革命性地改变你的YouTube剧本编写：使用LangChain和Streamlit解锁人工智能的力量，创造引人入胜的内容。

介绍：

在当前数字化时代，YouTube 已经成为一个领先的平台，供内容制作者向全世界数百万观众传播他们的思想，娱乐和信息。剧本是制作引人注目的 YouTube 视频的关键组成部分。通过使用写作精良的剧本，您可以保持组织有序并提高内容的整体质量和影响力。在本文中，我们将探讨 LangChain 和 Streamlit 强大的配对如何可能彻底改变您撰写 YouTube 剧本的方式。

了解LangChain:

OpenAI创造了一种复杂的人工智能语言模型，名为LangChain。它基于GPT-3.5架构，使其能够理解和产生类似于人类的文本。当涉及到制作YouTube脚本时，LangChain可能完全改变游戏规则，因为它使得想法、建议甚至整个脚本的撰写变得简单。

代码片段

导入库和包

import os 
import streamlit as st 
from langchain.llms import OpenAI
from langchain.prompts import PromptTemplate
from langchain.chains import LLMChain, SequentialChain 
from langchain.memory import ConversationBufferMemory
from langchain.utilities import WikipediaAPIWrapper 

os.environ['OPENAI_API_KEY'] = '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'

Streamlit 应用框架

st.title('?? YouTube Script Assistant')
st.image('./Youtube.png')
prompt = st.text_input('Plug in your prompt here')

提示模板

title_template = PromptTemplate(
    input_variables=['topic'], 
    template='write me a youtube video title about {topic}'
)

script_template = PromptTemplate(
    input_variables=['title', 'wikipedia_research'], 
    template='write me a youtube video script based on this title TITLE: {title} while leveraging this wikipedia research: {wikipedia_research}'
)

记忆

title_memory = ConversationBufferMemory(input_key='topic', memory_key='chat_history')
script_memory = ConversationBufferMemory(input_key='title', memory_key='chat_history')

WikipediaAPIWrapper（维基百科API包装器）

wiki = WikipediaAPIWrapper()

处理用户输入和生成输出

if prompt: 
    title = title_chain.run(prompt)
    wiki_research = wiki.run(prompt) 
    script = script_chain.run(title=title, wikipedia_research=wiki_research)

    st.write(title) 
    st.write(script) 

    with st.expander('Title History'): 
        st.info(title_memory.buffer)

    with st.expander('Script History'): 
        st.info(script_memory.buffer)

    with st.expander('Wikipedia Research'): 
        st.info(wiki_research)

以下是完整的代码，供您参考：—

# Import libraries & packages
import os 

import streamlit as st 
from langchain.llms import OpenAI
from langchain.prompts import PromptTemplate
from langchain.chains import LLMChain, SequentialChain 
from langchain.memory import ConversationBufferMemory
from langchain.utilities import WikipediaAPIWrapper 

os.environ['OPENAI_API_KEY'] = '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'

# Streamlet App framework
st.title('?? YouTube Script Assistant') # setting teh title 
st.image('./Youtube.png') # set the featured image of the web application 
prompt = st.text_input('Plug in your prompt here')  # The box for the text prompt

# Prompt templates
title_template = PromptTemplate(
    input_variables = ['topic'], 
    template='write me a youtube video title about {topic}'
)

script_template = PromptTemplate(
    input_variables = ['title', 'wikipedia_research'], 
    template='write me a youtube video script based on this title TITLE: {title} while leveraging this wikipedia reserch:{wikipedia_research} '
)

# Memory 
title_memory = ConversationBufferMemory(input_key='topic', memory_key='chat_history')
script_memory = ConversationBufferMemory(input_key='title', memory_key='chat_history')

# Llms
llm = OpenAI(temperature=0.9) 
title_chain = LLMChain(llm=llm, prompt=title_template, verbose=True, output_key='title', memory=title_memory)
script_chain = LLMChain(llm=llm, prompt=script_template, verbose=True, output_key='script', memory=script_memory)

wiki = WikipediaAPIWrapper()

# Show stuff to the screen if there's a prompt
if prompt: 
    title = title_chain.run(prompt)
    wiki_research = wiki.run(prompt) 
    script = script_chain.run(title=title, wikipedia_research=wiki_research)

    st.write(title) 
    st.write(script) 

    with st.expander('Title History'): 
        st.info(title_memory.buffer)

    with st.expander('Script History'): 
        st.info(script_memory.buffer)

    with st.expander('Wikipedia Research'): 
        st.info(wiki_research)