Update Dockerfile

Dockerfile

@@ -1,4 +1,4 @@
-FROM python:3.10-slim
+FROM python:3.10-slim AS builder
 
 ENV PYTHONDONTWRITEBYTECODE=1 \
     PYTHONUNBUFFERED=1 \
@@ -7,16 +7,44 @@ ENV PYTHONDONTWRITEBYTECODE=1 \
 
 WORKDIR /app
 
-# Runtime deps (libgomp helps llama-cpp wheels / OpenMP)
+# Build deps: required to compile llama-cpp-python from source on HF
 RUN apt-get update \
-    ; apt-get install -y --no-install-recommends libgomp1 ca-certificates \
+    ; apt-get install -y --no-install-recommends \
+        build-essential \
+        gcc \
+        g++ \
+        cmake \
+        git \
+        libgomp1 \
+        ca-certificates \
     ; rm -rf /var/lib/apt/lists/*
 
 # Install deps first (better layer caching)
 COPY requirements.txt ./
+
+# Optional: keep build simple and CPU-friendly
+ENV CMAKE_ARGS="-DLLAMA_BLAS=OFF"
+
 RUN pip install --upgrade pip \
     ; pip install -r requirements.txt
 
+
+FROM python:3.10-slim AS runtime
+
+ENV PYTHONDONTWRITEBYTECODE=1 \
+    PYTHONUNBUFFERED=1 \
+    PORT=7860
+
+WORKDIR /app
+
+# Runtime deps (OpenMP)
+RUN apt-get update \
+    ; apt-get install -y --no-install-recommends libgomp1 ca-certificates \
+    ; rm -rf /var/lib/apt/lists/*
+
+# Copy installed python packages from builder
+COPY --from=builder /usr/local /usr/local
+
 # Copy app and model files
 COPY app.py ./
 COPY *.gguf ./