # for add_link_options and implicitproject("llama.cpp" C CXX)include(CheckIncludeFileCXX)+### llama.cpp version+set(LLAMA_VERSION_MAJOR 0)+set(LLAMA_VERSION_MINOR 1)+set(LLAMA_VERSION_PATCH 0)+set(LLAMA_VERSION_BASE "${LLAMA_VERSION_MAJOR}.${LLAMA_VERSION_MINOR}.${LLAMA_VERSION_PATCH}")++# whether this is a development/nightly build+# set this to OFF when making a release from a release tag (vX.Y.Z)+# ref: https://github.com/ggml-org/ggml/discussions/1579+option(LLAMA_BUILD_IS_DEV "llama: dev build" ON)++if (LLAMA_BUILD_IS_DEV)+ set(LLAMA_VERSION "${LLAMA_VERSION_BASE}-dev")+else()+ # TODO: check that the current commit is tagged correctly according to the version specified above+ set(LLAMA_VERSION "${LLAMA_VERSION_BASE}")+endif()++message(STATUS "llama.cpp version: ${LLAMA_VERSION}")+#set(CMAKE_WARN_DEPRECATED YES)set(CMAKE_WARN_UNUSED_CLI YES)@@ -24,9 +44,6 @@ if (CMAKE_SOURCE_DIR STREQUAL CMAKE_CURRENT_SOURCE_DIR)set(LLAMA_STANDALONE ON)include(git-vars)-- # configure project version- # TODOelse()set(LLAMA_STANDALONE OFF)endif()@@ -139,7 +156,6 @@ endif()if (NOT DEFINED LLAMA_BUILD_COMMIT)set(LLAMA_BUILD_COMMIT ${BUILD_COMMIT})endif()-set(LLAMA_INSTALL_VERSION 0.0.${LLAMA_BUILD_NUMBER})# override ggml optionsset(GGML_ALL_WARNINGS ${LLAMA_ALL_WARNINGS})@@ -275,12 +291,12 @@ configure_package_config_file(LLAMA_BIN_INSTALL_DIR )write_basic_package_version_file(- ${CMAKE_CURRENT_BINARY_DIR}/llama-version.cmake- VERSION ${LLAMA_INSTALL_VERSION}+ ${CMAKE_CURRENT_BINARY_DIR}/llama-config-version.cmake+ VERSION ${LLAMA_VERSION}COMPATIBILITY SameMajorVersion)install(FILES ${CMAKE_CURRENT_BINARY_DIR}/llama-config.cmake- ${CMAKE_CURRENT_BINARY_DIR}/llama-version.cmake+ ${CMAKE_CURRENT_BINARY_DIR}/llama-config-version.cmakeDESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/llama)configure_file(cmake/llama.pc.indiff --git a/README.md b/README.mdindex fc7633a7c5..1b341e7fbc 100644--- a/README.md+++ b/README.md@@ -106,6 +106,7 @@ The `llama.cpp` project is build on top of the [ggml](https://github.com/ggml-or- [XCFramework](docs/xcframework.md)- [Completions](docs/completions.md)- [Models](docs/models.md)+- [Release process](docs/release.md)## Contributingdiff --git a/app/llama.cpp b/app/llama.cppindex 2cf1aa876c..3b7e46f20d 100644--- a/app/llama.cpp+++ b/app/llama.cpp@@ -1,5 +1,7 @@#include "build-info.h"+#include "llama.h"+#include <cstdio>#include <cstdlib>#include <string>@@ -77,12 +79,12 @@ static const command cmds[] = {#undef UPDATE_HIDDEN-static int version(int argc, char ** argv) {- printf("%s\n", llama_build_info());+static int version(int /*argc*/, char ** /*argv*/) {+ llama_print_build_info(llama_version());return 0;}-static int licenses(int argc, char ** argv) {+static int licenses(int /*argc*/, char ** /*argv*/) {for (int i = 0; LICENSES[i]; ++i) {printf("%s\n", LICENSES[i]);}diff --git a/cmake/llama-config.cmake.in b/cmake/llama-config.cmake.inindex b4defc76ff..6db73577ae 100644--- a/cmake/llama-config.cmake.in+++ b/cmake/llama-config.cmake.in@@ -1,4 +1,4 @@-set(LLAMA_VERSION @LLAMA_INSTALL_VERSION@)+set(LLAMA_VERSION @LLAMA_VERSION@)set(LLAMA_BUILD_COMMIT @LLAMA_BUILD_COMMIT@)set(LLAMA_BUILD_NUMBER @LLAMA_BUILD_NUMBER@)set(LLAMA_SHARED_LIB @BUILD_SHARED_LIBS@)diff --git a/cmake/llama.pc.in b/cmake/llama.pc.inindex 6fb58b5f68..31b043c0e3 100644--- a/cmake/llama.pc.in+++ b/cmake/llama.pc.in@@ -5,6 +5,6 @@ includedir=@CMAKE_INSTALL_FULL_INCLUDEDIR@Name: llamaDescription: Port of Facebook's LLaMA model in C/C++-Version: @LLAMA_INSTALL_VERSION@+Version: @LLAMA_VERSION@Libs: -L${libdir} -lggml -lggml-base -lllamaCflags: -I${includedir}diff --git a/common/CMakeLists.txt b/common/CMakeLists.txtindex 799d227519..d6cfc9a008 100644--- a/common/CMakeLists.txt+++ b/common/CMakeLists.txt@@ -121,8 +121,8 @@ add_library(${TARGET})set_target_properties(${TARGET} PROPERTIES- VERSION ${LLAMA_INSTALL_VERSION}- SOVERSION 0+ VERSION ${LLAMA_VERSION_BASE}+ SOVERSION ${LLAMA_VERSION_MAJOR}MACHO_CURRENT_VERSION 0 # keep macOS linker from seeing oversized version number)diff --git a/common/arg.cpp b/common/arg.cppindex cb314eee70..b2fddfe5f2 100644--- a/common/arg.cpp+++ b/common/arg.cpp@@ -1390,8 +1390,7 @@ common_params_context common_params_parser_init(common_params & params, llama_ex{"--version"},"show version and build info",[](common_params &) {- fprintf(stderr, "version: %d (%s)\n", llama_build_number(), llama_commit());- fprintf(stderr, "built with %s for %s\n", llama_compiler(), llama_build_target());+ llama_print_build_info(llama_version());exit(0);}));diff --git a/common/build-info.cpp.in b/common/build-info.cpp.inindex f888fd079f..4ec3397081 100644--- a/common/build-info.cpp.in+++ b/common/build-info.cpp.in@@ -29,7 +29,7 @@ const char * llama_build_info(void) {return s.c_str();}-void llama_print_build_info(void) {- fprintf(stderr, "%s: build = %d (%s)\n", __func__, llama_build_number(), llama_commit());- fprintf(stderr, "%s: built with %s for %s\n", __func__, llama_compiler(), llama_build_target());+void llama_print_build_info(const char * llama_version) {+ fprintf(stderr, "version: %s (build %d, commit %s)\n", llama_version, llama_build_number(), llama_commit());+ fprintf(stderr, "built with %s for %s\n", llama_compiler(), llama_build_target());}diff --git a/common/build-info.h b/common/build-info.hindex 382cfa7850..1e564591a6 100644--- a/common/build-info.h+++ b/common/build-info.h@@ -8,4 +8,4 @@ const char * llama_compiler(void);const char * llama_build_target(void);const char * llama_build_info(void);-void llama_print_build_info(void);+void llama_print_build_info(const char *);diff --git a/docs/release.md b/docs/release.mdnew file mode 100644index 0000000000..4335ef9d42--- /dev/null+++ b/docs/release.md@@ -0,0 +1,49 @@+# Release process++llama.cpp uses [semantic versioning](https://semver.org) (`MAJOR.MINOR.PATCH`).++## Version bump guidelines++| Change type | Version component |+|---|---|+| Breaking change to the public C API (`include/llama.h`) | `MAJOR` |+| Backward-compatible features, model support, or API addition | `MINOR` |+| Bug fix with no API change | `PATCH` |++The version is set in the three variables at the top of the root `CMakeLists.txt`:++```cmake+set(LLAMA_VERSION_MAJOR 0)+set(LLAMA_VERSION_MINOR 1)+set(LLAMA_VERSION_PATCH 0)+```++_A version bump should be included in the PR that introduces the change, or in a+dedicated bump commit merged before the release is cut._++_TODO: add PR labels (`semver: patch`, `semver: minor`, `semver: major`) to help+identify which PRs require a version bump before cutting a release._++## Making a release++Releases are created by running the [make-release](.github/workflows/make-release.yml)+which is a manual workflow.