LLM inference in C/C++
C++ ggml
ci: Clean up UI builds from releases (#27706)
Aleksander Grygier4d19b287691e
diff --git a/.github/workflows/docker.yml b/.github/workflows/docker.ymlindex be2f840b1a..1de25b522d 100644--- a/.github/workflows/docker.yml+++ b/.github/workflows/docker.yml@@ -64,7 +64,7 @@ jobs:needs: create_taguses: ./.github/workflows/ui-build.ymlwith:- hf_ui_version: ${{ needs.create_tag.outputs.source_tag }}+ ui_version: ${{ needs.create_tag.outputs.source_tag }}prepare_matrices:name: Prepare Docker matrices@@ -162,7 +162,7 @@ jobs:if: ${{ matrix.config.prebuilt_ui == true }}uses: actions/download-artifact@3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c # v8with:- name: ui-build+ name: llama-ui.zippath: tools/ui/dist- name: Set up QEMUdiff --git a/.github/workflows/release.yml b/.github/workflows/release.ymlindex 5815e5c475..ee1b6e8a50 100644--- a/.github/workflows/release.yml+++ b/.github/workflows/release.yml@@ -61,31 +61,8 @@ jobs:echo "should_release=false" >> $GITHUB_OUTPUTfi- get-version:- runs-on: ubuntu-slim- outputs:- ui_version: ${{ steps.version.outputs.ui_version }}- steps:- - uses: actions/checkout@v6- with:- fetch-depth: 0- - id: version- run: |- # Resolve UI version: BUILD_NUMBER from cmake/build-info.cmake > git hash + epoch > fallback- version=""- if grep -q "BUILD_NUMBER" cmake/build-info.cmake; then- build_number=$(grep "set(BUILD_NUMBER" cmake/build-info.cmake | grep -oP '\d+')- if [ -n "$build_number" ] && [ "$build_number" -gt 0 ]; then- version="b${build_number}"- fi- fi- if [ -z "$version" ]; then- version=$(git rev-parse --short HEAD)-$(date +%s)- fi- echo "ui_version=${version}" >> $GITHUB_OUTPUT-macos-cpu:- needs: [check-release, get-version]+ needs: [check-release, ui-build]if: ${{ needs.check-release.outputs.should_release == 'true' }}strategy:matrix:@@ -119,12 +96,11 @@ jobs:with:fetch-depth: 0- - name: Setup Node.js- uses: actions/setup-node@v6+ - name: Download UI build+ uses: actions/download-artifact@v7with:- node-version: "24"- cache: "npm"- cache-dependency-path: "tools/ui/package-lock.json"+ name: llama-ui.zip+ path: tools/ui/dist- name: ccacheuses: ggml-org/[email protected]@@ -141,7 +117,6 @@ jobs:-DCMAKE_BUILD_WITH_INSTALL_RPATH=ON \-DLLAMA_FATAL_WARNINGS=ON \-DLLAMA_BUILD_BORINGSSL=ON \- -DHF_UI_VERSION=${{ needs.get-version.outputs.ui_version }} \${{ env.CMAKE_ARGS }}cmake --build build --config Release -j $(sysctl -n hw.logicalcpu)@@ -167,7 +142,7 @@ jobs:key: release-${{ matrix.os }}-${{ matrix.arch }}ubuntu-cpu:- needs: [check-release, get-version]+ needs: [check-release, ui-build]if: ${{ needs.check-release.outputs.should_release == 'true' }}strategy:matrix:@@ -191,12 +166,11 @@ jobs:with:fetch-depth: 0- - name: Setup Node.js- uses: actions/setup-node@v6+ - name: Download UI build+ uses: actions/download-artifact@v7with:- node-version: "24"- cache: "npm"- cache-dependency-path: "tools/ui/package-lock.json"+ name: llama-ui.zip+ path: tools/ui/dist- name: Dependenciesid: depends@@ -227,7 +201,6 @@ jobs:-DGGML_NATIVE=OFF \-DGGML_CPU_ALL_VARIANTS=ON \-DLLAMA_FATAL_WARNINGS=ON \- -DHF_UI_VERSION=${{ needs.get-version.outputs.ui_version }} \${{ env.CMAKE_ARGS }}cmake --build build --config Release -j $(nproc)@@ -254,7 +227,7 @@ jobs:key: release-${{ matrix.os }}-cpuubuntu-vulkan:- needs: [check-release, get-version]+ needs: [check-release, ui-build]if: ${{ needs.check-release.outputs.should_release == 'true' }}strategy:@@ -277,12 +250,11 @@ jobs:with:fetch-depth: 0- - name: Setup Node.js- uses: actions/setup-node@v6+ - name: Download UI build+ uses: actions/download-artifact@v7with:- node-version: "24"- cache: "npm"- cache-dependency-path: "tools/ui/package-lock.json"+ name: llama-ui.zip+ path: tools/ui/dist- name: Dependenciesid: depends@@ -314,7 +286,6 @@ jobs:-DGGML_NATIVE=OFF \-DGGML_CPU_ALL_VARIANTS=ON \-DGGML_VULKAN=ON \- -DHF_UI_VERSION=${{ needs.get-version.outputs.ui_version }} \${{ env.CMAKE_ARGS }}cmake --build build --config Release -j $(nproc)@@ -340,7 +311,7 @@ jobs:key: release-${{ matrix.os }}-vulkanandroid-arm64:- needs: [check-release, get-version]+ needs: [check-release, ui-build]if: ${{ needs.check-release.outputs.should_release == 'true' }}runs-on: ubuntu-latest@@ -358,12 +329,11 @@ jobs:with:fetch-depth: 0- - name: Setup Node.js- uses: actions/setup-node@v6+ - name: Download UI build+ uses: actions/download-artifact@v7with:- node-version: "24"- cache: "npm"- cache-dependency-path: "tools/ui/package-lock.json"+ name: llama-ui.zip+ path: tools/ui/dist- name: Set up JDKuses: actions/setup-java@v5@@ -407,7 +377,6 @@ jobs:-DLLAMA_FATAL_WARNINGS=ON \-DGGML_OPENMP=OFF \-DLLAMA_BUILD_BORINGSSL=ON \- -DHF_UI_VERSION=${{ needs.get-version.outputs.ui_version }} \${{ env.CMAKE_ARGS }}cmake --build build --config Release -j $(nproc)@@ -433,7 +402,7 @@ jobs:name: llama-bin-android-arm64.tar.gzubuntu-24-openvino:- needs: [check-release, get-version]+ needs: [check-release, ui-build]if: ${{ needs.check-release.outputs.should_release == 'true' }}runs-on: ubuntu-24.04@@ -460,12 +429,11 @@ jobs:with:fetch-depth: 0- - name: Setup Node.js- uses: actions/setup-node@v6+ - name: Download UI build+ uses: actions/download-artifact@v7with:- node-version: "24"- cache: "npm"- cache-dependency-path: "tools/ui/package-lock.json"+ name: llama-ui.zip+ path: tools/ui/dist- name: ccacheuses: ggml-org/[email protected]@@ -508,7 +476,6 @@ jobs:-DGGML_OPENVINO=ON \-DCMAKE_INSTALL_RPATH='$ORIGIN' \-DCMAKE_BUILD_WITH_INSTALL_RPATH=ON \- -DHF_UI_VERSION=${{ needs.get-version.outputs.ui_version }} \${{ env.CMAKE_ARGS }}cmake --build build/ReleaseOV --config Release --parallel@@ -552,7 +519,7 @@ jobs:key: release-ubuntu-24.04-openvino-release-no-preset-v1windows-openvino:- needs: [check-release]+ needs: [check-release, ui-build]if: ${{ needs.check-release.outputs.should_release == 'true' }}runs-on: windows-2022@@ -577,12 +544,11 @@ jobs:with:fetch-depth: 0- - name: Setup Node.js- uses: actions/setup-node@v6+ - name: Download UI build+ uses: actions/download-artifact@v7with:- node-version: "24"- cache: "npm"- cache-dependency-path: "tools/ui/package-lock.json"+ name: llama-ui.zip+ path: tools/ui/dist- name: ccacheuses: ggml-org/[email protected]@@ -682,7 +648,7 @@ jobs:windows-cpu:name: windows-cpu / ${{ matrix.arch }}- needs: [check-release]+ needs: [check-release, ui-build]if: ${{ needs.check-release.outputs.should_release == 'true' }}runs-on: windows-2025-vs2026@@ -702,12 +668,11 @@ jobs:with:fetch-depth: 0- - name: Setup Node.js- uses: actions/setup-node@v6+ - name: Download UI build+ uses: actions/download-artifact@v7with:- node-version: "24"- cache: "npm"- cache-dependency-path: "tools/ui/package-lock.json"+ name: llama-ui.zip+ path: tools/ui/dist- name: Install Ninjarun: |@@ -749,8 +714,10 @@ jobs:with:key: release-windows-2025-vs2026-${{ matrix.arch }}-cpu+ # TODO: build only the ggml-hip backend like the other windows backend jobs+ # (windows-cuda, windows-sycl), then drop the ui-build dependencywindows-rocm:- needs: [check-release]+ needs: [check-release, ui-build]if: ${{ needs.check-release.outputs.should_release == 'true' }}runs-on: windows-2022@@ -769,6 +736,12 @@ jobs:with:fetch-depth: 0+ - name: Download UI build+ uses: actions/download-artifact@v7+ with:+ name: llama-ui.zip+ path: tools/ui/dist+- name: ccacheuses: ggml-org/[email protected]with:@@ -879,6 +852,8 @@ jobs:with:key: windows-rocm-${{ matrix.ROCM_VERSION }}-${{ matrix.build }}+ # note: builds only the backend library - llama-server (with the embedded UI)+ # is injected from the windows-cpu zip during the release "Merge artifacts" stepwindows:needs: [check-release]if: ${{ needs.check-release.outputs.should_release == 'true' }}@@ -909,13 +884,6 @@ jobs:id: checkoutuses: actions/checkout@v6- - name: Setup Node.js- uses: actions/setup-node@v6- with:- node-version: "24"- cache: "npm"- cache-dependency-path: "tools/ui/package-lock.json"-- name: Install Vulkan SDKid: get_vulkanif: ${{ matrix.backend == 'vulkan' }}@@ -978,6 +946,8 @@ jobs:path: llama-bin-win-${{ matrix.backend }}-${{ matrix.arch }}.zipname: llama-bin-win-${{ matrix.backend }}-${{ matrix.arch }}.zip+ # note: builds only the ggml-cuda backend - llama-server is injected from the+ # windows-cpu zip during the release "Merge artifacts" stepwindows-cuda:name: windows-cuda (${{ matrix.cuda }}, ${{ matrix.arch }})needs: [check-release]@@ -1006,13 +976,6 @@ jobs:id: checkoutuses: actions/checkout@v6- - name: Setup Node.js- uses: actions/setup-node@v6- with:- node-version: "24"- cache: "npm"- cache-dependency-path: "tools/ui/package-lock.json"-- name: Install Cuda Toolkituses: ./.github/actions/windows-setup-cudawith:@@ -1084,6 +1047,8 @@ jobs:with:key: release-windows-2022-${{ matrix.arch }}-cuda-${{ matrix.cuda }}+ # note: builds only the ggml-sycl backend - llama-server is injected from the+ # windows-cpu zip during the release "Merge artifacts" stepwindows-sycl:needs: [check-release]if: ${{ needs.check-release.outputs.should_release == 'true' }}@@ -1118,13 +1083,6 @@ jobs:Expand-Archive -Path "level-zero-win-sdk.zip" -DestinationPath "C:/level-zero-sdk" -Force"LEVEL_ZERO_V1_SDK_PATH=C:/level-zero-sdk" | Out-File -FilePath $env:GITHUB_ENV -Append- - name: Setup Node.js- uses: actions/setup-node@v6- with:- node-version: "24"- cache: "npm"- cache-dependency-path: "tools/ui/package-lock.json"-- name: ccacheuses: ggml-org/[email protected]with:@@ -1195,7 +1153,7 @@ jobs:key: release-windows-2022-x64-syclubuntu-24-sycl:- needs: [check-release]+ needs: [check-release, ui-build]if: ${{ needs.check-release.outputs.should_release == 'true' }}strategy:@@ -1237,12 +1195,11 @@ jobs:wget -q "https://github.com/oneapi-src/level-zero/releases/download/v${LEVEL_ZERO_VERSION}/level-zero-devel_${LEVEL_ZERO_VERSION}%2B${LEVEL_ZERO_UBUNTU_VERSION}_amd64.deb" -O level-zero-devel.debsudo apt-get install -y ./level-zero.deb ./level-zero-devel.deb- - name: Setup Node.js- uses: actions/setup-node@v6+ - name: Download UI build+ uses: actions/download-artifact@v7with:- node-version: "24"- cache: "npm"- cache-dependency-path: "tools/ui/package-lock.json"+ name: llama-ui.zip+ path: tools/ui/dist- name: ccacheuses: ggml-org/[email protected]@@ -1288,7 +1245,7 @@ jobs:key: release-ubuntu-24.04-sycl-${{ matrix.build }}ubuntu-24-rocm:- needs: [check-release, get-version]+ needs: [check-release, ui-build]if: ${{ needs.check-release.outputs.should_release == 'true' }}runs-on: ubuntu-24.04@@ -1310,12 +1267,11 @@ jobs:with:fetch-depth: 0- - name: Setup Node.js- uses: actions/setup-node@v6+ - name: Download UI build+ uses: actions/download-artifact@v7with:- node-version: "24"- cache: "npm"- cache-dependency-path: "tools/ui/package-lock.json"+ name: llama-ui.zip+ path: tools/ui/dist- name: Free up disk spaceuses: ggml-org/[email protected]@@ -1388,7 +1344,6 @@ jobs:-DGPU_TARGETS="${{ matrix.gpu_targets }}" \-DGGML_HIP=ON \-DHIP_PLATFORM=amd \- -DHF_UI_VERSION=${{ needs.get-version.outputs.ui_version }} \${{ env.CMAKE_ARGS }}cmake --build build --config Release -j $(nproc)@@ -1417,7 +1372,7 @@ jobs:key: release-ubuntu-24.04-rocm-${{ matrix.ROCM_VERSION }}-${{ matrix.build }}ios-xcode:- needs: [check-release, get-version]+ needs: [check-release]if: ${{ needs.check-release.outputs.should_release == 'true' }}runs-on: macos-26@@ -1445,8 +1400,7 @@ jobs:-DLLAMA_BUILD_SERVER=OFF \-DCMAKE_SYSTEM_NAME=iOS \-DCMAKE_OSX_DEPLOYMENT_TARGET=16.0 \- -DCMAKE_XCODE_ATTRIBUTE_DEVELOPMENT_TEAM=ggml \- -DHF_UI_VERSION=${{ needs.get-version.outputs.ui_version }}+ -DCMAKE_XCODE_ATTRIBUTE_DEVELOPMENT_TEAM=ggmlcmake --build build --config Release -j $(sysctl -n hw.logicalcpu) -- CODE_SIGNING_ALLOWED=NO- name: xcodebuild for swift package@@ -1569,11 +1523,9 @@ jobs:# name: llama-bin-${{ matrix.chip_type }}-openEuler-${{ matrix.arch }}${{ matrix.use_acl_graph == 'on' && '-aclgraph' || '' }}.tar.gzui-build:- needs: [check-release, get-version]+ needs: [check-release]if: ${{ needs.check-release.outputs.should_release == 'true' }}uses: ./.github/workflows/ui-build.yml- with:- hf_ui_version: ${{ needs.get-version.outputs.ui_version }}release:if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}@@ -1588,7 +1540,6 @@ jobs:runs-on: ubuntu-slimneeds:- - get-version- windows- windows-cpu- windows-cuda@@ -1628,24 +1579,27 @@ jobs:path: ./artifactmerge-multiple: true- - name: Move artifacts+ - name: Merge artifactsid: move_artifactsrun: |mkdir -p release- echo "Adding CPU backend files to existing zips..."