Ema Telegram Bot 🤖

Photo generated using https://firefly.adobe.com

Context:

Why I created Ema?
- In this age of AI, I've crafted a bot that's designed to offer my non-technical loved ones a natural and effortless way to access information quickly. By integrating it right into their chat group, I hope that they'll be able to tap into the power of AI without feeling overwhelmed or intimidated.
Why open source?
- Because I also like open-source projects, and also I hope to see better Ema versions created by people smarter than me
Why I created Ema from scratch instead of using an already existent bot project?
- Because as a nice quote says: "Life is a DIY project"

Ema is a Telegram bot that serves as a helpful female assistant named Ema. Ema is designed to answer all your questions or maintain a conversation. The bot is capable of processing text, voice messages, generating photos, and creating image variations. 📱🔊🖼️

Features 🌟

Receive text messages from Ema 💬
Receive voice messages from Ema 🎙️
Receive generated photos from Ema 📷
Receive image variations from Ema 🎨
Transcribe voice messages into text 📝

Installation 🔧

To use this bot, you need to have the following dependencies installed:

Python 3.10 🐍
telebot
openai
subprocess
langdetect
edge-tts
openai-whisper (this will use your GPU to transcribe the audio)
ffmpeg-python
imgcat
Pillow (Python Imaging Library) 📸

Usage 🚀

Please use conda to run this project. Check https://docs.anaconda.com/free/anaconda/install/ to see how to install conda on your machine.

Clone the repository.
Go into the repository directory and run this 3 commands
- conda env create --file ema.yaml
- conda activate ema_environment
- pip install -r requirements.txt
Go into config.py file
- Replace TELEGRAM_BOT_TOKEN with your TOKEN 🔑. Get a token from here
- Replace OPEN_API_KEY with your KEY. Get your 🔑 from here
- Optionally: Replace any other values. Make it your own.
Run the ema.py file to start the bot. 🏃
Open the Telegram app and start chatting with your new bot. 🔍