# Dockerfile — Cerebro F — TheLiquidator v6.0 (BitNet REAL — ik_llama.cpp)
# OrderFlowStressIndex+DeltaCVD

# ── Etapa 1: MODEL FETCHER ────────────────────────────────────────────────────
FROM python:3.11-slim AS model-fetcher

RUN apt-get update && apt-get install -y --no-install-recommends \
        ca-certificates wget \
    && rm -rf /var/lib/apt/lists/*

RUN pip install --no-cache-dir huggingface_hub

COPY download_model.py /tmp/download_model.py
RUN python3 /tmp/download_model.py

# ── Etapa 2: RUNTIME ──────────────────────────────────────────────────────────
FROM python:3.11-slim AS runtime

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

RUN wget -q --user-agent="Mozilla/5.0" -O /tmp/ik.zip \
        "https://github.com/Thireus/ik_llama.cpp/releases/download/main-b4801-b275691/ik_llama-main-b4801-b275691-bin-ubuntu-x64-avx512_vnni_vbmi.zip" \
    && mkdir -p /app && unzip -q /tmp/ik.zip -d /app && rm /tmp/ik.zip \
    && BINARY=$(find /app -name "llama-server" -type f | head -1) \
    && ln -sf "$BINARY" /usr/local/bin/llama-server && chmod +x "$BINARY"

ENV PATH="/app:$PATH"
ENV LD_LIBRARY_PATH="/app:$LD_LIBRARY_PATH"

WORKDIR /app

COPY --from=model-fetcher /models /models

RUN pip install --no-cache-dir \
        fastapi==0.111.0 \
        "uvicorn[standard]==0.30.1" \
        httpx==0.27.0 \
        numpy==1.26.4

ENV MODEL_PATH="/models/ggml-model-i2_s.gguf"
ENV N_CTX=2048
ENV N_THREADS=2
ENV BITNET_PORT=8080
ENV PORT=7860
ENV CEREBRO_ID=F
ENV BITNET_TIMEOUT=45.0
ENV KILL_THRESHOLD=0.82
ENV KILL_THRESHOLD_SCALP=0.80
ENV HOLD_THRESHOLD=0.18
ENV STAGNATION_BARS=5
ENV OSI_FAST_EXIT=80
ENV OSI_KILL_ZONE=60
ENV OSI_HOLD_ZONE=30
ENV OSI_FALLBACK_KILL=55

COPY server_F.py ./
COPY start_F.sh ./start.sh
RUN chmod +x start.sh

EXPOSE 7860

HEALTHCHECK --interval=30s --timeout=15s --start-period=300s --retries=5 \
    CMD curl -f http://localhost:7860/health || exit 1

CMD ["./start.sh"]