+ # the windows-cpu zip contains the full toolset (llama-server with the embedded+ # UI, ggml-cpu) - inject it into the other windows zips so that every archive+ # ships the same binaries, only with a different backend library on top+ echo "Injecting windows-cpu binaries (llama-server + CPU backend) into the backend zips..."for arch in x64 arm64; docpu_zip="artifact/llama-bin-win-cpu-${arch}.zip"temp_dir=$(mktemp -d)- echo "Extracting CPU backend for $arch..."+ echo "Extracting windows-cpu-${arch} package..."unzip "$cpu_zip" -d "$temp_dir"- echo "Adding CPU files to $arch zips..."+ echo "Merging into $arch zips..."for target_zip in artifact/llama-bin-win-*-${arch}.zip; doif [[ "$target_zip" == "$cpu_zip" ]]; thencontinuefi- echo "Adding CPU backend to $(basename "$target_zip")"+ echo "Injecting into $(basename "$target_zip")"realpath_target_zip=$(realpath "$target_zip")(cd "$temp_dir" && zip -r "$realpath_target_zip" .)done@@ -1669,7 +1623,7 @@ jobs:id: download_uiuses: actions/download-artifact@v7with:- name: ui-build+ name: llama-ui.zippath: ./ui-dist- name: Package UIdiff --git a/.github/workflows/server-sanitize.yml b/.github/workflows/server-sanitize.ymlindex 5d696282c7..a0ffc70c12 100644--- a/.github/workflows/server-sanitize.yml+++ b/.github/workflows/server-sanitize.yml@@ -73,13 +73,6 @@ jobs:fetch-depth: 0ref: ${{ github.event.inputs.sha || github.event.pull_request.head.sha || github.sha || github.head_ref || github.ref_name }}- - name: Setup Node.js- uses: actions/setup-node@v6- with:- node-version: "24"- cache: "npm"- cache-dependency-path: "tools/ui/package-lock.json"-- name: Buildid: cmake_buildrun: |diff --git a/.github/workflows/ui-build-self-hosted.yml b/.github/workflows/ui-build-self-hosted.ymlindex 7b7f8b6002..390a2f35f7 100644--- a/.github/workflows/ui-build-self-hosted.yml+++ b/.github/workflows/ui-build-self-hosted.yml@@ -31,6 +31,6 @@ jobs:- name: Upload built UIuses: actions/upload-artifact@v6with:- name: ui-build+ name: llama-ui.zippath: tools/ui/dist/retention-days: 1diff --git a/.github/workflows/ui-build.yml b/.github/workflows/ui-build.ymlindex 85642f3f4b..3fbd90c11c 100644--- a/.github/workflows/ui-build.yml+++ b/.github/workflows/ui-build.yml@@ -3,8 +3,8 @@ name: UI Buildon:workflow_call:inputs:- hf_ui_version:- description: 'Version string for version.json (e.g. 12345)'+ ui_version:+ description: 'Version string embedded in build.json (e.g. b1234); defaults to b<commit-count>'required: falsetype: string@@ -17,6 +17,17 @@ jobs:steps:- name: Checkout codeuses: actions/checkout@v6+ with:+ fetch-depth: 0++ - name: Resolve UI version+ id: version+ run: |+ version="${{ inputs.ui_version }}"+ if [ -z "$version" ]; then+ version="b$(git rev-list --count HEAD)"+ fi+ echo "ui_version=${version}" >> $GITHUB_OUTPUT- name: Setup Node.jsuses: actions/setup-node@v6@@ -31,8 +42,7 @@ jobs:- name: Build applicationenv:- HF_UI_VERSION: ${{ inputs.hf_ui_version || '' }}- LLAMA_BUILD_NUMBER: ${{ inputs.hf_ui_version || 'b0000' }}+ LLAMA_BUILD_NUMBER: ${{ steps.version.outputs.ui_version }}run: npm run buildworking-directory: tools/ui@@ -43,6 +53,6 @@ jobs:- name: Upload built UIuses: actions/upload-artifact@v6with:- name: ui-build+ name: llama-ui.zippath: tools/ui/dist/retention-days: 1diff --git a/.github/workflows/ui-publish.yml b/.github/workflows/ui-publish.ymlindex 99a6d8420f..e64ef32f80 100644--- a/.github/workflows/ui-publish.yml+++ b/.github/workflows/ui-publish.yml@@ -37,7 +37,7 @@ jobs:- name: Download UI build artifactuses: actions/download-artifact@v7with:- name: ui-build+ name: llama-ui.zippath: tools/ui/dist/- name: Create distribution archivediff --git a/.github/workflows/ui-self-hosted.yml b/.github/workflows/ui-self-hosted.ymlindex 79d7800d6b..63521ead2d 100644--- a/.github/workflows/ui-self-hosted.yml+++ b/.github/workflows/ui-self-hosted.yml@@ -64,7 +64,7 @@ jobs:- name: Download built UI artifactsuses: actions/download-artifact@v6with:- name: ui-build+ name: llama-ui.zippath: tools/ui/dist/- name: Run type checking@@ -106,7 +106,7 @@ jobs:- name: Download built UI artifactsuses: actions/download-artifact@v6with:- name: ui-build+ name: llama-ui.zippath: tools/ui/dist/- name: Build Storybookdiff --git a/.github/workflows/ui.yml b/.github/workflows/ui.ymlindex fa99a0cda2..00a0804af3 100644--- a/.github/workflows/ui.yml+++ b/.github/workflows/ui.yml@@ -63,7 +63,7 @@ jobs:- name: Download built UI artifactsuses: actions/download-artifact@v6with:- name: ui-build+ name: llama-ui.zippath: tools/ui/dist/- name: Install dependencies@@ -126,7 +126,7 @@ jobs:- name: Download built UI artifacts (reuses ui-build)uses: actions/download-artifact@v6with:- name: ui-build+ name: llama-ui.zippath: tools/ui/dist/- name: Install Playwright browsersdiff --git a/CMakeLists.txt b/CMakeLists.txtindex 730d5561fd..1d4bcf45f5 100644--- a/CMakeLists.txt+++ b/CMakeLists.txt@@ -134,8 +134,8 @@ option(LLAMA_BUILD_TOOLS "llama: build tools"option(LLAMA_BUILD_EXAMPLES "llama: build examples" ${LLAMA_STANDALONE})option(LLAMA_BUILD_SERVER "llama: build server example" ${LLAMA_STANDALONE})option(LLAMA_BUILD_APP "llama: build the unified binary" ${LLAMA_STANDALONE})-option(LLAMA_BUILD_UI "llama: build the embedded Web UI for server" ON)-option(LLAMA_USE_PREBUILT_UI "llama: use prebuilt UI from HF Bucket when available (requires LLAMA_BUILD_UI=ON)" ON)+option(LLAMA_BUILD_UI "llama: build the embedded Web UI for server" OFF)+option(LLAMA_USE_PREBUILT_UI "llama: use prebuilt UI from HF Bucket when available" ON)option(LLAMA_TOOLS_INSTALL "llama: install tools" ${LLAMA_TOOLS_INSTALL_DEFAULT})option(LLAMA_TESTS_INSTALL "llama: install tests" ON)