Enables testing for pull requests

triton-inference-server · Nov 23, 2024 · 0266303 · 0266303
1 parent 3423591
commit 0266303
Show file tree

Hide file tree

Showing 3 changed files with 59 additions and 6 deletions.
diff --git a/.github/workflows/build-and-test.yml b/.github/workflows/build-and-test.yml
@@ -0,0 +1,57 @@
+name: Build And Test
+
+on:
+    pull_request:
+      branches:
+        - main
+      types: [synchronize, opened, reopened, ready_for_review]
+
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    container:
+      image: nvcr.io/nvidia/tritonserver:24.10-py3
+      volumes:
+        - ${{ github.workspace }}:/core
+
+    steps:
+      # Based on https://stackoverflow.com/q/75536771
+      - name: Free disk space
+        run: |
+          sudo docker system prune -a -f
+          sudo rm -rf \
+            /usr/share/dotnet "$AGENT_TOOLSDIRECTORY" /usr/local/lib/android /opt/ghc \
+            /usr/local/share/powershell /usr/share/swift /usr/local/.ghcup \
+            /usr/lib/jvm
+
+          sudo apt-get purge microsoft-edge-stable || true
+          sudo apt-get purge google-cloud-cli || true
+          sudo apt-get purge dotnet-sdk-* || true
+          sudo apt-get purge google-chrome-stable || true
+
+          sudo apt-get autoremove -y
+          sudo apt-get autoclean -y
+
+      - uses: actions/checkout@v3
+
+      - name: Install dependencies
+        run: |
+          apt update
+          apt install -y --no-install-recommends clang-format-15 cmake libb64-dev rapidjson-dev libre2-dev
+          wget -O /tmp/boost.tar.gz https://archives.boost.io/release/1.80.0/source/boost_1_80_0.tar.gz && (cd /tmp && tar xzf boost.tar.gz) && mv /tmp/boost_1_80_0/boost /usr/include/boost
+          pip install build pytest
+
+      - name: Build
+        run: |
+          mkdir -p /core/build
+          cd /core/build
+          cmake -DCMAKE_INSTALL_PREFIX:PATH=`pwd`/install -DTRITON_CORE_HEADERS_ONLY=OFF ..
+          export TRITON_PYBIND="_c/triton_bindings.cpython-310-x86_64-linux-gnu.so"
+          make -j8
+
+      - name: Run tests with pytest
+        run: |
+          cd /core
+          python3 -m pip install --force-reinstall build/python/generic/wheel/dist/tritonserver-*.whl
+          pytest python/test -v
diff --git a/python/test/test_api.py b/python/test/test_api.py
@@ -357,11 +357,7 @@ def test_stop(self):
                     {
                         "backend": "python",
                         "parameters": {"decoupled": {"string_value": "False"}},
-                        # Keep instance count low for fast startup/cleanup.
-                        # Alternatively can use KIND_CPU here, but keeping gpus/count explicit.
-                        "instance_group": [
-                            {"kind": "KIND_GPU", "gpus": [0], "count": 1}
-                        ],
+                        "instance_group": [{"kind": "KIND_CPU"}],
                     }
                 )
             },

diff --git a/src/test/input_byte_size_test.cc b/src/test/input_byte_size_test.cc
@@ -378,7 +378,7 @@ TEST_F(InputByteSizeTest, InputByteSizeLarge)
       "setting request release callback");
 
   // Define input shape and data
-  size_t element_cnt = (1LL << 31) / sizeof(float);
+  int64_t element_cnt = (1LL << 31) / sizeof(float);
   std::vector<int64_t> shape{1, element_cnt};
   std::vector<float> input_data(element_cnt, 1);
   const auto input0_byte_size = sizeof(input_data[0]) * input_data.size();