LLM及LangChain開發筆記(14)_對話記憶（Conversational Memory）：介紹與簡介

 

記憶機制簡介

• ConversationBufferMemory :儲存完整的對話歷史
• ConversationSummaryBufferMemory :使用摘要方式儲存對話歷史
• ConversationBufferWindowMemory :僅保留最後幾輪對話
• ConversationalTokenBufferMemory :限制儲存的 Token 數量

範例程式

import openai
import os

from dotenv import load_dotenv, find_dotenv
_ = load_dotenv(find_dotenv())  # read local .env file

openai.api_key = os.getenv('OPENAI_API_KEY')
from langchain.chat_models import ChatOpenAI
chat = ChatOpenAI(model_name="gpt-3.5-turbo", temperature=0.0)

from langchain.callbacks import get_openai_callback

def count_tokens(chain, query):
    with get_openai_callback() as cb:
        result = chain.run(query)
        print(f'Spent a total of {cb.total_tokens} tokens')

    return result


from langchain.chains import ConversationChain

conversation_buf = ConversationChain(
    llm=chat, 
)

print(conversation_buf.prompt.template)

from langchain.memory import ConversationBufferMemory

memory=ConversationBufferMemory()

conversation_buf = ConversationChain(
    llm=chat,
    memory=memory
)
conversation_buf("早上好 AI！")
conversation_buf("我叫 Sara，你的名字是什麼！")
conversation_buf("你好，你還記得我叫什麼嗎？")