.github,docker/unified: include vulkan build (#599)

Update docker/unified scripts to support building both cuda and vulkan unified images.

.github/workflows/unified-docker.yml

@@ -30,6 +30,10 @@ permissions:
 jobs:
   build:
     runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        backend: [cuda, vulkan]
     steps:
       - name: Checkout code
         uses: actions/checkout@v4
@@ -61,24 +65,25 @@ jobs:
           username: ${{ github.actor }}
           password: ${{ secrets.GITHUB_TOKEN }}
 
-      - name: Build unified Docker image
+      - name: Build unified Docker image (${{ matrix.backend }})
         env:
           LLAMA_REF: ${{ inputs.llama_cpp_ref || 'master' }}
           WHISPER_REF: ${{ inputs.whisper_ref || 'master' }}
           SD_REF: ${{ inputs.sd_ref || 'master' }}
           LS_VERSION: ${{ inputs.llama_swap_version || 'main' }}
-          DOCKER_IMAGE_TAG: ghcr.io/mostlygeek/llama-swap:unified
+          DOCKER_IMAGE_TAG: ghcr.io/mostlygeek/llama-swap:unified-${{ matrix.backend }}
           # When running under act, use the local builder that has warm ccache.
           # On GitHub Actions, BUILDX_BUILDER is unset so docker uses the builder
           # created by setup-buildx-action above.
           BUILDX_BUILDER: ${{ env.ACT == 'true' && 'llama-swap-builder' || '' }}
         run: |
           chmod +x docker/unified/build-image.sh
-          docker/unified/build-image.sh
+          docker/unified/build-image.sh --${{ matrix.backend }}
 
       - name: Push to GitHub Container Registry
         if: ${{ !env.ACT }}
         run: |
-          docker push ghcr.io/mostlygeek/llama-swap:unified
+          docker push ghcr.io/mostlygeek/llama-swap:unified-${{ matrix.backend }}
-          docker tag ghcr.io/mostlygeek/llama-swap:unified ghcr.io/mostlygeek/llama-swap:unified-$(date -u +%Y-%m-%d)
+          DATE_TAG=$(date -u +%Y-%m-%d)
-          docker push ghcr.io/mostlygeek/llama-swap:unified-$(date -u +%Y-%m-%d)
+          docker tag ghcr.io/mostlygeek/llama-swap:unified-${{ matrix.backend }} ghcr.io/mostlygeek/llama-swap:unified-${{ matrix.backend }}-${DATE_TAG}
+          docker push ghcr.io/mostlygeek/llama-swap:unified-${{ matrix.backend }}-${DATE_TAG}

docker/unified/Dockerfile

@@ -1,14 +1,18 @@
-# Unified multi-stage Dockerfile for CUDA-accelerated AI inference tools
+# Unified multi-stage Dockerfile for AI inference tools
-# Includes: llama.cpp, whisper.cpp, stable-diffusion.cpp, llama-swap
+# Supports CUDA and Vulkan backends via BACKEND build arg
 #
 # Usage:
-#   docker buildx build -t llama-swap:unified .
+#   docker buildx build --build-arg BACKEND=cuda -t llama-swap:unified-cuda .
+#   docker buildx build --build-arg BACKEND=vulkan -t llama-swap:unified-vulkan .
 #
 # Each project has its own install script that handles cloning, building,
 # and installing binaries. Build stages are independent for cache efficiency.
 
