It's a very minimal cli prompt, where you can chat and keeping the conversation session memorable by chatgpt and keep chat history for future use.
Assuming you created a env variable with key named OPENAI_API_KEY
.
If you don't have a api key visit here and generate one.
To install ChatGPT-CLI, you'll first need to set up an OpenAI API key. If you don't have one yet, visit the OpenAI Website to sign up and generate an API key.
Next, add the following line to your ~/.bashrc
or ~/.zshrc
file:
export OPENAI_API_KEY=<YOUR OPENAI API KEY>
Be sure to replace with your actual API key. Then, source the file using:
Lastly source the file using source ~/.bashrc
or source ~/.zshrc
You may now install the package using pip:
$ pip3 install gpt-chatbot-cli
For arch users, this exact package is available on aur by the name chatgpt-cli-git
$ paru chatgpt-cli-git
Once you have installed ChatGPT-CLI, you can run it by typing:
$ gpt-chatbot-cli
This will start the CLI prompt, and you can begin chatting with the AI bot.
You can also pass various options to the gpt-chatbot-cli command to customize its behavior. Here are the available options:
-k, --api_key TEXT Openai API key. If not provided, will prompt for it
or use the environment variable OPENAI_API_KEY.
-m, --model TEXT Model to use for text generation | (default:
gpt-3.5-turbo)
-t, --temperature FLOAT Temperature for text generation | (default: 0.9)
-p, --preset TEXT Preset mode to use for text generation | (default:
Chat) Available presets: Chat, Q&A, Grammar
Correction, Eli5, Custom
-hs, --history Show chat history picker | (default: False)
-h, --help Show this message and exit.
ChatGPT-CLI depends on the following Python packages:
- Openai
- prompt-toolkit
- termcolor
- tinyDB
- click
Contributions are welcome! Please read the contributing guidelines first.