Dockerfile

# Multi-stage Dockerfile for GGUF Splitter
# Supports production, local development, and testing configurations

# Base stage with common dependencies
FROM python:3.11-slim as base

# Install system dependencies
RUN apt-get update && apt-get install -y --no-install-recommends \
    curl \
    wget \
    libgomp1 \
    && rm -rf /var/lib/apt/lists/*

# Build llama.cpp binaries (only needed for test stage)
FROM base as llama-builder

RUN apt-get update && apt-get install -y --no-install-recommends \
    git \
    build-essential \
    cmake \
    && rm -rf /var/lib/apt/lists/*

RUN git clone https://github.com/ggerganov/llama.cpp.git /tmp/llama.cpp
WORKDIR /tmp/llama.cpp
RUN cmake -B build -DCMAKE_BUILD_TYPE=Release
RUN cmake --build build --config Release -j --target llama-gguf-split

# Production stage
FROM base as production

# Download pre-built llama.cpp binaries
WORKDIR /tmp

RUN LLAMA_VERSION=$(wget -qO- https://api.github.com/repos/ggml-org/llama.cpp/releases/latest | grep -oP '"tag_name": "\K(.*)(?=")') && \
    wget -q "https://github.com/ggml-org/llama.cpp/releases/download/${LLAMA_VERSION}/llama-${LLAMA_VERSION}-bin-ubuntu-x64.tar.gz" && \
    tar -xzf "llama-${LLAMA_VERSION}-bin-ubuntu-x64.tar.gz" && \
    mkdir -p /usr/local/llama.cpp && \
    mv llama-${LLAMA_VERSION}/* /usr/local/llama.cpp/ && \
    rm -rf /tmp/*

ENV PATH="/usr/local/llama.cpp:${PATH}"
ENV LLAMA_GGUF_SPLIT_PATH=/usr/local/llama.cpp/llama-gguf-split

WORKDIR /home/user/app

COPY requirements.txt /home/user/app/requirements.txt

# Create virtual environment
RUN python3 -m venv /opt/venv
ENV PATH="/opt/venv/bin:$PATH"
RUN /opt/venv/bin/pip install --upgrade pip
RUN /opt/venv/bin/pip install -r /home/user/app/requirements.txt

ENV HF_HUB_HOME=/home/user/app
ENV GRADIO_HF_OFFLINE_SPACE_OK=1

COPY . /home/user/app

ENTRYPOINT ["python", "app.py"]

# Local development stage
FROM production as local

ENV GRADIO_SERVER_NAME="0.0.0.0"
ENV RUN_LOCALLY=1

EXPOSE 7860

CMD ["python", "app.py"]

# Test stage
FROM base as test

# Copy llama-gguf-split from builder stage
COPY --from=llama-builder /tmp/llama.cpp/build/bin/llama-gguf-split /usr/local/bin/
COPY --from=llama-builder /tmp/llama.cpp/build/bin/*.so* /usr/local/bin/

WORKDIR /home/user/app

COPY requirements.txt /home/user/app/requirements.txt

RUN pip install --no-cache-dir -r /home/user/app/requirements.txt

# Install test dependencies
COPY requirements-dev.txt /home/user/app/requirements-dev.txt
RUN pip install --no-cache-dir -r /home/user/app/requirements-dev.txt

ENV PYTHONPATH=/home/user/app
ENV RUN_LOCALLY=1
ENV LLAMA_GGUF_SPLIT_PATH=/usr/local/bin/llama-gguf-split
ENV LD_LIBRARY_PATH=/usr/local/bin

COPY . /home/user/app

CMD ["pytest", "tests/", "-v", "--tb=short"]