langchain-简单DEMO案例

Takake2023-09-182023-12-15

1.langchain简单运用

1.1.基础配置

1	!pip install langchain

1	import langchain

1	!pip install openai

Successfully installed openai-0.28.0

1 2	import os os.environ["OPENAI_API_KEY"] = "sk-OEHX9NSE2GUYSsene6SyT3BlbkFJMFsuXXXXXXXXX"

1	from langchain.llms import OpenAI

1	llm = OpenAI(temperature=0.9)

1 2	text = "What are 5 vacation destinations for someone who likes to eat pasta?" print(llm(text))

1. Rome, Italy
2. Venice, Italy
3. Bologna, Italy
4. Palermo, Sicily
5. Naples, Italy

1	from langchain.prompts import PromptTemplate

prompt = PromptTemplate(
    input_variables=["food"],
    template="What are 5 vacation destinations for someone who likes to eat {food}?"
)

1	print(prompt.format(food="sushi"))

What are 5 vacation destinations for someone who likes to eat sushi?

1	print(llm(prompt.format(food="sushi")))

1. Tokyo, Japan 
2. Osaka, Japan 
3. Sapporo, Japan 
4. Koh Samui, Thailand 
5. Hong Kong, China

1.2. Chains:Combine LLMs and prompts in multi-step workflows

1
2
3

from langchain.prompts import PromptTemplate
from langchain.llms import OpenAI
from langchain.chains import LLMChain

llm = OpenAI(temperature=0.9)
prompt = PromptTemplate(
    input_variables=["food"],
    template = "What are 5 vacation destinations for someone who likes to eat {food}?"
)

1	chain = LLMChain(llm=llm, prompt=prompt)

1	print(chain.run("fruit"))

1. Seychelles
2. Hawaii
3. Costa Rica
4. Thailand
5. The Caribbean Islands

1	!pip install google-search-results

Successfully installed google-search-results-2.4.2

1
2
3

from langchain.agents import load_tools
from langchain.agents import initialize_agent
from langchain.llms import OpenAI

1 2	# load the model llm = OpenAI(temperature=0)

1
2
3

# load in some tools to use
os.environ["SERPAPI_API_KEY"] = "46eca3f76a945e3c452903dea2641511a1c512a786cbb0eXXXXXXXXXXX"
tools = load_tools(["serpapi", "llm-math"], llm=llm)

# Finally let's initalize an agent with
# 1. The tools
# 2. The language model 
# 3. The type of agent we wangt to use
agent = initialize_agent(tools, llm, agent="zero-shot-react-description", verbose=True)

1	agent.run("哪个国家是世界上最强大的国家？现在最流行的大预言模型是什么模型？")

> Entering new AgentExecutor chain…
I need to find out which country is the most powerful and what the most popular big prediction model is.
Action: Search
Action Input: “most powerful country in the world” and “most popular big prediction model”
Observation:”The United States, China and Russia are seen as the world’s most powerful countries based on factors like economics, political influence and military …”]
Thought:3m I need to find out more about the most popular big prediction model
Action: Search
Action Input: “most popular big prediction model”
Observation: ‘Top 5 Predictive Analytics Models · Classification Model · Clustering Model · Forecast Model · Outliers Model · Time Series Model.
Thought:I now know the final answer
Final Answer: 美国、中国和俄罗斯被视为世界上最强大的国家，最流行的大预言模型是分类模型、聚类模型、预测模型、异常值模型和时间序列模型。

Finished chain.

‘美国、中国和俄罗斯被视为世界上最强大的国家，最流行的大预言模型是分类模型、聚类模型、预测模型、异常值模型和时间序列模型。’

1.3. Memory: Add state to chains and agents

1	from langchain import OpenAI, ConversationChain

1 2	llm = OpenAI(temperature=0) conversation = ConversationChain(llm=llm, verbose=True)

1	conversation.predict(input="Hi there!")

> 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 chain.

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

1	conversation.predict(input="What was the first thing i said to you?")

> 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 was the first thing i said to you?
AI:

> Finished chain.

' You said "Hi there!"'