🎙 English ↔ Tamil Speech Translator (Ollama LLM)

A real-time speech translation application that converts spoken English to Tamil and vice versa, powered by Ollama's local LLM (Llama 3.2) for translation, with text-to-speech playback functionality.

Features

• Speak in English or Tamil and get instant translation
• Local LLM-powered translation using Ollama (Llama 3.2)
• Automatic bidirectional translation (English ↔ Tamil)
• Listen to the translated speech output
• Save Translations - Store your translations to a local file
• Clean, modern UI using CustomTkinter
• Threaded processing for smooth performance
• Privacy-focused - All translation happens locally

What's New - Ollama Integration

This version replaces cloud-based translation APIs with local LLM translation using Ollama:

• 100% offline translation (after initial model download)
• No API costs or rate limits
• Enhanced privacy - your data never leaves your machine
• Customizable - Easy to switch between different Ollama models
• High-quality translations powered by Llama 3.2

📸 Demo / Output

The application provides a user-friendly graphical interface for real-time voice translation:

Prerequisites

Before running this application, ensure you have:

• Python 3.7 or higher
• A working microphone
• Audio output device for playback
• Ollama installed and running locally (see installation steps below)
• Internet connection (only for initial Ollama model download and speech recognition)

Installation

Step 1: Install Ollama

Ollama is required for local LLM-based translation.

1. Visit ollama.com and download the installer for your OS

2. Install Ollama following the platform-specific instructions:

   • Windows: Run the installer
   • macOS: Drag to Applications folder
   • Linux: Follow the command line installation

3. Verify Ollama installation:

ollama --version

4. Pull the Llama 3.2 model (used by default):

ollama pull llama3.2:latest

5. Start the Ollama server (if not already running):

ollama serve

Note: The Ollama server should be running at http://localhost:11434 (default). The application will connect to this endpoint.

Step 2: Clone the Repository

Option A: Using VS Code Terminal

1. Open Visual Studio Code
2. Open a new terminal (Terminal → New Terminal or `Ctrl+Shift+``)
3. Navigate to your desired directory:

cd path/to/your/desired/folder

4. Clone the repository:

git clone https://github.com/Sharan-Kumar-R/Talk2Translate.git

5. Open the project folder:

cd Talk2Translate

6. Open the project in VS Code:

code .

Option B: Using VS Code Git Integration

1. Open Visual Studio Code
2. Press Ctrl+Shift+P (Windows/Linux) or Cmd+Shift+P (Mac)
3. Type "Git: Clone" and select it
4. Paste the repository URL: https://github.com/Sharan-Kumar-R/Talk2Translate.git
5. Choose a folder location and click "Select Repository Location"
6. Click "Open" when prompted

Step 3: Create a Virtual Environment (Recommended)

1. In the VS Code terminal, create a virtual environment:

python -m venv venv

2. Activate the virtual environment:
   • Windows:

venv\Scripts\activate

• macOS/Linux:

source venv/bin/activate

Step 4: Install Python Dependencies

Install the required packages:

pip install -r requirements.txt

If you don't have a requirements.txt file, install packages individually:

pip install customtkinter SpeechRecognition gTTS pygame ollama

Step 5: Install PortAudio (Required for Microphone Input)

The SpeechRecognition library requires PortAudio for microphone functionality.

macOS:

brew install portaudio

Windows:

1. Download the appropriate PyAudio wheel file from Unofficial Windows Binaries
2. Install using pip:

pip install PyAudio-0.2.11-cp39-cp39-win_amd64.whl

Alternatively, install Microsoft C++ Build Tools if needed.

Linux (Debian/Ubuntu):

sudo apt-get install portaudio19-dev python3-pyaudio

Usage

1. Start Ollama Server (if not already running)

ollama serve

2. Run the Application

python Bilingual.py

3. Select Translation Mode

• Choose "English → Tamil" to speak in English
• Choose "Tamil → English" to speak in Tamil

4. Start Listening

• Click the "🎧 Start Listening" button
• Speak clearly into your microphone
• Wait for the LLM translation to appear

5. Save Translations

• Click "💾 Save Translation" to save the current translation
• Translations are saved to translations.txt in the application directory

Configuration

You can customize the Ollama configuration by editing these variables in Bilingual.py:

OLLAMA_HOST = "http://localhost:11434"  # Ollama server address
OLLAMA_MODEL = "llama3.2:latest"        # Model to use for translation

Available Ollama Models for Translation

You can experiment with different models:

# Smaller, faster model
ollama pull llama3.2:1b

# Default model (balanced)
ollama pull llama3.2:latest

# Larger, more capable model
ollama pull llama3.1:8b

# Or other multilingual models
ollama pull mistral

Update the OLLAMA_MODEL variable in the code to switch models.

How It Works

1. Speech Capture: The application listens to your microphone input
2. Speech-to-Text: Converts your speech to text using Google Speech Recognition
3. LLM Translation: Translates the text using local Ollama LLM (Llama 3.2)
4. Text-to-Speech: Converts the translated text to speech using gTTS
5. Playback: Plays the audio translation automatically

Project Structure

Talk2Translate/
│
├── Bilingual.py          # Main application file (uses google translator)
├── Bilingualllm.py       # Main application file (uses local llm)
├── Talk2Translate.png    # Demo image
├── requirements.txt      # Python dependencies
├── translations.txt      # Saved translations (auto-generated)
└── README.md             # This file

Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

In case of any queries, please leave a message or contact me via the email provided in my profile.

⭐ Star this repository if you found it helpful!