Dockerfile

# Base image: HF text-generation-inference server
FROM ghcr.io/huggingface/text-generation-inference:latest

# Environment variables
ENV MODEL_ID=meta-llama/Llama-2-7b-chat-hf
ENV PORT=7860

# Expose the port for the HF Space
EXPOSE 7860

# Launch the HF inference server
CMD ["text-generation-launcher", "--model-id", "meta-llama/Llama-2-7b-chat-hf", "--port", "7860"]