Skip to content

Latest commit

 

History

History
55 lines (37 loc) · 1.53 KB

README.md

File metadata and controls

55 lines (37 loc) · 1.53 KB
Raw

chatgpt_ros

chatgpt_ros_x264.mp4

Quick Start

First you need to get your api key from openai and set it as an environment variable.

export OPENAI_API_KEY=<your api key>

After that you can start the action server and thus establish communication with chatgpt.

ros2 run chatgpt_ros chatgpt_action_server

This command starts a node named /chatgpt_action_server, which provides the following services and actions.

Services:

  • /reset (std_srvs/srv/Empty): Reset chat history.
  • /set_prompt (chatgpt_ros_interfaces/srv/SetPrompt): Set system prompt.
  • /add_example (chatgpt_ros_interfaces/srv/AddExample): Add example.

Actions:

  • /chat (chatgpt_ros_interfaces/action/Chat): Chat with ChatGPT.

This is the description of the /chat action.

string messages
---
string content
---
string delta

The goal (messages) of the chat action is the user's input. The response from chatgpt will keep coming back from feedback (delta). When the chat ends, a full reply (content) is returned.

You can start an action client to test the /chat action.

ros2 run chatgpt_ros chatgpt_action_client

Turtlesim Demo

We also provide a demo on how to control the turtlesim in ros using chatgpt. You can start the demo with the following command:

ros2 launch chatgpt_ros_turtlesim demo.py

This will pop up a separate command window for entering commands to control the turtle. You can instruct the turtles to draw graphics through natural language.