FROM ghcr.io/ggml-org/llama.cpp:full

WORKDIR /app

# Python virtualenv
RUN apt-get update && apt-get install -y --no-install-recommends \
    python3-pip python3-venv \
    && rm -rf /var/lib/apt/lists/*

# Crear venv
RUN python3 -m venv /opt/venv
ENV PATH="/opt/venv/bin:$PATH"

# Instalar huggingface_hub
RUN pip install --no-cache-dir -U huggingface_hub

# Descargar modelo
RUN python3 -c 'from huggingface_hub import hf_hub_download; \
    repo="NovaCorp/Ultimate-RPG.System-3.2-1B-FP16-GGUF"; \
    hf_hub_download( \
        repo_id=repo, \
        filename="Ultimate-RPG.System-3.2-1B-FP16-Q4_K_M.gguf", \
        local_dir="/app" \
    )'

CMD ["--server", \
     "-m", "/app/Ultimate-RPG.System-3.2-1B-FP16-Q4_K_M.gguf", \
     "--host", "0.0.0.0", \
     "--port", "7860", \
     "-t", "2", \
     "--cache-type-k", "q8_0", \
     "--cache-type-v", "iq4_nl", \
     "-c", "512", \
     "-n", "2048"]