++The workflow creates an annotated git tag (e.g. `v0.1.0`) and pushes it to the+remote. No GitHub Release object is created, the tag is the release artifact.++## Building a release++By default, `LLAMA_BUILD_IS_DEV=ON` which appends a `-dev` suffix to `LLAMA_VERSION`,+marking the build as a nightly/development build. Distributors building from a+release tag must pass `-DLLAMA_BUILD_IS_DEV=OFF` to produce a clean version string+(e.g. `0.1.0` instead of `0.1.0-dev`).++## How releases reach users+Currently releases are not published to github releases, only nightly/development+builds are available there. The way users can access releases are using the following+channels:++- **llama-install.sh** — downloads pre-built binaries built from the release tag.+- **Package managers** — consume the git tag directly.+- **Build from source** — users clone the repo and check out the tag.diff --git a/examples/test-cmake/.gitignore b/examples/test-cmake/.gitignorenew file mode 100644index 0000000000..0ddff317a4--- /dev/null+++ b/examples/test-cmake/.gitignore@@ -0,0 +1,3 @@+llama-build-install+install+builddiff --git a/examples/test-cmake/CMakeLists.txt b/examples/test-cmake/CMakeLists.txtnew file mode 100644index 0000000000..ed5cb1f3c2--- /dev/null+++ b/examples/test-cmake/CMakeLists.txt@@ -0,0 +1,13 @@+cmake_minimum_required(VERSION 3.14)+project(llama-simple)++set(CMAKE_CXX_STANDARD 17)++find_package(llama 0.1.0 REQUIRED)++add_executable(test-cmake test-cmake.cpp)+target_link_libraries(test-cmake PRIVATE llama)+target_compile_definitions(test-cmake PRIVATE+ LLAMA_BUILD_NUMBER=${LLAMA_BUILD_NUMBER}+ LLAMA_BUILD_COMMIT="${LLAMA_BUILD_COMMIT}"+)diff --git a/examples/test-cmake/README.md b/examples/test-cmake/README.mdnew file mode 100644index 0000000000..21e5eb9607--- /dev/null+++ b/examples/test-cmake/README.md@@ -0,0 +1,36 @@+## cmake-test++This is just for manually testing/developing of a llama.cpp installation to+enable troubleshooting issues and exploration. The idea is that this can be used+after making changes to llama.cpp installation cmake configuration and then+verify it locally.++### Usage+The following will configure, build, and install llama.cpp++Configuring/build/install:+```console+./build-install.sh+```+The above command will create a directory named `install` in the current directory+which will have the follwing files in its lib directory:+```console+(venv) $ ls install/lib/+cmake libggml.so libllama-common.so.0 libllama.so.0.1.0 llama.cpp+libggml-base.so libggml.so.0 libllama-common.so.0.1.0 libmtmd.so pkgconfig+libggml-base.so.0 libggml.so.0.19.0 libllama.so libmtmd.so.0+libggml-base.so.0.19.0 libllama-common.so libllama.so.0 libmtmd.so.0.1.0+```++Build/run this project using the installation created above:+```console+(venv) $ ./build.sh+-- Configuring done (0.0s)+-- Generating done (0.0s)+-- Build files have been written to: /home/danbev/work/ai/llama.cpp/examples/test-cmake/build+[100%] Built target test-cmake+[test-cmake] Using llama.cpp version 0.1.0-dev-b10335+[test-cmake] Initializing backend...