自然語言處理從入門到應用——LangChain：記憶（Memory）-[自定義對話記憶與自定義記憶類]

分類目錄：《自然語言處理從入門到應用》總目錄

自定義對話記憶

本節介紹了幾種自定義對話記憶的方法：

from langchain.llms import OpenAI
from langchain.chains import ConversationChain
from langchain.memory import ConversationBufferMemoryllm = OpenAI(temperature=0)

AI前綴

第一種方法是通過更改對話摘要中的AI前綴來實現。默認情況下，它設置為AI，但你可以將其設置為任何你想要的內容。需要注意的是，如果我們更改了這個前綴，我們還應該相應地更改鏈條中使用的提示來反映這個命名更改。讓我們通過下面的示例來演示這個過程。

# Here it is by default set to "AI"
conversation = ConversationChain(llm=llm, verbose=True, memory=ConversationBufferMemory()
)
conversation.predict(input="Hi there!")

日志輸出：

> Entering new ConversationChain chain...
Prompt after formatting:
The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know.Current conversation:Human: Hi there!
AI:> Finished ConversationChain chain.

輸出：

" Hi there! It's nice to meet you. How can I help you today?"

輸入：

conversation.predict(input="What's the weather?")

日志輸出：

> Entering new ConversationChain chain...
Prompt after formatting:
The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know.Current conversation:Human: Hi there!
AI:  Hi there! It's nice to meet you. How can I help you today?
Human: What's the weather?
AI:> Finished ConversationChain chain.

輸出：

' The current weather is sunny and warm with a temperature of 75 degrees Fahrenheit. The forecast for the next few days is sunny with temperatures in the mid-70s.'

輸入：

# Now we can override it and set it to "AI Assistant"
from langchain.prompts.prompt import PromptTemplatetemplate = """The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know.Current conversation:
{history}
Human: {input}
AI Assistant:"""
PROMPT = PromptTemplate(input_variables=["history", "input"], template=template
)
conversation = ConversationChain(prompt=PROMPT,llm=llm, verbose=True, memory=ConversationBufferMemory(ai_prefix="AI Assistant")
)
conversation.predict(input="Hi there!")

日志輸出：

> Entering new ConversationChain chain...
Prompt after formatting:
The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know.Current conversation:Human: Hi there!
AI Assistant:> Finished ConversationChain chain.
" Hi there! It's nice to meet you. How can I help you today?"
conversation.predict(input="What's the weather?")
> Entering new ConversationChain chain...
Prompt after formatting:
The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know.Current conversation:Human: Hi there!
AI Assistant:  Hi there! It's nice to meet you. How can I help you today?
Human: What's the weather?
AI Assistant:> Finished ConversationChain chain.

輸出：

The current weather is sunny and warm with a temperature of 75 degrees Fahrenheit. The forecast for the rest of the day is sunny with a high of 78 degrees and a low of 65 degrees.'

人類前綴

第二種方法是通過更改對話摘要中的人類前綴來實現。默認情況下，它設置為Human，但我們可以將其設置為任何我們想要的內容。需要注意的是，如果我們更改了這個前綴，我們還應該相應地更改鏈條中使用的提示來反映這個命名更改。讓我們通過下面的示例來演示這個過程。

# Now we can override it and set it to "Friend"
from langchain.prompts.prompt import PromptTemplatetemplate = """The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know.Current conversation:
{history}
Friend: {input}
AI:"""
PROMPT = PromptTemplate(input_variables=["history", "input"], template=template
)
conversation = ConversationChain(prompt=PROMPT,llm=llm, verbose=True, memory=ConversationBufferMemory(human_prefix="Friend")
)
conversation.predict(input="Hi there!")

日志輸出：

> Entering new ConversationChain chain...
Prompt after formatting:
The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know.Current conversation:Friend: Hi there!
AI:> Finished ConversationChain chain.
" Hi there! It's nice to meet you. How can I help you today?"
conversation.predict(input="What's the weather?")
> Entering new ConversationChain chain...
Prompt after formatting:
The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know.Current conversation:Friend: Hi there!
AI:  Hi there! It's nice to meet you. How can I help you today?
Friend: What's the weather?
AI:> Finished ConversationChain chain.

輸出：

  ' The weather right now is sunny and warm with a temperature of 75 degrees Fahrenheit. The forecast for the rest of the day is mostly sunny with a high of 82 degrees.'

