langchain/docs/explanation/agents.md
2022-12-18 21:51:23 -05:00

1.3 KiB

Agents

Agents use an LLM to determine which actions to take and in what order. An action can either be using a tool and observing its output, or returning to the user. For a list of easily loadable tools, see here. Here are the agents available in LangChain.

For a tutorial on how to load agents, see here.

zero-shot-react-description

This agent uses the ReAct framework to determine which tool to use based solely on the tool's description. Any number of tools can be provided. This agent requires that a description is provided for each tool.

react-docstore

This agent uses the ReAct framework to interact with a docstore. Two tools must be provided: a Search tool and a Lookup tool (they must be named exactly as so). The Search tool should search for a document, while the Lookup tool should lookup a term in the most recently found document. This agent is equivalent to the original ReAct paper, specifically the Wikipedia example.

This agent utilizes a single tool that should be named Intermediate Answer. This tool should be able to lookup factual answers to questions. This agent is equivalent to the original self ask with search paper, where a Google search API was provided as the tool.