refactor: streamline Dockerfile and remove unused files for FluxSchnell (#506)

Author: gooishin

usecases/ai/microservices/text-to-image/Dockerfile

@@ -5,23 +5,30 @@ FROM debian:12-slim
 ARG DEBIAN_FRONTEND=noninteractive
 
 # Define the default allowed models and the default model
-ARG ALLOWED_MODELS="stable-diffusion-v3.5 stable-diffusion-xl flux-schnell"
 ARG DEFAULT_MODEL="stable-diffusion-v3.5"
+ARG HOST="0.0.0.0"
+ARG PORT="8100"
 
-# Set default model as a build argument and runtime environment variable
+# Set environment variables for the default model and allowed models
+ENV ALLOWED_MODELS="stable-diffusion-v3.5 stable-diffusion-xl flux-schnell"
 ENV DEFAULT_MODEL=${DEFAULT_MODEL}
+ENV HOST=${HOST}
+ENV PORT=${PORT}
+
+# Do checking if the default model is in the allowed models list
+RUN if [[ ! "${ALLOWED_MODELS}" =~ "${DEFAULT_MODEL}" ]]; then \
+        echo "Error: The default model '${DEFAULT_MODEL}' is not in the allowed models list." >&2; \
+        exit 1; \
+    fi
 
 # Ensure the container is running as root for package installations
 USER root
 
-# Set the working directory early for clarity and organization
-WORKDIR /usr/src/app
-
 # Install system dependencies, including Python 3.11 and venv
 RUN apt-get update \
     && apt-get install -y --no-install-recommends \
         ffmpeg \
-        wget \
+        curl \
         git \
         gnupg2 \
         libtbb12 \
@@ -30,23 +37,21 @@ RUN apt-get update \
     && apt-get clean && rm -rf /var/lib/apt/lists/*
 
 # Install Intel GPU drivers with error handling and version pinning
-RUN mkdir /tmp/neo \
-    && cd /tmp/neo \
-    && wget -q https://github.com/intel/intel-graphics-compiler/releases/download/igc-1.0.17791.9/intel-igc-core_1.0.17791.9_amd64.deb \
-    && wget -q https://github.com/intel/intel-graphics-compiler/releases/download/igc-1.0.17791.9/intel-igc-opencl_1.0.17791.9_amd64.deb \
-    && wget -q https://github.com/intel/compute-runtime/releases/download/24.39.31294.12/intel-level-zero-gpu_1.6.31294.12_amd64.deb \
-    && wget -q https://github.com/intel/compute-runtime/releases/download/24.39.31294.12/intel-opencl-icd_24.39.31294.12_amd64.deb \
-    && wget -q https://github.com/intel/compute-runtime/releases/download/24.39.31294.12/libigdgmm12_22.5.2_amd64.deb \
+WORKDIR /tmp/neo
+RUN curl -sSLO https://github.com/intel/intel-graphics-compiler/releases/download/igc-1.0.17791.9/intel-igc-core_1.0.17791.9_amd64.deb \
+    && curl -sSLO https://github.com/intel/intel-graphics-compiler/releases/download/igc-1.0.17791.9/intel-igc-opencl_1.0.17791.9_amd64.deb \
+    && curl -sSLO https://github.com/intel/compute-runtime/releases/download/24.39.31294.12/intel-level-zero-gpu_1.6.31294.12_amd64.deb \
+    && curl -sSLO https://github.com/intel/compute-runtime/releases/download/24.39.31294.12/intel-opencl-icd_24.39.31294.12_amd64.deb \
+    && curl -sSLO https://github.com/intel/compute-runtime/releases/download/24.39.31294.12/libigdgmm12_22.5.2_amd64.deb \
     && dpkg -i *.deb \
     && rm -rf /tmp/neo
 
