Update Dockerfile

Dockerfile

@@ -1,8 +1,5 @@
-# Use a base image that has Python and can support Ollama (e.g., Ubuntu or a CUDA image for GPU)
-# For GPU support, use nvidia/cuda base image (e.g., nvidia/cuda:12.1.1-devel-ubuntu22.04)
-# For CPU only, a regular Ubuntu or Python image should work.
-# Example for CPU (adjust as needed):
-FROM ubuntu:22.04
+# Base image with Python
+FROM python:3.10-slim
 
 # Set environment variables
 ENV OLLAMA_HOST=0.0.0.0:11434 # Ollama's API will be accessible on all interfaces
@@ -36,25 +33,25 @@ COPY --chown=user:user . .
 # The `ollama serve` command must be running in the background for `ollama pull` to work.
 # We'll use a trick with a shell script for this.
 
-# Create a script to start Ollama and pull models
+# Create a startup script
 RUN echo '#!/bin/bash' > /home/user/start_ollama_and_app.sh && \
-    echo 'nohup ollama serve &' >> /home/user/start_ollama_and_app.sh && \
+    echo 'echo "Starting Ollama..."' >> /home/user/start_ollama_and_app.sh && \
+    echo 'ollama serve & ' >> /home/user/start_ollama_and_app.sh && \
     echo 'sleep 5' >> /home/user/start_ollama_and_app.sh && \
     echo 'ollama pull nomic-embed-text' >> /home/user/start_ollama_and_app.sh && \
     echo 'ollama pull llama2' >> /home/user/start_ollama_and_app.sh && \
     echo 'ollama pull mistral' >> /home/user/start_ollama_and_app.sh && \
     echo 'ollama pull tinyllama' >> /home/user/start_ollama_and_app.sh && \
-    echo 'sleep 5' >> /home/user/start_ollama_and_app.sh && \
-    echo 'streamlit run app.py --server.port 7860 --server.address 0.0.0.0' >> /home/user/start_ollama_and_app.sh && \
+    echo 'echo "Launching Streamlit..."' >> /home/user/start_ollama_and_app.sh && \
+    echo 'streamlit run src/streamlit_app.py --server.port=8501 --server.address=0.0.0.0' >> /home/user/start_ollama_and_app.sh && \
     chmod +x /home/user/start_ollama_and_app.sh
 
-# Expose the port your Streamlit app listens on (as defined in README.md)
+# Expose Streamlit and Ollama ports
 EXPOSE 8501
-
-# Expose Ollama's API port (if your app directly connects to it)
 EXPOSE 11434
 
+# Healthcheck for Hugging Face Spaces
+HEALTHCHECK CMD curl --fail http://localhost:8501/_stcore/health || exit 1
 
-HEALTHCHECK CMD curl --fail http://localhost:8501/_stcore/health
-
-ENTRYPOINT ["streamlit", "run", "src/streamlit_app.py", "--server.port=8501", "--server.address=0.0.0.0"]
+# Start both Ollama and Streamlit via the script
+ENTRYPOINT ["/home/user/start_ollama_and_app.sh"]