Update Dockerfile

Dockerfile

@@ -1,26 +1,26 @@
-FROM ollama/ollama:latest
+FROM ollama/ollama
 
-# Create a user and set up permissions
-RUN useradd -ms /bin/bash ollama-user
-ENV HOME=/home/ollama-user
-WORKDIR $HOME
+# 1. User & Permissions setup
 
-# Ensure the user has access to the directory
-RUN mkdir -p $HOME/.ollama && chown -R ollama-user:ollama-user $HOME/.ollama
+RUN mkdir -p /home/user/.ollama
+RUN chown -R 1000:1000 /home/user
+ENV HOME=/home/user
 
-# Install netcat (nc) for checking server readiness
-RUN apt-get update && apt-get install -y netcat-openbsd
+# 2. Network Config
 
-
-# Copy the entrypoint script
-COPY entrypoint.sh /usr/local/bin/entrypoint.sh
-RUN chmod +x /usr/local/bin/entrypoint.sh
-
-# Switch to the non-root user
-USER ollama-user
-
-# Set Ollama to listen on the correct port
 ENV OLLAMA_HOST=0.0.0.0:7860
+ENV OLLAMA_ORIGINS=*
+
+# 3. CRITICAL OPTIMIZATIONS for 2-3 Users
+# Parallel: Allows 2 requests to run at the exact same time.
+ENV OLLAMA_NUM_PARALLEL=2
+# Keep Alive: Never unload the model from RAM (keeps it ready instantly).
+ENV OLLAMA_KEEP_ALIVE=-1
+# Max Models: Force Ollama to focus only on this one model to save RAM.
+ENV OLLAMA_MAX_LOADED_MODELS=1
+# 4. Switch to user
+USER 1000
 EXPOSE 7860
-
-ENTRYPOINT ["/usr/local/bin/entrypoint.sh"]
+# 5. Start, optimize, and keep running
+# We use 'qwen3:0.6b' (or the specific tag you need)
+ENTRYPOINT ["/bin/sh", "-c", "ollama serve & sleep 10 && ollama pull qwen3.5:0.8b && wait"]