Building with LangChain

Introduction

LangChain is the most popular framework for developing applications powered by language models. It provides standard interfaces for chains, agents, memory, and more, making it much easier to build complex AI systems.

Chains vs. Agents

In LangChain, a Chain is a sequence of hardcoded steps. For example: Prompt -> LLM -> Output Parser. It's predictable but rigid.

An Agent, however, uses the LLM to determine which actions to take and in what order. If a step fails, the agent can dynamically try another approach.

from langchain.agents import initialize_agent, Tool
from langchain.llms import OpenAI

# Define tools the agent can use
tools = [
    Tool(
        name="Calculator",
        func=calculator_function,
        description="Useful for when you need to answer questions about math"
    )
]

llm = OpenAI(temperature=0)
agent = initialize_agent(tools, llm, agent="zero-shot-react-description")
agent.run("What is 15 multiplied by 4?")

Assignment