-# Builder base: CUDA devel image with build tools
+ARG BACKEND=cuda
-FROM nvidia/cuda:12.4.0-devel-ubuntu22.04 AS builder-base
+
+# ── Builder bases ──────────────────────────────────────────────────────
+
+FROM nvidia/cuda:12.4.0-devel-ubuntu22.04 AS builder-base-cuda
 
 ENV DEBIAN_FRONTEND=noninteractive
 ENV CMAKE_CUDA_ARCHITECTURES="60;61;75;86;89"
@@ -23,42 +27,72 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
 
 WORKDIR /build
 
-# Build whisper.cpp (fastest build, run first)
+# ──
+
+FROM ubuntu:22.04 AS builder-base-vulkan
+
+ENV DEBIAN_FRONTEND=noninteractive
+ENV CCACHE_DIR=/ccache
+ENV CCACHE_MAXSIZE=2G
+ENV PATH="/usr/lib/ccache:${PATH}"
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    build-essential cmake git python3 python3-pip libssl-dev \
+    curl ca-certificates ccache make wget software-properties-common \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install LunarG Vulkan SDK (Ubuntu 22.04 repo headers are too old)
+RUN wget -qO /etc/apt/trusted.gpg.d/lunarg.asc https://packages.lunarg.com/lunarg-signing-key-pub.asc \
+    && echo "deb https://packages.lunarg.com/vulkan jammy main" > /etc/apt/sources.list.d/lunarg-vulkan.list \
+    && apt-get update && apt-get install -y --no-install-recommends vulkan-sdk \
+    && rm -rf /var/lib/apt/lists/*
+
+WORKDIR /build
+
+# ── Select builder base by BACKEND ────────────────────────────────────
+
+FROM builder-base-${BACKEND} AS builder-base
+
+# ── Build whisper.cpp (fastest build, run first) ──────────────────────
+
 FROM builder-base AS whisper-build
+ARG BACKEND=cuda
 ARG WHISPER_COMMIT_HASH=master
 COPY install-whisper.sh /build/
-RUN --mount=type=cache,id=ccache-cuda,target=/ccache \
+RUN --mount=type=cache,id=ccache-${BACKEND},target=/ccache \
-    --mount=type=cache,id=whisper-cuda,target=/src/whisper.cpp/build \
+    --mount=type=cache,id=whisper-${BACKEND},target=/src/whisper.cpp/build \
-    bash /build/install-whisper.sh "${WHISPER_COMMIT_HASH}"
+    BACKEND=${BACKEND} bash /build/install-whisper.sh "${WHISPER_COMMIT_HASH}"
+
+# ── Build stable-diffusion.cpp ────────────────────────────────────────
 
-# Build stable-diffusion.cpp
 FROM builder-base AS sd-build
+ARG BACKEND=cuda
 ARG SD_COMMIT_HASH=master
 COPY install-sd.sh /build/
-RUN --mount=type=cache,id=ccache-cuda,target=/ccache \
+RUN --mount=type=cache,id=ccache-${BACKEND},target=/ccache \
-    --mount=type=cache,id=sd-cuda,target=/src/stable-diffusion.cpp/build \
+    --mount=type=cache,id=sd-${BACKEND},target=/src/stable-diffusion.cpp/build \
-    bash /build/install-sd.sh "${SD_COMMIT_HASH}"
+    BACKEND=${BACKEND} bash /build/install-sd.sh "${SD_COMMIT_HASH}"
+
+# ── Build llama.cpp (slowest build, run last) ─────────────────────────
 
-# Build llama.cpp (slowest build, run last)
 FROM builder-base AS llama-build
+ARG BACKEND=cuda
 ARG LLAMA_COMMIT_HASH=master
 COPY install-llama.sh /build/
-RUN --mount=type=cache,id=ccache-cuda,target=/ccache \
+RUN --mount=type=cache,id=ccache-${BACKEND},target=/ccache \
-    --mount=type=cache,id=llama-cuda,target=/src/llama.cpp/build \
+    --mount=type=cache,id=llama-${BACKEND},target=/src/llama.cpp/build \
-    bash /build/install-llama.sh "${LLAMA_COMMIT_HASH}"
+    BACKEND=${BACKEND} bash /build/install-llama.sh "${LLAMA_COMMIT_HASH}"
+
+# ── Download llama-swap release binary ────────────────────────────────
 
