app.py

# -*- coding: utf-8 -*-
# 財政部財政資訊中心 江信宗

# pip install langchain transformers langchain-groq chromadb langchain-community langchain-huggingface gradio

import os
from dotenv import load_dotenv
load_dotenv()
os.environ["LANGCHAIN_COMMUNITY__USER_AGENT"] = "Taiwan_Tax_KB (Colab)"
from langchain_community.utils import user_agent
from langchain_groq import ChatGroq
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_huggingface import HuggingFaceEmbeddings
from langchain_community.vectorstores import Chroma
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.chains import RetrievalQA
from langchain_community.document_loaders import WebBaseLoader, TextLoader
from langchain.prompts import PromptTemplate
from langchain.schema import Document
import gradio as gr

def initialize_llm(api_key):
    os.environ["GROQ_API_KEY"] = api_key
    return ChatGroq(
        groq_api_key=api_key,
        model_name='llama-3.1-70b-versatile'
    )
print(f"成功初始化 ChatGroq 模型")

def load_documents(sources):
    documents = []
    for source in sources:
        try:
            if isinstance(source, str):
                if source.startswith('http'):
                    loader = WebBaseLoader(source)
                else:
                    loader = TextLoader(source)
                documents.extend(loader.load())
            elif isinstance(source, dict):
                documents.append(Document(page_content=source['content'], metadata=source.get('metadata', {})))
        except Exception as e:
            print(f"Error loading source {source}: {str(e)}")
    return documents

sources = [
    "/content/TaxQADataSet_kctax.txt",
    "/content/TaxQADataSet_chutax.txt",
    "/content/HouseTaxAct1130103.txt",
    "/content/VehicleLicenseTaxAct1101230.txt",
    "/content/TaxCollectionAct1101217.txt",
    "/content/LandTaxAct1100623.txt",
    "/content/AmusementTaxAct960523.txt",
    "/content/StampTaxAct910515.txt",
    "/content/DeedTaxAct990505.txt",
    "/content/ProgressiveHouseTaxRates1130701.txt",
    "https://www.tax.ntpc.gov.tw/lp-2158-1-1-20.html",
    "https://www.tax.ntpc.gov.tw/lp-2158-1-2-20.html",
    "https://www.tax.ntpc.gov.tw/lp-2158-1-3-20.html",
    "https://www.tax.ntpc.gov.tw/lp-2158-1-4-20.html",
    "https://www.tax.ntpc.gov.tw/lp-2158-1-5-20.html",
    "https://www.tax.ntpc.gov.tw/lp-2158-1-6-20.html",
    "https://www.tax.ntpc.gov.tw/lp-2158-1-7-20.html",
    "https://www.tax.ntpc.gov.tw/lp-2158-1-8-20.html",
    "https://www.tax.ntpc.gov.tw/lp-2158-1-9-20.html",
    "https://www.tax.ntpc.gov.tw/lp-2158-1-10-20.html",
    "https://www.tax.ntpc.gov.tw/lp-2158-1-11-20.html",
    "https://www.tax.ntpc.gov.tw/lp-2158-1-12-20.html"
]

documents = load_documents(sources)
print(f"成功載入 {len(documents)} 個網址或檔案")

text_splitter = RecursiveCharacterTextSplitter(
    chunk_size=512,
    chunk_overlap=50,
    length_function=len,
    separators=["\n\n\n","\n\n", "\n", "。"]
)

split_docs = text_splitter.split_documents(documents)
print(f"分割後的文檔數量：{len(split_docs)}")

embeddings = HuggingFaceEmbeddings(model_name="BAAI/bge-large-zh-v1.5")
print(f"\n成功初始化嵌入模型")

vectorstore = Chroma.from_documents(split_docs, embeddings, persist_directory="./Knowledge-base")
print(f"成功建立 Chroma 向量資料庫")

retriever = vectorstore.as_retriever()

template = """Let's work this out in a step by step way to be sure we have the right answer. Must reply to me in Taiwanese Traditional Chinese.
在回答之前，請仔細分析檢索到的上下文，確保你的回答準確完整反映了上下文中的訊息，而不是依賴先前的知識，但在回應答案中不要提到是根據提供的上下文回答。
如果檢索到的多個上下文之間存在聯繫，請整合這些訊息以提供全面的回答，但要避免過度推斷。
如果檢索到的上下文不包含足夠回答問題的訊息，請誠實的說明，不要試圖編造答案。

上下文： {context}

問題： {question}

答案："""

PROMPT = PromptTemplate(
    template=template, input_variables=["context", "question"]
)
print(f"成功定義 Prompt Template")

def create_chain(llm):
    return RetrievalQA.from_chain_type(
        llm=llm,
        chain_type="stuff",
        retriever=retriever,
        return_source_documents=True,
        chain_type_kwargs={"prompt": PROMPT}
    )
print(f"成功建立 RAG Chain")

def generate_insight_questions(answer, api_key):
    llm = initialize_llm(api_key)
    prompt = f"""
    根據以下回答,生成3個相關的洞見問題:

