HumLyric is an intelligent lyric-writing assistant that helps artists craft meaningful and rhythmically sound lyrics using AI. You provide seed words or syllable structures, and the tool generates lyrical drafts that follow your style and constraints.
- AI-powered lyric generation using Groq's large language models
- Syllable-aware prompts to guide rhythmic accuracy
- Pinned words to enforce topic consistency across lines
- Regenerates lyrics on demand with stylistic variance
- Stores multiple drafts for review or editing
- Backend: Python
groqAPI for LLM access- Custom syllable counter (vowel-based pattern detection)
- Async generation pipeline
- Frontend: (optional — React, Flask, etc.)
- Clone the repository
git clone https://github.com/yourusername/humlyric.git
cd humlyric- Install dependencies
pip install -r requirements.txt- Set your Groq API key
export GROQ_API_KEY=your-api-key- Run the server
python main.py| File | Purpose |
|---|---|
services/lyric.py |
Core logic for generating lyrics |
utils/syllables.py |
Syllable counter for rhythm shaping |
main.py |
Entry point for testing lyric prompts |
requirements.txt |
Python dependencies |
The system uses:
- Seed words for inspiration (e.g., jot, drop, let)
- Pinned words to anchor recurring themes
- Syllable count templates to structure verses (like melody scaffolding)
In the stillness, I'll inscribe your name in journal
Jot it down in my journal, next to things I’ll do
That summer breeze, it whispers low and sweet
- Add rhyme-scheme detection
- UI for interactive line editing
- Melody input → automatic syllable template generation
- Export lyrics to
.txtor.lrcfile
Built by Shihab Jamal
Inspired by the intersection of music, language, and AI.
MIT License – do what you want with it, just give credit.