- The following Python IDE program is required for test, download and install from below links.
- Deploying firmware to the device. you can use the below firmware.
If you need a detailed explanation of firmware and Thonny setup, please refer to the video below.
- Combine WIZnet Ethernet HAT with Raspberry Pi Pico.
- Connect ethernet cable to Ethernet HAT ethernet port.
- Connect Raspberry Pi Pico to desktop or laptop using 5 pin micro USB cable.
If you use W5100S-EVB-Pico, you can skip '1. Combine...'
To test the Open AI example, minor settings shall be done in code.
- Check COMport in Device Manager and then open Thonny Python IDE.
- Setup SPI and Reset pin in chat_with_openai.py or send_prompt.py file.
If you are not using the WIZnet PICO Board, please adjust the SPI pin configuration according to your hardware setup.
spi=SPI(0,2_000_000, mosi=Pin(19),miso=Pin(16),sck=Pin(18))
- Enter OpenAI information(version, API key) into the chatGPT.py file.
If you have not generated an API key, please generate one on the following site https://platform.openai.com/account/api-keys
api_key = "insert your API-KEY"
chatgpt_url = "https://api.openai.com/v1/chat/completions"
chatgpt_ver= "gpt-3.5-turbo"
The function 'send_prompt_to_chatGPT' sends the prompt to openAI and receives a response from the model. If the response is successful, the generated text is printed serial program.
Enter the prompt as a parameter for the [send_prompt_to_chatGPT] function.
send_prompt_to_chatGPT("Hello")
The function 'chat_with_chatGPT' enables communication with the chatGPT model using the OpenAI API via a serial program. It user enter a message and sends it to OpenAI as a prompt, then prints the response.
The default exit command is ">exit". To modify the exit command, please modify the relevant section in the chatGPT.py file.
end_command= ">exit"