-# Download llama-swap release binary
 FROM builder-base AS llama-swap-download
 ARG LS_VERSION=latest
 COPY install-llama-swap.sh /build/
 RUN bash /build/install-llama-swap.sh "${LS_VERSION}"
 
-# Runtime image (no build tooling)
+# ── Runtime bases ─────────────────────────────────────────────────────
-FROM nvidia/cuda:12.4.0-runtime-ubuntu22.04 AS runtime
 
-ARG LLAMA_COMMIT_HASH=unknown
+FROM nvidia/cuda:12.4.0-runtime-ubuntu22.04 AS runtime-cuda
-ARG WHISPER_COMMIT_HASH=unknown
-ARG SD_COMMIT_HASH=unknown
 
 ENV DEBIAN_FRONTEND=noninteractive
 ENV LD_LIBRARY_PATH="/usr/local/cuda/lib64:${LD_LIBRARY_PATH}"
@@ -69,8 +103,29 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
     && rm -rf /var/lib/apt/lists/*
 
 # CUDA stub drivers for container compatibility
-COPY --from=builder-base /usr/local/cuda/lib64/stubs/libcuda.so /usr/local/cuda/lib64/stubs/libcuda.so
+COPY --from=builder-base-cuda /usr/local/cuda/lib64/stubs/libcuda.so /usr/local/cuda/lib64/stubs/libcuda.so
-COPY --from=builder-base /usr/local/cuda/lib64/stubs/libcuda.so /usr/local/cuda/lib64/stubs/libcuda.so.1
+COPY --from=builder-base-cuda /usr/local/cuda/lib64/stubs/libcuda.so /usr/local/cuda/lib64/stubs/libcuda.so.1
+
+# ──
+
+FROM ubuntu:22.04 AS runtime-vulkan
+
+ENV DEBIAN_FRONTEND=noninteractive
+ENV PATH="/usr/local/bin:${PATH}"
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    libgomp1 libvulkan1 mesa-vulkan-drivers \
+    python3 python3-pip curl ca-certificates git \
+    && rm -rf /var/lib/apt/lists/*
+
+# ── Select runtime base by BACKEND ────────────────────────────────────
+
+FROM runtime-${BACKEND} AS runtime
+
+ARG BACKEND=cuda
+ARG LLAMA_COMMIT_HASH=unknown
+ARG WHISPER_COMMIT_HASH=unknown
+ARG SD_COMMIT_HASH=unknown
 
 RUN pip3 install --no-cache-dir numpy sentencepiece
 
@@ -109,7 +164,7 @@ RUN echo "llama.cpp: ${LLAMA_COMMIT_HASH}" > /versions.txt && \
     echo "whisper.cpp: ${WHISPER_COMMIT_HASH}" >> /versions.txt && \
     echo "stable-diffusion.cpp: ${SD_COMMIT_HASH}" >> /versions.txt && \
     echo "llama-swap: $(cat /tmp/llama-swap-version)" >> /versions.txt && \
-    echo "backend: cuda" >> /versions.txt && \
+    echo "backend: ${BACKEND}" >> /versions.txt && \
     echo "build_timestamp: $(date -u +%Y-%m-%dT%H:%M:%SZ)" >> /versions.txt
 
