fix dockerfile

Dockerfile

@@ -1,40 +1,52 @@
-FROM python:3.12-slim
+FROM debian:bookworm-slim AS builder
 
-# Install build dependencies for llama-cpp-python
+# Install build dependencies
 RUN apt-get update && apt-get install -y --no-install-recommends \
-    gcc \
-    g++ \
+    git \
+    build-essential \
     cmake \
+    ca-certificates \
     && rm -rf /var/lib/apt/lists/*
 
-RUN pip install --no-cache-dir --root-user-action=ignore uv
+# Clone and build llama.cpp
+WORKDIR /build
+RUN git clone https://github.com/ggerganov/llama.cpp.git && \
+    cd llama.cpp && \
+    cmake -B build -DCMAKE_BUILD_TYPE=Release -DGGML_NATIVE=OFF -DGGML_AVX2=OFF && \
+    cmake --build build --config Release --target llama-server -j$(nproc)
 
-WORKDIR /app
+# Runtime stage
+FROM debian:bookworm-slim
 
-COPY pyproject.toml /app/pyproject.toml
-
-# Install dependencies (llama-cpp-python will compile but with minimal features)
-#ENV CMAKE_ARGS="-DLLAMA_BLAS=OFF -DLLAMA_AVX2=OFF -DLLAMA_FMA=OFF -DLLAMA_F16C=OFF"
-RUN uv pip install --system -r pyproject.toml
-
-COPY . /app
+# Install runtime dependencies
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    curl \
+    ca-certificates \
+    && rm -rf /var/lib/apt/lists/*
 
-# RUN uv pip install --system --no-cache .
+# Copy llama-server binary from builder
+COPY --from=builder /build/llama.cpp/build/bin/llama-server /usr/local/bin/llama-server
 
-# Create a non-root user
-# explicitly create the HF_HOME directory and give the user ownership
-# so they can save the downloaded model there.
+# Create non-root user
 RUN useradd -m -u 1000 user && \
-    mkdir -p /home/user/.cache/huggingface && \
+    mkdir -p /home/user/.cache/llama.cpp && \
     chown -R user:user /home/user
 
 USER user
+WORKDIR /home/user
 
+# Set environment variables
 ENV HOME=/home/user \
-    PATH=/home/user/.local/bin:$PATH \
-    HF_HOME=/home/user/.cache/huggingface
+    LLAMA_CACHE=/home/user/.cache/llama.cpp \
+    PATH=/home/user/.local/bin:$PATH
 
 EXPOSE 7860
 
-# Start the application
-CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]
+# Start llama-server with HuggingFace model
+# Using TheBloke's DeepSeek Coder GGUF model
+CMD ["llama-server", \
+     "-hf", "TheBloke/deepseek-coder-6.7B-instruct-GGUF:deepseek-coder-6.7b-instruct.Q4_K_M.gguf", \
+     "--host", "0.0.0.0", \
+     "--port", "7860", \
+     "-c", "2048", \
+     "--metrics"]