🎵 LyricMind-AI: Creative Song Lyrics Generator

[Project Banner Image]

LyricMind-AI is an advanced deep learning model that generates creative and contextually relevant song lyrics. Built with PyTorch and Flask, it uses LSTM architecture to understand and generate human-like lyrics across different musical genres.

📋 Table of Contents

• Features
• Installation
• Usage
• API Reference
• Model Architecture
• Contributing
• License

🌟 Features

• Creative Lyrics Generation: Generate unique and contextually relevant lyrics from any prompt
• Genre-Aware: Supports multiple music genres including Pop, Rock, Hip Hop, Country, and Jazz
• Temperature Control: Adjust creativity vs. coherence with temperature parameter
• User-Friendly Interface: Clean, modern web interface for easy interaction
• Real-Time Generation: Fast response times with asynchronous processing

[Screenshot of Web Interface]

🚀 Installation

Prerequisites

• Python 3.8 or higher
• pip (Python package manager)
• Git
• 4GB+ RAM recommended
• CUDA-capable GPU (optional, for faster training)

Step-by-Step Installation

1. Clone the Repository

   git clone https://github.com/AMIRHaytham/lyricMind-AI.git
   cd lyricMind-AI

2. Set Up Virtual Environment

   # Windows
   python -m venv venv
   .\venv\Scripts\activate
   
   # macOS/Linux
   python3 -m venv venv
   source venv/bin/activate

3. Install Dependencies

   pip install -r requirements.txt

4. Download Required Data

   • Download the datasets:
     • Spotify Million Song Dataset
     • Billboard Top 500
   • Place the files in the project root directory

💻 Usage

Web Interface

1. Start the server:

   python app.py

2. Open http://localhost:5000 in your browser
3. Enter your prompt and adjust parameters
4. Click "Generate Lyrics"

Python API

from lyricmind import LyricGenerator

# Initialize generator
generator = LyricGenerator(
    genre='pop',
    model_path='best_model.pth',
    vocab_path='vocab.json'
)

# Generate lyrics
lyrics = generator.generate(
    prompt="In the midnight hour",
    temperature=0.7,
    max_length=100
)

🔌 API Reference

RESTful API Endpoints

POST /generate

Generate lyrics from a prompt.

Request Body:

{
    "prompt": "In the midnight hour",
    "temperature": 0.7,
    "max_length": 100,
    "genre": "pop"
}

Response:

{
    "lyrics": "Generated lyrics...",
    "error": null
}

For detailed API documentation, see our API Reference Guide.

🧠 Model Architecture

Overview

LyricMind-AI uses a deep learning architecture based on LSTM networks:

Input Text → Embedding → LSTM → Dropout → Linear → Softmax → Output

Components

• Embedding Layer: 64 dimensions, ~50,000 word vocabulary
• LSTM Layer: 128 hidden units, single layer, 0.3 dropout
• Output Layer: Linear transformation with softmax activation

Training Details

• Dataset: 1M+ lyrics across multiple genres
• Training Parameters:
  • Batch Size: 64
  • Learning Rate: 0.001
  • Epochs: 50
  • Optimizer: Adam

👥 Contributing

We welcome contributions! Please see our Contributing Guidelines for details on:

• Code of Conduct
• Development Process
• Pull Request Process
• Style Guidelines

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

Created by Amir Haytham | GitHub