 WORKDIR /models

docker/unified/build-image.sh

@@ -1,32 +1,45 @@
 #!/bin/bash
 #
-# Build script for unified CUDA container with version pinning
+# Build script for unified container with version pinning
 #
 # Usage:
-#   ./build-image.sh                               # Build with auto-detected versions
+#   ./build-image.sh --cuda                              # Build CUDA image
-#   ./build-image.sh --no-cache                    # Build without cache
+#   ./build-image.sh --vulkan                            # Build Vulkan image
-#   LLAMA_REF=b1234 ./build-image.sh               # Pin llama.cpp to a commit hash
+#   ./build-image.sh --cuda --no-cache                   # Build without cache
-#   LLAMA_REF=v1.2.3 ./build-image.sh              # Pin llama.cpp to a tag
+#   LLAMA_REF=b1234 ./build-image.sh --vulkan            # Pin llama.cpp to a commit hash
-#   LLAMA_REF=my-branch ./build-image.sh           # Pin llama.cpp to a branch
+#   LLAMA_REF=v1.2.3 ./build-image.sh --cuda             # Pin llama.cpp to a tag
-#   WHISPER_REF=v1.0.0 ./build-image.sh            # Pin whisper.cpp to a tag
+#   WHISPER_REF=v1.0.0 ./build-image.sh --vulkan         # Pin whisper.cpp to a tag
-#   SD_REF=master ./build-image.sh                 # Pin stable-diffusion.cpp to a branch
+#   SD_REF=master ./build-image.sh --cuda                # Pin stable-diffusion.cpp to a branch
-#   LS_VERSION=170 ./build-image.sh                # Override llama-swap version
+#   LS_VERSION=170 ./build-image.sh --cuda               # Override llama-swap version
 #
 
 set -euo pipefail
 
+BACKEND=""
 NO_CACHE=false
 
 for arg in "$@"; do
     case $arg in
+        --cuda)
+            BACKEND="cuda"
+            ;;
+        --vulkan)
+            BACKEND="vulkan"
+            ;;
         --no-cache)
             NO_CACHE=true
             ;;
         --help|-h)
-            echo "Usage: ./build-image.sh [--no-cache]"
+            echo "Usage: ./build-image.sh --cuda|--vulkan [--no-cache]"
+            echo ""
+            echo "Options:"
+            echo "  --cuda      Build CUDA image (NVIDIA GPUs)"
+            echo "  --vulkan    Build Vulkan image (AMD GPUs and compatible hardware)"
+            echo "  --no-cache  Force rebuild without using Docker cache"
+            echo "  --help, -h  Show this help message"
             echo ""
             echo "Environment variables:"
-            echo "  DOCKER_IMAGE_TAG     Set custom image tag (default: llama-swap:unified)"
+            echo "  DOCKER_IMAGE_TAG     Set custom image tag (default: llama-swap:unified-cuda or llama-swap:unified-vulkan)"
             echo "  LLAMA_REF            Pin llama.cpp to a commit, tag, or branch"
             echo "  WHISPER_REF          Pin whisper.cpp to a commit, tag, or branch"
             echo "  SD_REF               Pin stable-diffusion.cpp to a commit, tag, or branch"
@@ -36,7 +49,14 @@ for arg in "$@"; do
     esac
 done
 
-DOCKER_IMAGE_TAG="${DOCKER_IMAGE_TAG:-llama-swap:unified}"
+if [[ -z "$BACKEND" ]]; then
+    echo "Error: No backend specified. Please use --cuda or --vulkan."
+    echo ""
+    echo "Usage: ./build-image.sh --cuda|--vulkan [--no-cache]"
+    exit 1
+fi
+
+DOCKER_IMAGE_TAG="${DOCKER_IMAGE_TAG:-llama-swap:unified-${BACKEND}}"
 
 # Git repository URLs
 LLAMA_REPO="https://github.com/ggml-org/llama.cpp.git"
@@ -89,7 +109,7 @@ get_latest_hash() {
 }
 
 echo "=========================================="
-echo "llama-swap Unified CUDA Build"
+echo "llama-swap Unified Build (${BACKEND})"
 echo "=========================================="
 echo ""
 
@@ -154,6 +174,7 @@ echo ""
 SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
 
