Dockerfile

FROM python:3.11-slim

ENV DEBIAN_FRONTEND=noninteractive \
    MODEL_REPO=yuxinlu1/gemma-4-12B-agentic-fable5-composer2.5-v2-3.5x-tau2-GGUF \
    MODEL_FILE=gemma4-v2-Q4_K_M.gguf \
    MODEL_DIR=/data/models/gemma4-coder \
    LLAMA_VERSION=b9592 \
    LLAMA_DIR=/opt/llama.cpp \
    LLAMA_SERVER_BIN=/opt/llama.cpp/llama-server \
    LD_LIBRARY_PATH=/opt/llama.cpp \
    LLAMA_HOST=0.0.0.0 \
    LLAMA_PORT=7860 \
    THREADS=4 \
    CTX_SIZE=2048 \
    BATCH_SIZE=default \
    UBATCH_SIZE=default \
    FLASH_ATTN=default \
    CACHE_TYPE_K=default \
    CACHE_TYPE_V=default \
    GPU_LAYERS=0 \
    TEMPERATURE=0.2 \
    TOP_P=0.95 \
    TOP_K=64 \
    REPEAT_PENALTY=1.08 \
    HF_XET_HIGH_PERFORMANCE=1 \
    PYTHONUNBUFFERED=1

RUN apt-get update && apt-get install -y --no-install-recommends \
    ca-certificates \
    curl \
    libgomp1 \
    libstdc++6 \
    && rm -rf /var/lib/apt/lists/*

RUN mkdir -p "${LLAMA_DIR}" \
    && curl -fL "https://github.com/ggml-org/llama.cpp/releases/download/${LLAMA_VERSION}/llama-${LLAMA_VERSION}-bin-ubuntu-x64.tar.gz" \
        | tar -xz --strip-components=1 -C "${LLAMA_DIR}" \
    && chmod +x "${LLAMA_SERVER_BIN}"

RUN pip install --no-cache-dir \
    huggingface_hub

WORKDIR /app
COPY app.py /app/app.py

EXPOSE 7860
CMD ["python", "app.py"]