Skip to content

Latest commit

 

History

History
333 lines (249 loc) · 13.2 KB

README.md

File metadata and controls

333 lines (249 loc) · 13.2 KB
Raw

🤖 GPT-Telegramus v5

GPT-Telegramus logo

The best free Telegram bot for ChatGPT, Microsoft Copilot (aka Bing AI / Sidney / EdgeGPT), Microsoft Copilot Designer (aka BingImageCreator), Gemini and Groq with stream writing, requests with images, multiple languages, admin control, data logging and more!

YouTube Bandcamp Spotify SoundCloud

Project banner

🚧 GPT-Telegramus is under development

😔 Currently, GPT-Telegramus doesn't have paid ChatGPT and DALL-E support

📈 GPT-Telegramus v5 is an updated and highly refactored version of the old GPT-Telegramus

📄 Documentation is also under development! Consider reading docstring for now

🐛 If you find a bug in GPT-Telegramus, please create an Issue

p.s. Due to my studies, I don't have much time to work on the project 😔

😋 Support project

💜 Please support the project so that I can continue to develop it

  • BTC: bc1qd2j53p9nplxcx4uyrv322t3mg0t93pz6m5lnft

  • ETH: 0x284E6121362ea1C69528eDEdc309fC8b90fA5578

  • ZEC: t1Jb5tH61zcSTy2QyfsxftUEWHikdSYpPoz

  • Or by my music on 🟦 bandcamp

  • Or message me if you would like to donate in other way 💰

Star History Chart

🤗 Contributors

  • 💜 Sprav04ka - Tofii'skovyi' language, Testing, Super beautiful poster, Project Logo, Motivation
  • 💜 Hanssen - Markdown parsing, bard images, /chat command, caption fix, loading emoji, dynamic splitting, code block splitting, Gemini module, Docker fix, GitHub actions fix and much much more
  • 💜 Sergey Krashevich - Docker, GitHub Actions
  • 💜 Wahit Fitriyanto - Indonesian language
  • 💜 Alexander Fadeyev - EdgeGPT Fix
  • 💜 AnthroAsja - Belarusian language
  • 💜 Anonymous Samurai - Ukrainian language
  • 💜 Dunya Jafari - Persian language
  • 💜 Dedy Rudney - French language

📨 Project channel and demo

⚠️ Please do not overload the bot and use it only as a demo version

🏗️ Requirements

  • Python 3.10 / 3.11 (not tested on other versions)
  • Unblocked access to the telegram bot official API
  • Other requirements specified in the requirements.txt file

📙 Project based on

❓ Get started

See 🐧 Running as service on linux, 🍓 Running on Raspberry Pi (ARM), 🐋 Running in Docker sections for more info

  1. Install Python 3.10 / 3.11 (not tested on other versions), venv and pip
  2. Download source code (clone repo)
  3. Create venv python -m venv venv / python3 -m venv venv / python3.10 -m venv venv / python3.11 -m venv venv
  4. Activate venv source venv/bin/activate / venv\Scripts\activate.bat
  5. Check python version using python --version command
  6. Install requirements pip install -r requirements.txt --upgrade
  7. Carefully change all the settings in config.json file and in each *.json file inside module_configs directory. If you have questions regarding any setting, open an issue, I'll try to add a more detailed description
  8. Run main script python main.py
  • 💬 ChatGPT
    • Free browser-like Chat-GPT. Currently, without extensions and image requests (text only) (because I don't have a paid account to test it)
    • Stream response support
    • Chat history support
    • See https://github.com/F33RNI/LlM-Api-Open for more info
  • 🟦 Microsoft Copilot (aka EdgeGPT aka Bing AI aka Sydney)
    • Supports conversation style /style
    • Stream response support
    • Chat history support
    • Web-browsing (probably) and sources (attributions) support
    • Accepts image requests (for lmao_ms_copilot only)
    • Send generated images (for lmao_ms_copilot only)
    • Suggestions (for lmao_ms_copilot only)
    • See https://github.com/F33RNI/LlM-Api-Open for more info
    • NOTE: Non-LMAO API (the old one ms_copilot) is deprecated! Please use lmao_ms_copilot instead
  • 🎨 Microsoft Copilot Designer (DEPRECATED)
    • Bing Image Generator. Used as a separate module due to issues with the EdgeGPT module
    • Free and unlimited
  • ♊ Gemini
    • Google's AI using the Gemini Pro model
    • Chat history support
    • Requests with images (you can send an image with text to it)
    • Requests with images will not be recorded in the chat history since Google hasn't support this
  • 🔴 Groq
    • Official Python API
    • Chat history support
    • Very fast response
    • Multiple models (see /model command)

🐧 Running as service on linux

  1. Install Python 3.10 / 3.11 (not tested on other versions), venv and pip

  2. Clone repo

    1. git clone https://github.com/F33RNI/GPT-Telegramus.git
    2. cd GPT-Telegramus

  3. Create venv python -m venv venv / python3 -m venv venv / python3.10 -m venv venv / python3.11 -m venv venv

  4. Carefully change all the settings in config.json file and in each *.json file inside module_configs directory

  5. Install systemd

    1. sudo apt-get install -y systemd

  6. Create new service file

    1. sudo nano /etc/systemd/system/gpt-telegramus.service

      [Unit]
Description=GPT-Telegramus service
After=multi-user.target

[Service]
Type=simple
Restart=on-failure
RestartSec=5

WorkingDirectory=YOUR DIRECTORY HERE/GPT-Telegramus
ExecStart=YOUR DIRECTORY HERE/GPT-Telegramus/run.sh

