python-services/Retrieve/config/app_config_dev.yaml

qa-llm:
  main:
    model: deepseek-r1
    api_key: sk-sk-*************
    base_url: https://dashscope.aliyuncs.com/compatible-mode/v1
    max_tokens: 1024
    temperature: 0.7
  backup:
    model: qwen-plus-latest
    api_key: sk-sk-*************
    base_url: https://dashscope.aliyuncs.com/compatible-mode/v1
    max_tokens: 1024
    temperature: 0.7

rewrite-llm:
  backup:
    model: gpt-4o
    api_key: sk-**********
    base_url: https://openai.sohoyo.io/v1
    max_tokens: 1024
    temperature: 0.7
  main:
    model: qwen-plus-latest
    api_key: sk-sk-*************
    base_url: https://dashscope.aliyuncs.com/compatible-mode/v1
    max_tokens: 1024
    temperature: 0.7

recall:
  pubmed_topk: 30
  es_topk: 30

qa-topk:
  personal_vector: 40
  pubmed: 10
  web: 5

qa-prompt-max-token:
  max_tokens: 120000


chat:
  rag_prompt: |
    # The following contents are the search results related to the user's message:
        {search_results}
        In the search results I provide to you, each result is formatted as [document X begin]...[document X end], where X represents the numerical index of each article. 
        When responding, please keep the following points in mind:
        - Today is {cur_date}.
        - Not all content in the search results is closely related to the user's question. You need to evaluate and filter the search results based on the question.
        - If all the search results are irrelevant, please answer the question by yourself  professionally and concisely.
        - The search results may focus only on a few points, use the information it provided, but do not favor those points in your answer, reason and answer by yourself all-sidedly with full consideration. 
        - For listing-type questions (e.g., listing all flight information), try to limit the answer to 10 key points and inform the user that they can refer to the search sources for complete information. Prioritize providing the most complete and relevant items in the list. Avoid mentioning content not provided in the search results unless necessary.
        - If the response is lengthy, structure it well and summarize it in paragraphs. If a point-by-point format is needed, try to limit it to 5 points and merge related content.
        - For objective Q&A, if the answer is very brief, you may add one or two related sentences to enrich the content.
        - Choose an appropriate and visually appealing format for your response based on the user's requirements and the content of the answer, ensuring strong readability.
        - Your answer should synthesize information from multiple relevant documents.
        - Unless the user requests otherwise, your response should be in the same language as the user's question.
        # The user's message is:
        {question}
        - The content should be concise and direct, and you MUST include proper citations using ONLY "[bdd-rag-citation:X]" format reference marks to indicate the sources of your information. Do NOT use any other citation formats such as [document X], [Author, Year], or parenthetical bibliographical references.