Merge pull request #647 from basetenlabs/bump-version-0.7.2

Release 0.7.2

Author: squidarth

.github/workflows/commit_new_release_to_main.yml

@@ -13,6 +13,8 @@ jobs:
     steps:
     - name: Check out code
       uses: actions/checkout@v3
+      with:
+        token: ${{ secrets.BASETENBOT_GITHUB_TOKEN }}
 
     - name: Configure Git user as basetenbot
       run: |
@@ -22,12 +24,13 @@ jobs:
     - name: Fetch all branches
       run: |
         git fetch --all --unshallow
+        git checkout release
         git pull origin release
 
     - name: Merge release into main with priority on main changes
       run: |
         git checkout main
-        git merge --strategy-option=ours release -m "Merge release into main prioritizing main changes"
+        git merge --strategy-option=ours release
         git push origin main
       env:
         GH_TOKEN: ${{ secrets.BASETENBOT_GITHUB_TOKEN }}

docker/base_images/vllm.Dockerfile

@@ -0,0 +1,26 @@
+# Use an official CUDA runtime as a parent image
+FROM nvidia/cuda:11.8.0-devel-ubuntu22.04
+
+ENV DEBIAN_FRONTEND=noninteractive
+ENV SHELL=/bin/bash
+
+# Update, upgrade, install packages and clean up
+RUN apt-get update --yes && \
+    apt-get upgrade --yes && \
+    apt install --yes --no-install-recommends git wget curl bash software-properties-common nginx && \
+    apt install python3.10-dev python3.10-venv -y --no-install-recommends && \
+    apt-get autoremove -y && \
+    apt-get clean && \
+    rm -rf /var/lib/apt/lists/* && \
+    echo "en_US.UTF-8 UTF-8" > /etc/locale.gen
+
+
+# Set up Python and pip
+RUN ln -s /usr/bin/python3.10 /usr/bin/python && \
+    rm /usr/bin/python3 && \
+    ln -s /usr/bin/python3.10 /usr/bin/python3 && \
+    curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py && \
+    python get-pip.py
+
+# Install vllm and pandas (dependency of Ray)
+RUN pip install --upgrade --no-cache-dir vllm pandas>=1.3 && rm -rf /root/.cache/pip

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "truss"
-version = "0.7.1"
+version = "0.7.2"
 description = "A seamless bridge from model development to model delivery"
 license = "MIT"
 readme = "README.md"

truss/templates/server/model_wrapper.py

@@ -366,7 +366,7 @@ def _handle_exception():
     # Note that logger.exception logs the stacktrace, such that the user can
     # debug this error from the logs.
     logging.exception("Internal Server Error")
-    raise HTTPException(status_code=500, detail={"message": "Internal Server Error"})
+    raise HTTPException(status_code=500, detail="Internal Server Error")
 
 
 def _intercept_exceptions_sync(func):

truss/templates/vllm/vllm.Dockerfile.jinja

@@ -1,4 +1,4 @@
-FROM baseten/vllm:v0.1
+FROM baseten/vllm:v0.3
 
 EXPOSE 8080-9000
 

truss/tests/test_model_inference.py

@@ -424,7 +424,7 @@ def predict(self, request):
 
         assert_logs_contain_error(container.logs(), "ValueError: error")
 
-        assert "Internal Server Error" in response.json()["error"]["message"]
+        assert "Internal Server Error" in response.json()["error"]
 
     model_preprocess_error = """
     class Model:
@@ -452,7 +452,7 @@ def predict(self, request):
         assert "error" in response.json()
 
         assert_logs_contain_error(container.logs(), "ValueError: error")
-        assert "Internal Server Error" in response.json()["error"]["message"]
+        assert "Internal Server Error" in response.json()["error"]
 
     model_postprocess_error = """
     class Model:
@@ -479,7 +479,7 @@ def postprocess(self, response):
         assert response.status_code == 500
         assert "error" in response.json()
         assert_logs_contain_error(container.logs(), "ValueError: error")
-        assert "Internal Server Error" in response.json()["error"]["message"]
+        assert "Internal Server Error" in response.json()["error"]
 
     model_async = """
     class Model:
@@ -505,7 +505,7 @@ async def predict(self, request):
 
         assert_logs_contain_error(container.logs(), "ValueError: error")
 
-        assert "Internal Server Error" in response.json()["error"]["message"]
+        assert "Internal Server Error" in response.json()["error"]
 
 
 @pytest.mark.integration