GO Whisper Telegram Bot

This project is a Dockerized Telegram bot written in Go, designed to handle speech transcription and synthesis via a Python backend. The backend code is directly included in this repository, and is based on components from two of my earlier projects.

---

🔁 Code Reuse Acknowledgement

The Python backend in this repo reuses and adapts functionality from:

• Flutter_Python_Speech_App
• TG_WebApp_Flutter_WCTTS

These repositories provided the foundation.

---

🧠 Features

• Telegram bot built with Go (go-telegram-bot-api)
• Two main actions triggered via buttons:
  • Transcribe: Upload a small audio or video file and receive a transcript
  • Synthesize: Upload a voice sample, enter text, choose output language, receive speech audio
• Python backend handles:
  • File validation
  • Whisper transcription
  • Language detection and voice cloning
• Dockerized and runnable via docker-compose

---

⚙️ Environment Setup

📁 frontend/bot/.env

Set your Telegram bot token:

TELEGRAM_BOT_TOKEN=your_token_here

---

📁 backend/.env

Configure backend settings:

PORT=5000
MAX_FILE_MB=20
MAX_FILE_DURATION_SEC=360
CORS_ORIGIN=http://localhost:8080

---

🚀 Running the Bot

Prerequisites

• Docker
• Docker Compose

Start the system:

docker-compose up --build

---