Advanced LangChain: Chains, Memory, and Agents

Advanced LangChain Patterns

Custom Chains

from langchain.chains import LLMChain
from langchain_openai import ChatOpenAI
from langchain.prompts import PromptTemplate

# Sequential chain
from langchain.chains import SequentialChain

chain1 = LLMChain(
    llm=ChatOpenAI(),
    prompt=PromptTemplate.from_template("Summarize: {text}"),
    output_key="summary"
)

chain2 = LLMChain(
    llm=ChatOpenAI(),
    prompt=PromptTemplate.from_template("Translate to Spanish: {summary}"),
    output_key="translation"
)

overall_chain = SequentialChain(
    chains=[chain1, chain2],
    input_variables=["text"],
    output_variables=["summary", "translation"]
)

Memory Types

# Conversation buffer memory
from langchain.memory import ConversationBufferMemory

memory = ConversationBufferMemory()
memory.save_context({"input": "hi"}, {"output": "Hello!"})

# Conversation summary memory
from langchain.memory import ConversationSummaryMemory

memory = ConversationSummaryMemory(llm=ChatOpenAI())

# Vector store memory
from langchain.memory import VectorStoreRetrieverMemory

memory = VectorStoreRetrieverMemory(
    retriever=vectorstore.as_retriever(search_kwargs={"k": 3})
)

Custom Agents

from langchain.agents import Tool, AgentExecutor, create_react_agent

tools = [
    Tool(
        name="Search",
        func=search.run,
        description="Search for information"
    )
]

agent = create_react_agent(llm, tools, prompt)
agent_executor = AgentExecutor(agent=agent, tools=tools)

---

Found an issue? Open an issue!