[Install]
WantedBy=multi-user.target

  7. Reload systemctl daemon

    1. sudo systemctl daemon-reload

  8. Enable and start service

    1. sudo systemctl enable gpt-telegramus
    2. sudo systemctl start gpt-telegramus

  9. Note: Please use sudo systemctl kill gpt-telegramus and then sudo systemctl stop gpt-telegramus if only sudo systemctl stop gpt-telegramus not working

🍓 Running on Raspberry Pi (ARM)

  1. Install Python 3.11 or later (not tested) if not installed

    1. sudo apt-get update
sudo apt-get install -y build-essential tk-dev libncurses5-dev libncursesw5-dev libreadline6-dev libdb5.3-dev libgdbm-dev libsqlite3-dev libssl-dev libbz2-dev libexpat1-dev liblzma-dev zlib1g-dev libffi-dev
wget https://www.python.org/ftp/python/3.11.8/Python-3.11.8.tgz
sudo tar zxf Python-3.11.8.tgz
cd Python-3.11.8
sudo ./configure --enable-optimizations
sudo make -j 4
sudo make altinstall

    2. Check version by typing python3.11 -V. After this, you should use python3.11 command instead of python or you can add it to the bashrc by typing echo "alias python=/usr/local/bin/python3.11" >> ~/.bashrc

  2. Follow the 🐧 Running as service on linux guide

🐋 Running in Docker

From GitHub Package

  1. Clone repo or download config.json and module_configs and langs

  2. Edit the config.json, set options in the files section to the path in the container (/app/config/<FILE_NAME>)

  3. Run the container

    docker run -d -e TELEGRAMUS_CONFIG_FILE="/app/config/config.json" -v <YOUR_CONFIG_FOLDER>:/app/config --name gpt-telegramus --restart on-failure ghcr.io/f33rni/gpt-telegramus:latest

    If you want to try the preview version

    docker run -d -e TELEGRAMUS_CONFIG_FILE="/app/config/config.json" -v <YOUR_CONFIG_FOLDER>:/app/config --name gpt-telegramus --restart on-failure ghcr.io/f33rni/gpt-telegramus:edge

Build Manually

  1. Install Docker

  2. Clone repo

  3. Build container

    docker buildx build -t telegramus --load -f Dockerfile .

  4. Run the container

    docker run -d --name gpt-telegramus --restart on-failure telegramus

    or if you want to use a custom config

    docker run -d -e TELEGRAMUS_CONFIG_FILE="/app/config/config.json" -v <YOUR_CONFIG_FOLDER>:/app/config --name gpt-telegramus --restart on-failure telegramus

🌐 Bot messages

Currently available languages

  • 🇺🇸 English
  • 🇷🇺 Русский
  • ‍☠️ Тофийсковый
  • 🇮🇩 Bahasa Indonesia
  • 🇨🇳 简体中文
  • 🇧🇾 Беларуская
  • 🇺🇦 Українська
  • فارسی 🇮🇷
  • 🇪🇸 Español
  • 🇻🇳 Vietnamese
  • 🇫🇷 Français

You can add a new language. For that:

  1. Copy any existing language file (inside langs directory)
  2. Rename it according to Set2/T (3-letters code) ISO 639 https://en.wikipedia.org/wiki/List_of_ISO_639_language_codes
  3. Translate each entry
  4. Create a pull request 💜

You can add new lines by adding \n

⚠️ Please make sure you haven't messed up the string formatting structure {this_type}

🤖 Telegram bot commands

  • 📄 /start - Welcome message and bot version
  • ❓ /help - Show help message
  • ↕️ /module - Change module to chat with
  • 🧹 /clear - Clear chat history
  • 🌎 /lang - Change the language
  • 🆔 /chatid - Show your chat_id
  • ⚙️ /model - Change model of module (Currently for 🔴 Groq module only)
  • /style - Bing AI conversation style
  • /chat - Send request in group chat
  • Other direct module commands (please check automatically-generated list of commands inside bot)

Admin commands

  • 💬 /queue - Show requests queue
  • 🔃 /restart [module name, optional] - Restart specific module (and it's config) or every module and all configs, languages and bot commands
    • Please see bot_command_restart function in bot_handler.py file for more info
  • 👤 /users - Show list of all users
  • 🔨 /ban <id> [reason] - Ban a user by their id with reason (optional)
  • 🔓 /unban <id> - Unban a user by their id
  • 📢 /broadcast <message> - Send text message to everyone except banned users

📜 Data collecting

GPT-Telegramus has a built-in data collecting function (saves requests and responses in a files)

  • For text requests / responses will be saved as plain text
  • For image requests / responses will be saved as Base64-encoded image (in the same text file)

You can enable and configure data collection in config.json in data_collecting section

⚠️ Please make sure you notify your bot users that you're collecting data

📝 TODO

  • Paid version of ChatGPT
  • DALL-E (from OpenAI)
  • Some other LLMs (and maybe some free GPT-4 model)

✨ Contribution

  • Anyone can contribute! Just create a pull request