Update Dockerfile

Dockerfile

@@ -1,29 +1,22 @@
-FROM python:3.10-bookworm
+# This image comes with llama-cpp-python PRE-INSTALLED and optimized
+FROM ghcr.io/abetlen/llama-cpp-python:latest
 
-# 1. Install system requirements
-RUN apt-get update && apt-get install -y \
-    curl git procps libopenblas-dev \
-    && rm -rf /var/lib/apt/lists/*
+# Switch to root to set up the HF user
+USER root
+RUN apt-get update && apt-get install -y curl git procps && rm -rf /var/lib/apt/lists/*
 
-# 2. Set up user (HF requires UID 1000)
+# Setup HF compatible user
 RUN useradd -m -u 1000 user
-ENV HOME=/home/user \
-    PATH=/home/user/.local/bin:$PATH
-WORKDIR $HOME/app
+WORKDIR /home/user/app
 
-# 3. Pre-create model directory
-RUN mkdir -p $HOME/app/model_cache && chown -R user:user $HOME/app
-USER user
-
-# 4. Install Gradio first to establish the web server
-RUN pip install --no-cache-dir gradio huggingface-hub hf-transfer numpy
+# Install the few extra things we need
+RUN pip install --no-cache-dir gradio huggingface-hub hf-transfer
 
-# 5. Install llama-cpp-python using the OFFICIAL CPU INDEX
-# This is safer than a random GitHub link as it auto-matches your OS
-RUN pip install llama-cpp-python \
-    --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cpu
+# Give permissions to the user
+RUN chown -R user:user /home/user/app
+USER user
 
 COPY --chown=user . .
 
-# Use -u for unbuffered logs (so you see the download progress)
+# Use the same app.py logic
 CMD ["python", "-u", "app.py"]