Langchain을 활용한 인공지능 영어과외 업그레이드

import json

history = [] # 대화 내용을 저장하는 메모리

def talk_to_gpt(user_input):
  gpt_response = get_gpt_response(user_input, history[-10:])
  gpt_response = json.loads(gpt_response)
  gpt_response = gpt_response['response']
  history.extend([
      {"role": "user", "content": user_input}, 
      {"role": "assistant", "content": gpt_response}
  ])

# Now we can override it and set it to "AI Assistant"
from langchain.prompts.prompt import PromptTemplate

template = """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!")

prompt_template = """
The following is a friendly conversation between a human and an AI.
The AI a top-tier English tutor with years of experience.
The AI is talking to a student who wants to practice speaking English. 
The AI is to help the student practice speaking English by having a conversation. 

The AI should feel free to correct the student's grammar and pronunciation and/or suggest different words or phrases to use whenever the AI feels needed.
And when the AI corrects the student, the AI must start the sentence with "it is better to put it this way"
But even when you correct the student, try to make a conversation first, and then correct the student

Current conversation:
{history}
Human: {input}
AI Tutor:"""

# Import necessary classes from the langchain and langchain_openai libraries
from langchain.chains import ConversationChain
from langchain.memory import ConversationBufferMemory
from langchain.prompts.prompt import PromptTemplate
from langchain_openai import OpenAI

# Initialize the OpenAI model with a specified temperature.
# Temperature set to 0 for deterministic, consistent responses.
llm = OpenAI(
    temperature=0
)

# Create a conversation buffer memory to keep track of the conversation.
# This includes prefixes to distinguish between the AI tutor and the human user.
memory = ConversationBufferMemory(
    ai_prefix="AI Tutor:",
    human_prefix="Human:",
)

# Function to get and configure the conversation chain
def get_chain():
    # Define a template for the conversation prompt.
    # This template sets the context for the conversation and instructions for the AI.
    ...
    # Create a PromptTemplate object with the defined prompt template.
    # This template includes variables for the conversation history and the latest human input.
    conversation_prompt = PromptTemplate(input_variables=["history", "input"], template=prompt_template)

    # Initialize the conversation chain.
    # This chain uses the defined prompt, the language model (llm), and the conversation memory.
    conversation_chain = ConversationChain(
        prompt=conversation_prompt,
        llm=llm,
        verbose=True,
        memory=memory,
    )
    
    # Return the configured conversation chain.
    return conversation_chain

def get_gpt_response(transcript):
    # Talk to the AI Tutor via langchain 
    conversation = get_chain()
    answer = conversation.predict(input=transcript)
    
    # Return the AI's message content
    return answer