Enable RPC support and build RPC server images

EZForever · Sep 18, 2024 · 8d1073b · 8d1073b
1 parent f3eeb5d
commit 8d1073b
Show file tree

Hide file tree

Showing 3 changed files with 37 additions and 11 deletions.
diff --git a/.devops/rpc-server.Dockerfile b/.devops/rpc-server.Dockerfile
@@ -0,0 +1,4 @@
+FROM scratch
+COPY --chown=0:0 --chmod=755 rpc-server /rpc-server
+ENV LC_ALL=C.utf8
+ENTRYPOINT ["/rpc-server"]
diff --git a/.github/workflows/build-oneapi.yml b/.github/workflows/build-oneapi.yml
@@ -49,18 +49,18 @@ jobs:
       - name: CMake
         run: |
           . /opt/intel/oneapi/setvars.sh
-          cmake repo -B build -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_COMPILER=icx -DCMAKE_CXX_COMPILER=icpx -DCMAKE_C_FLAGS="-Ofast -xCORE-AVX2" -DCMAKE_CXX_FLAGS="-Ofast -xCORE-AVX2" -DCMAKE_EXE_LINKER_FLAGS="-static" -DGGML_NATIVE=OFF -DGGML_STATIC=ON -DGGML_OPENMP=OFF -DLLAMA_BUILD_SERVER=ON -DBUILD_SHARED_LIBS=OFF
-          cmake --build build --config Release --target llama-server -j ${{ steps.cpu-cores.outputs.count }}
-          strip build/bin/llama-server
+          cmake repo -B build -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_COMPILER=icx -DCMAKE_CXX_COMPILER=icpx -DCMAKE_C_FLAGS="-Ofast -xCORE-AVX2" -DCMAKE_CXX_FLAGS="-Ofast -xCORE-AVX2" -DCMAKE_EXE_LINKER_FLAGS="-static" -DGGML_NATIVE=OFF -DGGML_STATIC=ON -DGGML_RPC=ON -DGGML_OPENMP=OFF -DLLAMA_BUILD_SERVER=ON -DBUILD_SHARED_LIBS=OFF
+          cmake --build build --config Release --target llama-server --target rpc-server -j ${{ steps.cpu-cores.outputs.count }}
+          strip build/bin/*
 
       - name: Upload build artifact
         uses: actions/upload-artifact@v4
         with:
           name: server-${{ matrix.ssl }}-oneapi
-          path: build/bin/llama-server
+          path: build/bin/*
           if-no-files-found: error
 
-      - name: Build and push Docker image
+      - name: Build and push llama-server Docker image
         uses: macbre/push-to-ghcr@v13
         with:
           dockerfile: .devops/llama-server.Dockerfile
@@ -71,3 +71,14 @@ jobs:
           docker_io_user: ${{ vars.DOCKER_IO_USER }}
           docker_io_token: ${{ secrets.DOCKER_IO_ACCESS_TOKEN }}
 
+      - name: Build and push rpc-server Docker image
+        uses: macbre/push-to-ghcr@v13
+        with:
+          dockerfile: .devops/rpc-server.Dockerfile
+          context: build/bin
+          image_name: ${{ github.repository }}
+          image_tag: rpc-server-${{ matrix.ssl }}-oneapi
+          github_token: ${{ secrets.GHCR_ACCESS_TOKEN }}
+          docker_io_user: ${{ vars.DOCKER_IO_USER }}
+          docker_io_token: ${{ secrets.DOCKER_IO_ACCESS_TOKEN }}
+
diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
@@ -58,20 +58,20 @@ jobs:
 
       - name: CMake
         run: |
-          cmake repo -B build -DCMAKE_BUILD_TYPE=Release -DCMAKE_EXE_LINKER_FLAGS="-static" -DGGML_NATIVE=OFF -DGGML_STATIC=ON -DGGML_OPENMP=OFF -DLLAMA_BUILD_SERVER=ON -DBUILD_SHARED_LIBS=OFF ${{ matrix.ssl_defines }} ${{ matrix.avx_defines }}
-          cmake --build build --config Release --target llama-server -j ${{ steps.cpu-cores.outputs.count }}
-          strip build/bin/llama-server
+          cmake repo -B build -DCMAKE_BUILD_TYPE=Release -DCMAKE_EXE_LINKER_FLAGS="-static" -DGGML_NATIVE=OFF -DGGML_STATIC=ON -DGGML_RPC=ON -DGGML_OPENMP=OFF -DLLAMA_BUILD_SERVER=ON -DBUILD_SHARED_LIBS=OFF ${{ matrix.ssl_defines }} ${{ matrix.avx_defines }}
+          cmake --build build --config Release --target llama-server --target rpc-server -j ${{ steps.cpu-cores.outputs.count }}
+          strip build/bin/*
         shell: alpine.sh {0}
 
       - name: Upload build artifact
         uses: actions/upload-artifact@v4
         with:
           name: server-${{ matrix.ssl }}-${{ matrix.avx }}
-          path: build/bin/llama-server
+          path: build/bin/*
           if-no-files-found: error
 
-      - name: Build and push Docker image
-        uses: macbre/push-to-ghcr@v13
+      - name: Build and push llama-server Docker image
+        uses: macbre/push-to-ghcr@v14
         with:
           dockerfile: .devops/llama-server.Dockerfile
           context: build/bin
@@ -81,3 +81,14 @@ jobs:
           docker_io_user: ${{ vars.DOCKER_IO_USER }}
           docker_io_token: ${{ secrets.DOCKER_IO_ACCESS_TOKEN }}
 
+      - name: Build and push rpc-server Docker image
+        uses: macbre/push-to-ghcr@v14
+        with:
+          dockerfile: .devops/rpc-server.Dockerfile
+          context: build/bin
+          image_name: ${{ github.repository }}
+          image_tag: rpc-server-${{ matrix.ssl }}-${{ matrix.avx }}
+          github_token: ${{ secrets.GHCR_ACCESS_TOKEN }}
+          docker_io_user: ${{ vars.DOCKER_IO_USER }}
+          docker_io_token: ${{ secrets.DOCKER_IO_ACCESS_TOKEN }}
+