🎙️ Personal Storyteller with LLaMA 3.1 (English & Hindi)

Personal Storyteller is a Python application that generates personalized stories using either voice input or typed prompts. It leverages offline speech recognition with Vosk, story generation via LLaMA 3.1 (Ollama), and converts text-to-speech using gTTS, all wrapped in a Gradio web interface for easy interaction.

---

🚀 Features

• Voice Input: Record your voice and transcribe it into text offline using Vosk.
• Text Input: Optionally type your story prompt.
• Story Generation: Generate creative stories with LLaMA 3.1 via Ollama.
• Text-to-Speech: Convert generated stories into audio using gTTS.
• Bilingual Support: Works with English and Hindi.
• Interactive UI: User-friendly web interface powered by Gradio.

---

🛠️ Installation

1. Clone the repository

git clone https://github.com/yourusername/personal-storyteller.git
cd personal-storyteller
2. Set up Python environment
Create a virtual environment and activate it:

Linux/Mac:
python -m venv .venv && source .venv/bin/activate

Windows:
python -m venv .venv && .venv\Scripts\activate

Then install dependencies:

2.Copy code pip install -r requirements.txt Dependencies include vosk, sounddevice, gradio, gtts, numpy, scipy.

3. Download Vosk Models Download the required models from Vosk Models and place them in the models/ folder:

vosk-model-small-en-us-0.15 (English)

vosk-model-small-hi-0.22 (Hindi)

4. Install Ollama & LLaMA 3.1 Download and install Ollama. Ensure ollama.exe is in your system PATH, or update the path in app.py.

🎯 Usage Run the application:

bash Copy code python app.py Open the Gradio web interface (usually http://127.0.0.1:7860).

Select language (English or Hindi).

Record your voice or type a prompt.

Click 🎤 Record & Tell Story.

View the transcript, generated story, and play the audio output.

📂 Project Structure
bash
Copy code
personal-storyteller/
├── app.py             # Main application script
├── models/            # Vosk language models
├── output/            # Generated audio files  
└── README.md

⚙️ How It Works

Input: Accepts user voice or typed prompt.

STT (Speech-to-Text): Converts voice to text using Vosk.

Story Generation: LLaMA 3.1 generates a story from the prompt via Ollama.

TTS (Text-to-Speech): Converts generated story to audio using gTTS.

Output: Displays transcript, story text, and audio playback in Gradio.

🧰 Key Libraries & Tools

Library Purpose vosk Offline speech recognition gtts Text-to-speech gradio Interactive web interface sounddevice Capture microphone audio subprocess Run Ollama CLI for story generation numpy & scipy Audio resampling and processing pathlib & shutil File handling and path management

🤝 Contributing

Contributions are welcome! To contribute:

Fork the repository.

Create a feature branch (git checkout -b feature-name).

Commit your changes (git commit -m "Add feature").

Push to the branch (git push origin feature-name).

Open a Pull Request.

📄 License

This project is licensed under the MIT License. © Tejas