+load_backend: loaded CPU backend from /home/danbev/work/ai/llama.cpp/examples/test-cmake/install/lib/llama.cpp/libggml-cpu-alderlake.so+[test-cmake] Backend initialized.+```diff --git a/examples/test-cmake/build-install.sh b/examples/test-cmake/build-install.shnew file mode 100755index 0000000000..77a6713d67--- /dev/null+++ b/examples/test-cmake/build-install.sh@@ -0,0 +1,19 @@+#!/bin/bash++set -e++rm -rf llama-build-install install++cmake --fresh -S ../../. -B llama-build-install -DCMAKE_BUILD_TYPE=Release \+ -DBUILD_SHARED_LIBS=ON \+ -DGGML_BACKEND_DL=ON \+ -DGGML_CPU_ALL_VARIANTS=ON \+ -DLLAMA_TESTS_INSTALL=OFF \+ -DCMAKE_INSTALL_PREFIX="${PWD}/install" \+ -DGGML_BACKEND_DIR="${PWD}/install/lib/llama.cpp" \+ -DGGML_LIB_INSTALL_DIR="${PWD}/install/lib/llama.cpp" \+ -DLLAMA_LIB_INSTALL_DIR="${PWD}/install/lib/llama.cpp" \+ -DLLAMA_TOOLS_INSTALL=OFF++cmake --build llama-build-install --parallel 12+cmake --install llama-build-installdiff --git a/examples/test-cmake/build.sh b/examples/test-cmake/build.shnew file mode 100755index 0000000000..a212732b89--- /dev/null+++ b/examples/test-cmake/build.sh@@ -0,0 +1,7 @@+#!/bin/bash++set -e++cmake -S . -B build -DCMAKE_PREFIX_PATH="${PWD}/install"+cmake --build build+LD_LIBRARY_PATH="${PWD}/install/lib/llama.cpp:${PWD}/install/lib${LD_LIBRARY_PATH:+:$LD_LIBRARY_PATH}" ./build/test-cmakediff --git a/examples/test-cmake/test-cmake.cpp b/examples/test-cmake/test-cmake.cppnew file mode 100644index 0000000000..c5c4765b43--- /dev/null+++ b/examples/test-cmake/test-cmake.cpp@@ -0,0 +1,12 @@+#include "llama.h"+#include <cstdio>++int main(void) {+ printf("[test-cmake] version: %s, build: %d (%s)\n",+ llama_version(), LLAMA_BUILD_NUMBER, LLAMA_BUILD_COMMIT);+ printf("[test-cmake] Initializing backend...\n");+ llama_backend_init();+ printf("[test-cmake] Backend initialized.\n");+ llama_backend_free();+ return 0;+}diff --git a/include/llama.h b/include/llama.hindex ef278c9c32..177fc10a91 100644--- a/include/llama.h+++ b/include/llama.h@@ -457,6 +457,8 @@ extern "C" {// lora adapterstruct llama_adapter_lora;+ LLAMA_API const char * llama_version(void);+// Helpers for getting default parameters// TODO: update API to start accepting pointers to params structs (https://github.com/ggml-org/llama.cpp/discussions/9172)LLAMA_API struct llama_model_params llama_model_default_params(void);diff --git a/scripts/make-release-checks.sh b/scripts/make-release-checks.shnew file mode 100755index 0000000000..c8c6322841--- /dev/null+++ b/scripts/make-release-checks.sh@@ -0,0 +1,83 @@+#!/bin/bash+# Run all pre-release checks and determine the release version.