創建自定義記憶類

盡管在LangChain中有幾種預定義的記憶類型，但我們很可能希望添加自己的記憶類型，以使其適用于我們的應用程序。在本節中，我們將向ConversationChain添加一個自定義的記憶類型。為了添加自定義的記憶類，我們需要導入基本的記憶類并對其進行子類化。

from langchain import OpenAI, ConversationChain
from langchain.schema import BaseMemory
from pydantic import BaseModel
from typing import List, Dict, Any

在這個示例中，我們將編寫一個自定義的記憶類，使用spacy提取實體并將有關它們的信息保存在一個簡單的哈希表中。然后，在對話過程中，我們將查看輸入文本，提取任何實體，并將關于它們的任何信息放入上下文中。需要注意的是，這種實現相當簡單且脆弱，可能在生產環境中不太有用。它的目的是展示我們可以添加自定義的記憶實現。為此，我們需要首先安裝spacy。

# !pip install spacy
# !python -m spacy download en_core_web_lg
import spacy
nlp = spacy.load('en_core_web_lg')
class SpacyEntityMemory(BaseMemory, BaseModel):"""Memory class for storing information about entities."""# Define dictionary to store information about entities.entities: dict = {}# Define key to pass information about entities into prompt.memory_key: str = "entities"def clear(self):self.entities = {}@propertydef memory_variables(self) -> List[str]:"""Define the variables we are providing to the prompt."""return [self.memory_key]def load_memory_variables(self, inputs: Dict[str, Any]) -> Dict[str, str]:"""Load the memory variables, in this case the entity key."""# Get the input text and run through spacydoc = nlp(inputs[list(inputs.keys())[0]])# Extract known information about entities, if they exist.entities = [self.entities[str(ent)] for ent in doc.ents if str(ent) in self.entities]# Return combined information about entities to put into context.return {self.memory_key: "\n".join(entities)}def save_context(self, inputs: Dict[str, Any], outputs: Dict[str, str]) -> None:"""Save context from this conversation to buffer."""# Get the input text and run through spacytext = inputs[list(inputs.keys())[0]]doc = nlp(text)# For each entity that was mentioned, save this information to the dictionary.for ent in doc.ents:ent_str = str(ent)if ent_str in self.entities:self.entities[ent_str] += f"\n{text}"else:self.entities[ent_str] = text

我們現在定義一個提示，其中包含有關實體的信息以及用戶的輸入：

from langchain.prompts.prompt import PromptTemplatetemplate = """The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know. You are provided with information about entities the Human mentions, if relevant.Relevant entity information:
{entities}Conversation:
Human: {input}
AI:"""
prompt = PromptTemplate(input_variables=["entities", "input"], template=template
)

現在，我們把它們整合起來：

llm = OpenAI(temperature=0)
conversation = ConversationChain(llm=llm, prompt=prompt, verbose=True, memory=SpacyEntityMemory())

在第一個例子中，由于對Harrison沒有先前的了解，"Relevant entity information"部分是空的：

conversation.predict(input="Harrison likes machine learning")

日志輸出：

> Entering new ConversationChain chain...
Prompt after formatting:
The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know. You are provided with information about entities the Human mentions, if relevant.Relevant entity information:Conversation:
Human: Harrison likes machine learning
AI:> Finished ConversationChain chain.

輸出：

" That's great to hear! Machine learning is a fascinating field of study. It involves using algorithms to analyze data and make predictions. Have you ever studied machine learning, Harrison?"

現在在第二個例子中，我們可以看到它提取了關于Harrison的信息。

conversation.predict(input="What do you think Harrison's favorite subject in college was?")

日志輸出：

> Entering new ConversationChain chain...
Prompt after formatting:
The following is a friendly conversation between a human and an AI. The AI is talkative and provides lots of specific details from its context. If the AI does not know the answer to a question, it truthfully says it does not know. You are provided with information about entities the Human mentions, if relevant.Relevant entity information:
Harrison likes machine learningConversation:
Human: What do you think Harrison's favorite subject in college was?
AI:> Finished ConversationChain chain.

輸出：

' From what I know about Harrison, I believe his favorite subject in college was machine learning. He has expressed a strong interest in the subject and has mentioned it often.'

這個實現方式相對簡單且容易出錯，可能在實際生產環境中沒有太大的用途，但它展示了我們可以添加自定義的內存實現方式。

參考文獻：
[1] LangChain官方網站：https://www.langchain.com/
[2] LangChain 🦜?🔗 中文網，跟著LangChain一起學LLM/GPT開發：https://www.langchain.com.cn/
[3] LangChain中文網 - LangChain 是一個用于開發由語言模型驅動的應用程序的框架：http://www.cnlangchain.com/