123 lines
4.5 KiB
Python
123 lines
4.5 KiB
Python
# -*- coding: utf-8 -*-
|
||
import time
|
||
import json
|
||
|
||
|
||
def create_neutral_debator(llm):
|
||
"""
|
||
建立一個中立的風險辯論員節點。
|
||
|
||
這個節點在風險評估辯論中扮演中立派的角色。
|
||
其目標是提供一個平衡的視角,權衡交易員決策的潛在利益和風險。
|
||
它會挑戰過於樂觀或過於謹慎的觀點,並倡導一個溫和、可持續的策略。
|
||
|
||
Args:
|
||
llm: 用於生成回應的語言模型。
|
||
|
||
Returns:
|
||
function: 一個代表中立辯論員節點的函式,可在 langgraph 中使用。
|
||
"""
|
||
|
||
def neutral_node(state) -> dict:
|
||
"""
|
||
中立辯論員節點的執行函式。
|
||
|
||
Args:
|
||
state (dict): 當前的圖狀態。
|
||
|
||
Returns:
|
||
dict: 更新後的狀態,包含新的風險辯論狀態。
|
||
"""
|
||
# 從狀態中獲取風險辯論的相關資訊
|
||
risk_debate_state = state["risk_debate_state"]
|
||
history = risk_debate_state.get("history", "")
|
||
neutral_history = risk_debate_state.get("neutral_history", "")
|
||
|
||
# 獲取其他辯論者的最新回應
|
||
current_risky_response = risk_debate_state.get("current_risky_response", "")
|
||
current_safe_response = risk_debate_state.get("current_safe_response", "")
|
||
|
||
# 從狀態中獲取各類分析報告
|
||
market_research_report = state["market_report"]
|
||
sentiment_report = state["sentiment_report"]
|
||
news_report = state["news_report"]
|
||
fundamentals_report = state["fundamentals_report"]
|
||
|
||
# 獲取交易員的決策
|
||
trader_decision = state["trader_investment_plan"]
|
||
|
||
# 定義文本截斷函數以避免超過 token 限制
|
||
def truncate_text(text, max_chars):
|
||
"""截斷文本到指定字符數"""
|
||
if len(text) <= max_chars:
|
||
return text
|
||
return text[:max_chars] + "\n...(內容已截斷)"
|
||
|
||
# 截斷各類輸入以控制 token 使用量
|
||
# 模型限制: 8192 tokens,目標: < 3500 字符
|
||
market_research_report = truncate_text(market_research_report, 500)
|
||
sentiment_report = truncate_text(sentiment_report, 500)
|
||
news_report = truncate_text(news_report, 600)
|
||
fundamentals_report = truncate_text(fundamentals_report, 600)
|
||
trader_decision = truncate_text(trader_decision, 800)
|
||
history = truncate_text(history, 400)
|
||
current_risky_response = truncate_text(current_risky_response, 300)
|
||
current_safe_response = truncate_text(current_safe_response, 300)
|
||
|
||
# 建立提示 (prompt)
|
||
prompt = f"""**重要:您必須使用繁體中文(Traditional Chinese)回覆所有內容。**
|
||
|
||
【專業身份】
|
||
您是中立風險分析師,專注於風險與回報的平衡。
|
||
|
||
【論證重點】
|
||
1. **平衡視角**:權衡上檔與下檔
|
||
2. **條件式建議**:特定條件下的可行性
|
||
3. **情境分析**:不同環境下的表現
|
||
4. **務實建議**:分批進場、對沖
|
||
|
||
【可用資訊】
|
||
- 交易員計畫:{trader_decision}
|
||
- 各類報告:{market_research_report}, {sentiment_report}, {news_report}, {fundamentals_report}
|
||
- 辯論歷史:{history}
|
||
- 對手觀點:{current_risky_response}, {current_safe_response}
|
||
|
||
【輸出要求】
|
||
**長度**:300-500字
|
||
**結構**:
|
||
1. 核心中立論點(80字)
|
||
2. 風險回報平衡分析(150字)
|
||
3. 評論對手觀點(100字)
|
||
4. 投資建議(70字)
|
||
|
||
**注意**:
|
||
- 尋求最佳平衡點
|
||
- 客觀評估雙方
|
||
- 提供穩健策略
|
||
|
||
請提供平衡且客觀的投資論證!"""
|
||
|
||
# 呼叫 LLM 生成回應
|
||
response = llm.invoke(prompt)
|
||
|
||
# 格式化論點
|
||
argument = f"中立分析師:{response.content}"
|
||
|
||
# 更新風險辯論狀態
|
||
new_risk_debate_state = {
|
||
"history": history + "\n" + argument,
|
||
"risky_history": risk_debate_state.get("risky_history", ""),
|
||
"safe_history": risk_debate_state.get("safe_history", ""),
|
||
"neutral_history": neutral_history + "\n" + argument,
|
||
"latest_speaker": "Neutral", # 記錄最新的發言者
|
||
"current_risky_response": risk_debate_state.get(
|
||
"current_risky_response", ""
|
||
),
|
||
"current_safe_response": risk_debate_state.get("current_safe_response", ""),
|
||
"current_neutral_response": argument,
|
||
"count": risk_debate_state["count"] + 1,
|
||
}
|
||
|
||
return {"risk_debate_state": new_risk_debate_state}
|
||
|
||
return neutral_node |