    回答: {answer}

    請提供3個簡短但有深度的問題,這些問題應該:
    1. 與原始回答緊密相關
    2. 能夠引導更深入的討論
    3. 涵蓋不同的方面或角度

    請直接列出這3個問題,每個問題一行,不要添加編號或其他文字。
    """
    try:
        response = llm.invoke(prompt)
        if hasattr(response, 'content'):
            questions = response.content.split('\n')
        else:
            questions = str(response).split('\n')

        # 確保我們有至少3個問題
        while len(questions) < 3:
            questions.append("需要更多資訊嗎？")

        return questions[:3]  # 只返回前3個問題
    except Exception as e:
        print(f"Error generating insight questions: {str(e)}")
        return ["需要更多資訊嗎？", "有其他問題嗎？", "還有什麼想了解的嗎？"]

def answer_question(query, api_key):
    try:
        llm = initialize_llm(api_key)
        chain = create_chain(llm)
        result = chain({"query": query})
        answer = result["result"]

        insight_questions = generate_insight_questions(answer, api_key)

        # 確保有三個問題，如果不足則添加默認問題
        while len(insight_questions) < 3:
            insight_questions.append("需要更多資訊嗎？")

        # 分開返回答案和洞見問題
        return answer, insight_questions[:3]
    except Exception as e:
        return f"抱歉，處理您的問題時發生錯誤：{str(e)}", []

def handle_interaction(query, api_key, state):
    if state is None:
        state = {"history": []}

    answer, insight_questions = answer_question(query, api_key)

    state["history"].append((query, answer))

    insight_questions = [q if q.strip() else "需要更多資訊" for q in insight_questions]

    return answer, insight_questions[0], insight_questions[1], insight_questions[2], state, query

custom_css = """
body {
    background-color: #e8f5e9;
}
#answer-box textarea, #query-input textarea {
    font-size: 18px !important;
    background-color: #ffffff;
    border: 1px solid #81c784;
    border-radius: 8px;
}
.center-text {
    text-align: center !important;
    color: #2e7d32 !important;
}
.gradio-container {
    background-color: #c8e6c9 !important;
    border-radius: 15px !important;
    padding: 20px !important;
    box-shadow: 0 4px 6px rgba(0, 0, 0, 0.1) !important;
}
.gr-button {
    color: white !important;
    border: none !important;
    border-radius: 20px !important;
    transition: all 0.3s ease !important;
    font-weight: bold !important;
}
.gr-button:hover {
    transform: translateY(-2px) !important;
    box-shadow: 0 4px 8px rgba(0, 0, 0, 0.2) !important;
}
#submit-btn {
    background-color: #ff4081 !important;
}
#submit-btn:hover {
    background-color: #f50057 !important;
}
.insight-btn {
    background-color: #00bcd4 !important;
}
.insight-btn:hover {
    background-color: #00acc1 !important;
}
.gr-form {
    background-color: #e8f5e9 !important;
    padding: 15px !important;
    border-radius: 10px !important;
}
"""

with gr.Blocks(theme=gr.themes.Monochrome(), css=custom_css) as iface:
    gr.Markdown("# 地方稅知識庫系統 - 財政部財政資訊中心", elem_classes=["center-text"])
    gr.Markdown("※ RAG-based Q&A Web系統，建置：江信宗，LLM：Llama-3.1-70B，目前僅示範地方稅各稅目問答。", elem_classes=["center-text"])

    with gr.Row():
        query_input = gr.Textbox(lines=2, placeholder="請輸入您的問題...", label="輸入您的問題，系統將基於學習到的知識資料提供相關答案。", elem_id="query-input")
        api_key_input = gr.Textbox(type="password", placeholder="請輸入您的 API Key", label="API authentication key for large language models")

    answer_output = gr.Textbox(lines=6, label="答案：", elem_id="answer-box")

    with gr.Row():
        insight_q1 = gr.Button("洞見問題 1", visible=False, elem_classes=["insight-btn"])
        insight_q2 = gr.Button("洞見問題 2", visible=False, elem_classes=["insight-btn"])
        insight_q3 = gr.Button("洞見問題 3", visible=False, elem_classes=["insight-btn"])

    state = gr.State()
    current_question = gr.Textbox(lines=2, label="當前問題", visible=False)

    submit_btn = gr.Button("提交", elem_id="submit-btn")

    def update_ui(answer, q1, q2, q3, state, current_q):
        return [
            answer,
            gr.update(value=q1, visible=bool(q1)),
            gr.update(value=q2, visible=bool(q2)),
            gr.update(value=q3, visible=bool(q3)),
            state,
            current_q
        ]

    submit_btn.click(
        fn=handle_interaction,
        inputs=[query_input, api_key_input, state],
        outputs=[answer_output, insight_q1, insight_q2, insight_q3, state, current_question]
    ).then(
        fn=update_ui,
        inputs=[answer_output, insight_q1, insight_q2, insight_q3, state, current_question],
        outputs=[answer_output, insight_q1, insight_q2, insight_q3, state, current_question]
    )

    for btn in [insight_q1, insight_q2, insight_q3]:
        btn.click(
            lambda x: x,
            inputs=[btn],
            outputs=[query_input]
        )

if __name__ == "__main__":
    iface.launch(share=True, debug=True)