 BUILD_ARGS=(
+    --build-arg "BACKEND=${BACKEND}"
     --build-arg "LLAMA_COMMIT_HASH=${LLAMA_HASH}"
     --build-arg "WHISPER_COMMIT_HASH=${WHISPER_HASH}"
     --build-arg "SD_COMMIT_HASH=${SD_HASH}"
@@ -166,7 +187,7 @@ if [[ "$NO_CACHE" == true ]]; then
     BUILD_ARGS+=(--no-cache)
     echo "Note: Building without cache"
 elif [[ "${GITHUB_ACTIONS:-}" == "true" && "${ACT:-}" != "true" ]]; then
-    CACHE_REF="ghcr.io/mostlygeek/llama-swap:unified-cache"
+    CACHE_REF="ghcr.io/mostlygeek/llama-swap:unified-${BACKEND}-cache"
     BUILD_ARGS+=(
         --cache-from "type=registry,ref=${CACHE_REF}"
         --cache-to "type=registry,ref=${CACHE_REF},mode=max"
@@ -196,7 +217,7 @@ if [[ ${#MISSING_BINARIES[@]} -gt 0 ]]; then
     done
     echo ""
     echo "Try running with --no-cache flag:"
-    echo "  ./build-image.sh --no-cache"
+    echo "  ./build-image.sh --${BACKEND} --no-cache"
     exit 1
 fi
 
@@ -215,5 +236,13 @@ echo "  whisper.cpp:         ${WHISPER_HASH}"
 echo "  stable-diffusion.cpp: ${SD_HASH}"
 echo "  llama-swap:          $(docker run --rm --entrypoint cat "${DOCKER_IMAGE_TAG}" /versions.txt | grep llama-swap | cut -d' ' -f2-)"
 echo ""
-echo "Run with:"
+if [[ "$BACKEND" == "vulkan" ]]; then
-echo "  docker run -it --rm --gpus all ${DOCKER_IMAGE_TAG}"
+    echo "Run with:"
+    echo "  docker run -it --rm --device /dev/dri:/dev/dri ${DOCKER_IMAGE_TAG}"
+    echo ""
+    echo "Note: For AMD GPUs, you may also need:"
+    echo "  docker run -it --rm --device /dev/dri:/dev/dri --group-add video ${DOCKER_IMAGE_TAG}"
+else
+    echo "Run with:"
+    echo "  docker run -it --rm --gpus all ${DOCKER_IMAGE_TAG}"
+fi

docker/unified/install-llama.sh

@@ -1,9 +1,10 @@
 #!/bin/bash
-# Install llama.cpp - clone, build with CUDA, and install binaries
+# Install llama.cpp - clone, build, and install binaries
-# Usage: ./install-llama.sh <commit_hash>
+# Usage: BACKEND=cuda|vulkan ./install-llama.sh <commit_hash>
 set -e
 
 COMMIT_HASH="${1:-master}"
+BACKEND="${BACKEND:-cuda}"
 
 mkdir -p /install/bin /install/lib
 
@@ -18,26 +19,36 @@ fi
 git fetch --depth=1 origin "${COMMIT_HASH}"
 git checkout FETCH_HEAD
 
-# CUDA cmake flags + llama-specific flags
+# Common cmake flags
 CMAKE_FLAGS=(
     -DGGML_NATIVE=OFF
     -DCMAKE_BUILD_TYPE=Release
-    -DGGML_CUDA=ON
-    -DGGML_VULKAN=OFF
-    "-DCMAKE_CUDA_ARCHITECTURES=${CMAKE_CUDA_ARCHITECTURES:-60;61;75;86;89}"
-    "-DCMAKE_CUDA_FLAGS=-allow-unsupported-compiler"
-    "-DCMAKE_EXE_LINKER_FLAGS=-Wl,-rpath-link,/usr/local/cuda/lib64/stubs -lcuda"
-    "-DCMAKE_SHARED_LINKER_FLAGS=-Wl,-rpath-link,/usr/local/cuda/lib64/stubs -lcuda"
     -DCMAKE_C_COMPILER_LAUNCHER=ccache
     -DCMAKE_CXX_COMPILER_LAUNCHER=ccache
     -DLLAMA_BUILD_TESTS=OFF
 )
 
