Dockerfile

FROM ghcr.io/ggml-org/llama.cpp:server
USER root
RUN apt-get update && apt-get install -y curl
RUN mkdir -p /models && \
    curl -L https://huggingface.co/unsloth/gemma-3n-E2B-it-GGUF/resolve/main/gemma-3n-E2B-it-Q4_K_M.gguf -o /models/model.gguf && \
    chown -R 1000:1000 /models
USER 1000
ENV LLAMA_ARG_MODEL=/models/model.gguf
ENV LLAMA_ARG_HOST=0.0.0.0
ENV LLAMA_ARG_PORT=7860
ENV LLAMA_ARG_THREADS=8
ENV LLAMA_ARG_BATCH_SIZE=2048
ENV LLAMA_ARG_UBATCH_SIZE=512
ENV LLAMA_ARG_CTX_SIZE=8196 
ENV LLAMA_ARG_FLASH_ATTN=true
ENV LLAMA_ARG_NO_MMAP=false 
ENV LLAMA_ARG_MLOCK=true 
HEALTHCHECK --interval=30s --timeout=15s --start-period=10s --retries=3 \
    CMD curl -f http://localhost:7860/health || exit 1
ENTRYPOINT ["/app/llama-server"]