ChatGPT + Raspberry Pi - A ChatGPT powered Raspberry Pi chatbot

A beginner-friendly way to combine Chat GPT & Raspberry Pi.

0:00 Building
5:36 Demo Chatbot
Guide:

Terminal commands:

1. Installing libraries/dependencies

sudo apt update
sudo apt upgrade

python3 -m pip install python-dotenv
pip3 install openai dotenv SpeechRecognition pyttsx3 gtts PyAudio
sudo apt install python3-pyaudio flac python3-espeak espeak python3-dotenv
pip uninstall dotenv
pip install python-dotenv

*Change audio to HDMI

2. Python script

Python script:

import os
import openai
from dotenv import load_dotenv
import time
import speech_recognition as sr
import pyttsx3
import numpy as np
from gtts import gTTS

mytext = 'Welcome to me'
language = 'en'
# ^ matplotlib is great for visualising data and for testing purposes but usually not needed for production
load_dotenv()
model = 'gpt-3.5-turbo'
# Set up the speech recognition and text-to-speech engines
r = sr.Recognizer()
name = "YOUR NAME HERE"
greetings = [f"whats up master {name}",
"yeah?",
"Well, hello there, Master of Puns and Jokes - how's it going today?",
f"Ahoy there, Captain {name}! How's the ship sailing?",
f"Bonjour, Monsieur {name}! Comment ça va? Wait, why the hell am I speaking French?" ]

# Listen for the wake word "hey pos"
def listen_for_wake_word(source):
print("Listening for 'Hey'...")

while True:
try:
print("Wake word detected.")
listen_and_respond(source)
break
except sr.UnknownValueError:
pass
# Listen for input and respond with OpenAI API
def listen_and_respond(source):
print("Listening...")

while True:
try:
print(f"You said: {text}")
if not text:
continue

# Send input to OpenAI API
print(response_text)
#myobj = gTTS(text = response_text, lang = language, slow = False)
# Speak the response
print("speaking")

if not audio:
listen_for_wake_word(source)
except sr.UnknownValueError:
print("Silence found, shutting up, listening...")
listen_for_wake_word(source)
break

except sr.RequestError as e:
print(f"Could not request results; {e}")
listen_for_wake_word(source)
break

# Use the default microphone as the audio source
with sr.Microphone() as source:
listen_for_wake_word(source)

[geoffreygeo]

I tried to install all the libraries. Only with PyAudio did this go wrong. Eventually, I got it to work. By the following commands:

sudo apt install libasound-dev portaudio19-dev
libportaudio2 libportaudiocpp0

After that sudo apt install ffmpeg and than run

python3 -m pip install pyaudio

Everything works great

[alexpapou]

Just came across, a totally amazing project. Perhaps the ultimate stage of this would be a bot that activates only on a certain voice command so what you chat in a room does not get sent to gpt unless you ask for this, and a real non-robotic voice of your choice. I am impressed, thank you!

[JohnDoe-ogkz]

Amazing work dude. Also congrats on accomplishing 100 subs. Have a great day!
- your 100th subscriber

[Skibidioo]

Thanks this worked well i can now use chatgpt as a voice command to turn on the laser

[lzcell]

The first time I saw this video, I was intrigued, but now ChatGPT has its own speech conversation function.

[WINTERMUTE_AI]

GENIUS! This is exactly what I was looking for, thanks!

[charlesmarchand]

ill start my project with yours as guidance to make my own, awesome and simple. Thank you!

[betrayedsoul]

Thx this is so cool your 1000th subscriber

[jtr]

Great ChatGPT chatbot! Awesome! Hi Thomas I have a Raspberry PI 5, a ollama with tinyollama setup on my system with an ollama webui as frontend. It works and the responses are good enough as far as speed is concerned. I already installed piper and the voice sounds natural and perhaps add whisper next for an all out local project. Do you know a way on how I can this all integrated together. I'll be looking forward to your next video.🌟🌟👍

[mercy]

I have been thinking of this for a few days now ! Thanks Thomas !!

[jacquesdupontd]

Perfect, now it needs to be also a media player for spotify and other stream and be able to control home assistant devices :) Thanks

[pilqpj]

thank you it very interesting to see somone of your age do a project like this

[davyponte]

That’s amazing! Bravo👏🏻 congrats… but is possible change the chatbot voice? For example my voice?

[Kataroobungaloo]

Wow incredible work here!! Very inspiring!!!

[drewski]

Wow, thanks for this video man.
I was windering, once installed, how can you install this on a rc car?
Have the system roam around with comouter vision, lidar, and other systems for obstacle avoidance, roaming around my property?
Thanks again for this informative video.

[markboweringphotography]

Superb project. Just what i was looking for.

[TheCryptoCollective]

When I copy and paste the libraries/dependencies code I get a error then I go into a virtual environment then do it again it works but then I get more errors stating that it can't find dotenv idk what I'm doing wrong but I followed this video to a T and nothing's working

[elvenkim]

most interesting project! didn't know raspberry pi can run chatgpt

[johnmatheson]

this is awesome ive been trying to do the same project over the last week. i set up your code to use as an example for mine, but its running slow do you have any suggestions on how to speed it up? im running it on a rasp pi 4 8gb btw.

[carterkraber]

Not running well, it keeps saying command not found when I type in python 3. gpt.py and I just can’t get it to run