+if [ "$BACKEND" = "cuda" ]; then
+    CMAKE_FLAGS+=(
+        -DGGML_CUDA=ON
+        -DGGML_VULKAN=OFF
+        "-DCMAKE_CUDA_ARCHITECTURES=${CMAKE_CUDA_ARCHITECTURES:-60;61;75;86;89}"
+        "-DCMAKE_CUDA_FLAGS=-allow-unsupported-compiler"
+        "-DCMAKE_EXE_LINKER_FLAGS=-Wl,-rpath-link,/usr/local/cuda/lib64/stubs -lcuda"
+        "-DCMAKE_SHARED_LINKER_FLAGS=-Wl,-rpath-link,/usr/local/cuda/lib64/stubs -lcuda"
+    )
+elif [ "$BACKEND" = "vulkan" ]; then
+    CMAKE_FLAGS+=(
+        -DGGML_CUDA=OFF
+        -DGGML_VULKAN=ON
+    )
+fi
+
 TARGETS=(llama-cli llama-server)
 
 rm -rf build/CMakeCache.txt build/CMakeFiles 2>/dev/null || true
 
-echo "=== Building llama.cpp for CUDA ==="
+echo "=== Building llama.cpp for ${BACKEND} ==="
 cmake -B build "${CMAKE_FLAGS[@]}"
 cmake --build build --config Release -j"$(nproc)" --target "${TARGETS[@]}"
 

docker/unified/install-sd.sh

@@ -1,9 +1,10 @@
 #!/bin/bash
-# Install stable-diffusion.cpp - clone and build with CUDA, install binaries and library
+# Install stable-diffusion.cpp - clone, build, and install binaries and library
-# Usage: ./install-sd.sh <commit_hash>
+# Usage: BACKEND=cuda|vulkan ./install-sd.sh <commit_hash>
 set -e
 
 COMMIT_HASH="${1:-master}"
+BACKEND="${BACKEND:-cuda}"
 
 mkdir -p /install/bin /install/lib
 
@@ -19,27 +20,37 @@ git fetch --depth=1 origin "${COMMIT_HASH}"
 git checkout FETCH_HEAD
 git submodule update --init --recursive --depth=1
 
-# CUDA cmake flags + sd-specific flags
+# Common cmake flags
 CMAKE_FLAGS=(
     -DGGML_NATIVE=OFF
     -DCMAKE_BUILD_TYPE=Release
-    -DGGML_CUDA=ON
-    -DGGML_VULKAN=OFF
-    "-DCMAKE_CUDA_ARCHITECTURES=${CMAKE_CUDA_ARCHITECTURES:-60;61;75;86;89}"
-    "-DCMAKE_CUDA_FLAGS=-allow-unsupported-compiler"
-    "-DCMAKE_EXE_LINKER_FLAGS=-Wl,-rpath-link,/usr/local/cuda/lib64/stubs -lcuda"
-    "-DCMAKE_SHARED_LINKER_FLAGS=-Wl,-rpath-link,/usr/local/cuda/lib64/stubs -lcuda"
     -DCMAKE_C_COMPILER_LAUNCHER=ccache
     -DCMAKE_CXX_COMPILER_LAUNCHER=ccache
     -DSD_BUILD_EXAMPLES=ON
-    -DSD_CUDA=ON
 )
 
+if [ "$BACKEND" = "cuda" ]; then
+    CMAKE_FLAGS+=(
+        -DGGML_CUDA=ON
+        -DGGML_VULKAN=OFF
+        "-DCMAKE_CUDA_ARCHITECTURES=${CMAKE_CUDA_ARCHITECTURES:-60;61;75;86;89}"
+        "-DCMAKE_CUDA_FLAGS=-allow-unsupported-compiler"
+        "-DCMAKE_EXE_LINKER_FLAGS=-Wl,-rpath-link,/usr/local/cuda/lib64/stubs -lcuda"
+        "-DCMAKE_SHARED_LINKER_FLAGS=-Wl,-rpath-link,/usr/local/cuda/lib64/stubs -lcuda"
+        -DSD_CUDA=ON
+    )
+elif [ "$BACKEND" = "vulkan" ]; then
+    CMAKE_FLAGS+=(
+        -DGGML_CUDA=OFF
+        -DGGML_VULKAN=ON
+    )
+fi
+
 TARGETS=(stable-diffusion sd-cli sd-server)
 