+#+# Usage: make-release-checks.sh [--dry-run]+# --dry-run: warn on failures instead of aborting+#+# Env (when running in GitHub Actions): GH_TOKEN, GITHUB_REPOSITORY, GITHUB_OUTPUT+set -euo pipefail++SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"+REPO_ROOT="$(cd "$SCRIPT_DIR/.." && pwd)"++DRY_RUN=false+for arg in "$@"; do+ case "$arg" in+ --dry-run) DRY_RUN=true ;;+ *) echo "Unknown argument: $arg"; exit 1 ;;+ esac+done++MAJOR=$(grep "set(LLAMA_VERSION_MAJOR" "$REPO_ROOT/CMakeLists.txt" | grep -oP '\d+')+MINOR=$(grep "set(LLAMA_VERSION_MINOR" "$REPO_ROOT/CMakeLists.txt" | grep -oP '\d+')+PATCH=$(grep "set(LLAMA_VERSION_PATCH" "$REPO_ROOT/CMakeLists.txt" | grep -oP '\d+')+VERSION="v${MAJOR}.${MINOR}.${PATCH}"+echo "Determined version: ${VERSION}"+if [[ -n "${GITHUB_OUTPUT:-}" ]]; then+ echo "version=${VERSION}" >> "$GITHUB_OUTPUT"+fi++echo "Checking that tag ${VERSION} does not already exist..."+if git ls-remote --tags origin "${VERSION}" | grep -q "${VERSION}"; then+ echo "Error: tag ${VERSION} already exists on remote"+ exit 1+fi+echo "Tag ${VERSION} does not exist on remote - OK"++SHA=$(git rev-parse HEAD)+echo "Checking release.yml status for commit ${SHA}..."+if [[ -z "${GITHUB_REPOSITORY:-}" ]]; then+ echo "Warning: GITHUB_REPOSITORY not set - skipping CI check (local run)"+else+ RUNS=$(gh api "repos/${GITHUB_REPOSITORY}/actions/workflows/release.yml/runs" \+ --jq "[.workflow_runs[] | select(.head_sha == \"${SHA}\" and .conclusion == \"success\")] | length")+ if [[ "$RUNS" -eq 0 ]]; then+ if [[ "$DRY_RUN" == "true" ]]; then+ echo "Warning: no successful release.yml run found for HEAD (${SHA}) (dry run, continuing)."+ else+ echo "Error: no successful release.yml run found for HEAD (${SHA})"+ echo "The nightly build must complete successfully before making a release."+ exit 1+ fi+ else+ echo "Found successful release.yml run for HEAD."+ fi+fi++MAJOR=$(grep "set(GGML_VERSION_MAJOR" "$REPO_ROOT/ggml/CMakeLists.txt" | grep -oP '\d+')+MINOR=$(grep "set(GGML_VERSION_MINOR" "$REPO_ROOT/ggml/CMakeLists.txt" | grep -oP '\d+')+PATCH=$(grep "set(GGML_VERSION_PATCH" "$REPO_ROOT/ggml/CMakeLists.txt" | grep -oP '\d+')+GGML_VERSION="v${MAJOR}.${MINOR}.${PATCH}"+echo "Local ggml version: ${GGML_VERSION}"++if ! git clone --depth 1 --branch "${GGML_VERSION}" https://github.com/ggml-org/ggml.git upstream-ggml 2>/dev/null; then+ echo "Warning: tag ${GGML_VERSION} not found in upstream ggml - skipping comparison"+else+ echo "Comparing local ggml/ src and include with upstream ${GGML_VERSION}..."+ DIFF=$(diff -rq "$REPO_ROOT/ggml/src" upstream-ggml/src 2>&1 || true)+ DIFF+=$(diff -rq "$REPO_ROOT/ggml/include" upstream-ggml/include 2>&1 || true)+ DIFF+=$(diff "$REPO_ROOT/ggml/CMakeLists.txt" upstream-ggml/CMakeLists.txt 2>&1 || true)+ rm -rf upstream-ggml+ if [[ -n "$DIFF" ]]; then+ echo "local ggml/ differs from upstream ${GGML_VERSION}:"+ echo "$DIFF"+ if [[ "$DRY_RUN" == "true" ]]; then+ echo "Warning: would abort release due to ggml mismatch (dry run, continuing)."+ else+ echo "Error: ggml must match upstream before making a release."