Update: v1.9

Dockerfile

@@ -13,30 +13,36 @@ RUN apt-get update && apt-get install -y git git-lfs && rm -rf /var/lib/apt/list
 RUN useradd -m -u 1000 user
 WORKDIR /home/user/app
 
-# Create necessary directories
-RUN mkdir -p /home/user/model /home/user/data
+# Create necessary directories and set permissions
+RUN mkdir -p /home/user/model /home/user/data /home/user/.cache /usr/local/lib/python3.10/site-packages/nltk_data && \
+    chown -R user:user /home/user /usr/local/lib/python3.10/site-packages/nltk_data
 
 # Set environment variables
 ENV HF_ENDPOINT=https://hf-mirror.com
 ENV PATH="/home/user/.local/bin:$PATH"
 ENV PYTHONPATH="/home/user/app:$PYTHONPATH"
+ENV NLTK_DATA="/home/user/nltk_data"
+
+# Switch to user for pip installations
+USER user
 
 # Copy requirements first
 COPY --chown=user requirements.txt .
 
-# Install Python dependencies as root to ensure proper installation
+# Install Python dependencies
 RUN pip install --no-cache-dir -r requirements.txt && \
     pip install --no-cache-dir transformers torch sentence-transformers streamlit
 
-# Download paraphrase-multilingual-MiniLM-L12-v2
+# Download models and data
 RUN cd /home/user/model && \
     git clone https://www.modelscope.cn/Ceceliachenen/paraphrase-multilingual-MiniLM-L12-v2.git && \
     cd /home/user/model/paraphrase-multilingual-MiniLM-L12-v2 && \
     git lfs install && \
     git lfs pull
 
-# Download NLTK data
-RUN cd /home/user && \
+# Create NLTK data directory and download data
+RUN mkdir -p /home/user/nltk_data && \
+    cd /home/user && \
     git clone https://gitee.com/yzy0612/nltk_data.git --branch gh-pages && \
     cd /home/user/nltk_data && \
     mv /home/user/nltk_data/packages/* /home/user/nltk_data && \
@@ -46,21 +52,16 @@ RUN cd /home/user && \
     unzip averaged_perceptron_tagger.zip
 
 # Download demo data
-RUN cd /home/user/data && \
+RUN mkdir -p /home/user/data && \
+    cd /home/user/data && \
     git clone https://github.com/Kit-Hung/demos && \
     cp /home/user/data/demos/k8s/1-deploy/3-k8s_install/2-kubeadm/kubeadm.md /home/user/data
 
 # Copy application files
 COPY --chown=user . .
 
-# Change ownership of all files to user
-RUN chown -R user:user /home/user
-
-# Switch to non-root user
-USER user
-
 # Expose Streamlit port
 EXPOSE 8501
 
-# Run streamlit with proper path
+# Run streamlit
 CMD ["streamlit", "run", "app.py"]