Dockerfile

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

RUN apt update && apt install -y wget python3 python3-pip
RUN pip install gradio requests

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

CMD ["bash", "-c", "\
if [ ! -f /gpt-oss-20b_Q4_0.gguf ]; then \
  echo 'Downloading model...'; \
  wget -q 'https://huggingface.co/unsloth/gpt-oss-20b-GGUF/resolve/main/gpt-oss-20b_Q4_0.gguf' -O /gpt-oss-20b_Q4_0.gguf; \
fi; \
llama.cpp --server -m /gpt-oss-20b_Q4_0.gguf --port 7860 --host 0.0.0.0 -n 512 & \
python3 /app/app.py"]