+ exit 1+ fi+ else+ echo "local ggml/ matches upstream ${GGML_VERSION}"+ fi+fidiff --git a/src/CMakeLists.txt b/src/CMakeLists.txtindex 24f05cc916..39ba3061f7 100644--- a/src/CMakeLists.txt+++ b/src/CMakeLists.txt@@ -45,11 +45,16 @@ add_library(llama)set_target_properties(llama PROPERTIES- VERSION ${LLAMA_INSTALL_VERSION}- SOVERSION 0+ VERSION ${LLAMA_VERSION_BASE}+ SOVERSION ${LLAMA_VERSION_MAJOR}MACHO_CURRENT_VERSION 0 # keep macOS linker from seeing oversized version number)+target_compile_definitions(llama PRIVATE+ LLAMA_VERSION="${LLAMA_VERSION}"+ LLAMA_COMMIT="${LLAMA_BUILD_COMMIT}"+)+target_include_directories(llama PRIVATE .)target_include_directories(llama PUBLIC ../include)target_compile_features (llama PRIVATE cxx_std_17) # don't bumpdiff --git a/src/llama.cpp b/src/llama.cppindex 94c8f60e0c..9ff1902fc1 100644--- a/src/llama.cpp+++ b/src/llama.cpp@@ -114,6 +114,10 @@ bool llama_supports_rpc(void) {return ggml_backend_reg_by_name("RPC") != nullptr;}+const char * llama_version(void) {+ return LLAMA_VERSION;+}+void llama_backend_init(void) {ggml_time_init();diff --git a/tests/test-quantize-stats.cpp b/tests/test-quantize-stats.cppindex c655575340..e07d75b7e7 100644--- a/tests/test-quantize-stats.cpp+++ b/tests/test-quantize-stats.cpp@@ -301,7 +301,7 @@ int main(int argc, char ** argv) {return 1;}- llama_print_build_info();+ llama_print_build_info(llama_version());// load the modelfprintf(stderr, "Loading model\n");diff --git a/tools/cvector-generator/cvector-generator.cpp b/tools/cvector-generator/cvector-generator.cppindex 8c6b3d868d..558c37e612 100644--- a/tools/cvector-generator/cvector-generator.cpp+++ b/tools/cvector-generator/cvector-generator.cpp@@ -421,7 +421,7 @@ int main(int argc, char ** argv) {params.cb_eval_user_data = &cb_data;params.warmup = false;- llama_print_build_info();+ llama_print_build_info(llama_version());llama_backend_init();llama_numa_init(params.numa);diff --git a/tools/gguf-split/gguf-split.cpp b/tools/gguf-split/gguf-split.cppindex 5cafcc9aa9..c6cdbb98e2 100644--- a/tools/gguf-split/gguf-split.cpp+++ b/tools/gguf-split/gguf-split.cpp@@ -106,7 +106,7 @@ static void split_params_parse_ex(int argc, const char ** argv, split_params & psplit_print_usage(argv[0]);exit(0);} else if (arg == "--version") {- fprintf(stderr, "version: %d (%s)\n", llama_build_number(), llama_commit());+ fprintf(stderr, "version: %s (build %d, commit %s)\n", llama_version(), llama_build_number(), llama_commit());fprintf(stderr, "built with %s for %s\n", llama_compiler(), llama_build_target());exit(0);} else if (arg == "--dry-run") {diff --git a/tools/mtmd/CMakeLists.txt b/tools/mtmd/CMakeLists.txtindex 3f4a6c670d..769a44e0b7 100644--- a/tools/mtmd/CMakeLists.txt+++ b/tools/mtmd/CMakeLists.txt@@ -72,8 +72,8 @@ add_library(mtmd)set_target_properties(mtmd PROPERTIES- VERSION ${LLAMA_INSTALL_VERSION}- SOVERSION 0+ VERSION ${LLAMA_VERSION_BASE}+ SOVERSION ${LLAMA_VERSION_MAJOR}MACHO_CURRENT_VERSION 0 # keep macOS linker from seeing oversized version number)diff --git a/tools/quantize/quantize.cpp b/tools/quantize/quantize.cppindex 15ef64c4b0..8d03c8fcd4 100644--- a/tools/quantize/quantize.cpp+++ b/tools/quantize/quantize.cpp@@ -611,7 +611,7 @@ int llama_quantize(int argc, char ** argv) {}}- llama_print_build_info();+ llama_print_build_info(llama_version());if (params.dry_run) {fprintf(stderr, "%s: calculating quantization size for '%s' as %s", __func__, fname_inp.c_str(), ftype_str.c_str());