 rm -rf build/CMakeCache.txt build/CMakeFiles 2>/dev/null || true
 
-echo "=== Building stable-diffusion.cpp for CUDA ==="
+echo "=== Building stable-diffusion.cpp for ${BACKEND} ==="
 cmake -B build "${CMAKE_FLAGS[@]}"
 cmake --build build --config Release -j"$(nproc)" --target "${TARGETS[@]}"
 

docker/unified/install-whisper.sh

@@ -1,9 +1,10 @@
 #!/bin/bash
-# Install whisper.cpp - clone, build with CUDA, and install binaries
+# Install whisper.cpp - clone, build, and install binaries
-# Usage: ./install-whisper.sh <commit_hash>
+# Usage: BACKEND=cuda|vulkan ./install-whisper.sh <commit_hash>
 set -e
 
 COMMIT_HASH="${1:-master}"
+BACKEND="${BACKEND:-cuda}"
 
 mkdir -p /install/bin /install/lib
 
@@ -18,25 +19,35 @@ fi
 git fetch --depth=1 origin "${COMMIT_HASH}"
 git checkout FETCH_HEAD
 
-# CUDA cmake flags
+# Common cmake flags
 CMAKE_FLAGS=(
     -DGGML_NATIVE=OFF
     -DCMAKE_BUILD_TYPE=Release
-    -DGGML_CUDA=ON
-    -DGGML_VULKAN=OFF
-    "-DCMAKE_CUDA_ARCHITECTURES=${CMAKE_CUDA_ARCHITECTURES:-60;61;75;86;89}"
-    "-DCMAKE_CUDA_FLAGS=-allow-unsupported-compiler"
-    "-DCMAKE_EXE_LINKER_FLAGS=-Wl,-rpath-link,/usr/local/cuda/lib64/stubs -lcuda"
-    "-DCMAKE_SHARED_LINKER_FLAGS=-Wl,-rpath-link,/usr/local/cuda/lib64/stubs -lcuda"
     -DCMAKE_C_COMPILER_LAUNCHER=ccache
     -DCMAKE_CXX_COMPILER_LAUNCHER=ccache
 )
 
+if [ "$BACKEND" = "cuda" ]; then
+    CMAKE_FLAGS+=(
+        -DGGML_CUDA=ON
+        -DGGML_VULKAN=OFF
+        "-DCMAKE_CUDA_ARCHITECTURES=${CMAKE_CUDA_ARCHITECTURES:-60;61;75;86;89}"
+        "-DCMAKE_CUDA_FLAGS=-allow-unsupported-compiler"
+        "-DCMAKE_EXE_LINKER_FLAGS=-Wl,-rpath-link,/usr/local/cuda/lib64/stubs -lcuda"
+        "-DCMAKE_SHARED_LINKER_FLAGS=-Wl,-rpath-link,/usr/local/cuda/lib64/stubs -lcuda"
+    )
+elif [ "$BACKEND" = "vulkan" ]; then
+    CMAKE_FLAGS+=(
+        -DGGML_CUDA=OFF
+        -DGGML_VULKAN=ON
+    )
+fi
+
 TARGETS=(whisper-cli whisper-server)
 
 rm -rf build/CMakeCache.txt build/CMakeFiles 2>/dev/null || true
 
-echo "=== Building whisper.cpp for CUDA ==="
+echo "=== Building whisper.cpp for ${BACKEND} ==="
 cmake -B build "${CMAKE_FLAGS[@]}"
 cmake --build build --config Release -j"$(nproc)" --target "${TARGETS[@]}"