Agents - Make OpenAI Do Things For you

Agents - Make OpenAI Do Things For you#

Source

  • Agent - 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.

Parameters when creating an agent:

  • Tool: A function that performs a specific duty. This can be things like: Google Search, Database lookup, Python REPL, other chains. The interface for a tool is currently a function that is expected to have a string as an input, with a string as an output.

  • LLM: The language model powering the agent.

  • Agent: The agent to use. This should be a string that references a support agent class. Because this notebook focuses on the simplest, highest level API, this only covers using the standard supported agents. If you want to implement a custom agent, see the documentation for custom agents (coming soon).

# Unzip data folder

import zipfile
with zipfile.ZipFile('../../data.zip', 'r') as zip_ref:
    zip_ref.extractall('..')

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

List of Tools#

  1. python_repl - A Python shell. Use this to execute python commands. Input should be a valid python command. If you expect output it should be printed out.

  2. serpapi [Complete] - A search engine. Useful for when you need to answer questions about current events. Input should be a search query.

  3. wolfram-alpha [Complete] - A wolfram alpha search engine. Useful for when you need to answer questions about Math, Science, Technology, Culture, Society and Everyday Life. Input should be a search query.

  4. requests - A portal to the internet. Use this when you need to get specific content from a site. Input should be a specific url, and the output will be all the text on that page.

  5. terminal - Executes commands in a terminal. Input should be valid commands, and the output will be any output from running that command.

  6. pal-math - A language model that is excellent at solving complex word math problems. Input should be a fully worded hard word math problem.

  7. pal-colored-objects - A language model that is wonderful at reasoning about position and the color attributes of objects. Input should be a fully worded hard reasoning problem. Make sure to include all information about the objects AND the final question you want to answer.

  8. llm-math - Useful for when you need to answer questions about math.

  9. open-meteo-api - Useful for when you want to get weather information from the OpenMeteo API. The input should be a question in natural language that this API can answer.

  10. news-api - Use this when you want to get information about the top headlines of current news stories. The input should be a question in natural language that this API can answer.

  11. tmdb-api - Useful for when you want to get information from The Movie Database. The input should be a question in natural language that this API can answer.

  12. google-search - A wrapper around Google Search. Useful for when you need to answer questions about current events. Input should be a search query.

2. SERP API#

from langchain.agents import load_tools

# import os
# os.environ['OPENAI_API_KEY'] = "..."
# os.environ['SERPAPI_API_KEY'] = "..."

llm = OpenAI(temperature=0)

tool_names = ["serpapi"]
tools = load_tools(tool_names)

agent = initialize_agent(tools, llm, agent="zero-shot-react-description", verbose=True)

agent.run("What is LangChain?")

'LangChain is a platform that provides a standard interface for chains, lots of integrations with other tools, and end-to-end chains for common applications. Data Augmented Reality (DAR) is also supported.'

# Input should be a search query.
agent.run("who is the ceo of pipe?")

'Luke Voiles is the CEO of Pipe.'

3. Wolfram Alpha#

from langchain.agents import load_tools

# import os
# pip install wolframalpha
# os.environ['OPENAI_API_KEY'] = "..."
# os.environ['WOLFRAM_ALPHA_APPID'] = ".."

llm = OpenAI(temperature=0)

tool_names = ["wolfram-alpha"]
tools = load_tools(tool_names)

agent = initialize_agent(tools, llm, agent="zero-shot-react-description", verbose=True)

# Input should be a search query.

agent.run("What is the asthenosphere?")

'The asthenosphere is the lower layer of the crust.'
Contents