 # Install Intel NPU drivers with error handling and version pinning
-RUN mkdir /tmp/npu-driver \
-    && cd /tmp/npu-driver \
-    && wget -q https://github.com/intel/linux-npu-driver/releases/download/v1.10.0/intel-driver-compiler-npu_1.10.0.20241107-11729849322_ubuntu22.04_amd64.deb \
-    && wget -q https://github.com/intel/linux-npu-driver/releases/download/v1.10.0/intel-fw-npu_1.10.0.20241107-11729849322_ubuntu22.04_amd64.deb \
-    && wget -q https://github.com/intel/linux-npu-driver/releases/download/v1.10.0/intel-level-zero-npu_1.10.0.20241107-11729849322_ubuntu22.04_amd64.deb \
-    && wget -q https://github.com/oneapi-src/level-zero/releases/download/v1.17.6/level-zero_1.17.6+u22.04_amd64.deb \
+WORKDIR /tmp/npu-driver
+RUN curl -sSLO https://github.com/intel/linux-npu-driver/releases/download/v1.10.0/intel-driver-compiler-npu_1.10.0.20241107-11729849322_ubuntu22.04_amd64.deb \
+    && curl -sSLO https://github.com/intel/linux-npu-driver/releases/download/v1.10.0/intel-fw-npu_1.10.0.20241107-11729849322_ubuntu22.04_amd64.deb \
+    && curl -sSLO https://github.com/intel/linux-npu-driver/releases/download/v1.10.0/intel-level-zero-npu_1.10.0.20241107-11729849322_ubuntu22.04_amd64.deb \
+    && curl -sSLO https://github.com/oneapi-src/level-zero/releases/download/v1.17.6/level-zero_1.17.6+u22.04_amd64.deb \
     && dpkg -i *.deb \
     && rm -rf /tmp/npu-driver
 
@@ -57,46 +62,25 @@ RUN groupadd -r openvino && useradd -r -g openvino -G video openvino
 RUN mkdir -p /home/openvino && \
     chown -R openvino:openvino /home/openvino
 
-# Copy application files and adjust permissions (excluding the virtual environments)
-COPY . /usr/src/app
-RUN find /usr/src/app -not -path "/usr/src/app/*/.venv*" -exec chown openvino:openvino {} + \
-    && find /usr/src/app -not -path "/usr/src/app/*/.venv*" -exec chmod 755 {} +
-
-# Remove existing virtual environments to ensure a clean install
-RUN rm -rf /usr/src/app/*/.venv
-
-# Create a Python virtual environment for each model and install dependencies
-RUN for model in $ALLOWED_MODELS; do \
-        python3.11 -m venv /usr/src/app/$model/.venv; \
-        /usr/src/app/$model/.venv/bin/python -m pip install --no-cache-dir --upgrade pip; \
-        if [ -f "/usr/src/app/$model/requirements.txt" ]; then \
-            /usr/src/app/$model/.venv/bin/python -m pip install --no-cache-dir -r /usr/src/app/$model/requirements.txt; \
-        fi; \
-        chown -R openvino:openvino /usr/src/app/$model/.venv; \
-    done
-
-# Set the environment variable for the virtual environment based on the selected model
-ENV PATH="/usr/src/app/${SELECTED_MODEL}/.venv/bin:$PATH"
-
-# Switch to the non-root user for security
 USER openvino
+WORKDIR /usr/src/app/${DEFAULT_MODEL}
 
-# Set the working directory based on the selected model
-WORKDIR /usr/src/app/${SELECTED_MODEL}
+# Create a virtual environment for the default model and install dependencies
+RUN python3 -m venv .venv
+ENV PATH="/usr/src/app/${DEFAULT_MODEL}/.venv/bin:$PATH"
+COPY --chown=openvino:openvino ./${DEFAULT_MODEL}/requirements.txt .
+RUN python3 -m pip install --no-cache-dir --upgrade pip \
+    && python3 -m pip install --no-cache-dir -r requirements.txt
+
+# Copy the model files and scripts into the container
+COPY --chown=openvino:openvino ./${DEFAULT_MODEL} .
 
 # Expose port for the microservice
-EXPOSE 8100
-
-# Allow runtime injection of Hugging Face Token and model selection
-# If HF_TOKEN is not provided, models requiring it will fail securely.
-CMD ["bash", "-c", \
-     "SELECTED_MODEL=${MODEL:-$DEFAULT_MODEL} && \
-      export PATH=/usr/src/app/${SELECTED_MODEL}/.venv/bin:$PATH && \
-      export HF_TOKEN=${HF_TOKEN:-''} && \
-      echo Using model: $SELECTED_MODEL && \
-      echo HF_TOKEN set to: ${HF_TOKEN:0:5}****** && \
-      /usr/src/app/${SELECTED_MODEL}/.venv/bin/python /usr/src/app/${SELECTED_MODEL}/backend/server.py"]
+EXPOSE ${PORT}
+
+# Set the entrypoint to the server script
+CMD ["bash", "-c", "python3 backend/server.py"]
 
 # Add a basic health check
 HEALTHCHECK --interval=30s --timeout=5s --start-period=10s \
-  CMD exit 0
+  CMD curl -f http://localhost:${PORT}/health || exit 1