node-red-contrib-eksido-chatgpt

A Node-RED module that integrates with OpenAI's GPT-3 API and OpenAI Chat API.

This module includes two nodes:

• chatgpt: A node that interacts with GPT-3 API for generating text completions.
• openai-chat: A node that interacts with OpenAI Chat API for multi-turn conversations.

Installation

Run the following command in your Node-RED user directory (typically ~/.node-red):

npm install node-red-contrib-eksido-chat

Usage

chatgpt Node

1. Add the chatgpt node to your Node-RED flow.
2. Double-click on the node to configure the required parameters:
   • API Key: Your OpenAI API key.
   • Max Tokens: The maximum number of tokens (words) in the generated text.
   • Num Responses: The number of responses to generate.
   • Stop Sequences: The character sequences to stop generating text.
   • Temperature: Controls the randomness of the generated text.
3. Connect the input and output of the chatgpt node to other nodes as needed.
4. Deploy your flow.

openai-chat Node

1. Add the openai-chat node to your Node-RED flow 2. Double-click on the node to configure the required parameters:
   • API Key: Your OpenAI API key.
   • Model: The model to use for generating responses (e.g., "gpt-3.5-turbo").
2. Connect the input and output of the openai-chat node to other nodes as needed.
3. Deploy your flow.

Input

For both nodes, the input message payload should be a string.

chatgpt Node

• The input message payload is the prompt for the GPT-3 API.

openai-chat Node

• The input message payload is the content of the user message to the Chat API.

Output

The output message payload will be an array of generated text for the chatgpt node and an array of generated messages for the openai-chat node. You can use other nodes in your flow to process these outputs as needed.

Example

Here's a simple example of using the chatgpt node in a flow:

1. Add an inject node with a string payload (e.g., "Translate the following English text to French: 'Hello, how are you?'").
2. Connect the inject node to a chatgpt node configured with your OpenAI API key and desired parameters.
3. Connect the chatgpt node to a debug node to display the generated text.
4. Deploy your flow and click on the inject node to trigger the GPT-3 API request.

For the openai-chat node, the flow would be similar, with the input message payload containing the content of the user message for the Chat API.

1. Add an inject node with a string payload (e.g., "Tell me a joke.").
2. Connect the inject node to an openai-chat node configured with your OpenAI API key and desired model.
3. Connect the openai-chat node to a debug node to display the generated chat messages.
4. Deploy your flow and click on the inject node to trigger the OpenAI Chat API request.

License

MIT License