endpoints:
  - name: completions
    endpoint_type: llm/v1/completions
    model:
      provider: "huggingface-text-generation-inference"
      name: falcon-7b-instruct
      config:
        hf_server_url: http://127.0.0.1:8080