# Healing Music Classifier

This is a machine learning model that predicts the healing potential of music tracks. The model analyzes various audio features to determine how likely a piece of music is to have healing properties.

## Model Details

- **Type**: Random Forest Classifier
- **Features**: MFCC, Spectral Features, Chroma Features
- **Input**: Audio file (MP3 or WAV format)
- **Output**: Healing probability score (0-1)

## Usage

```python
import joblib
from huggingface_hub import hf_hub_download

# Download model files
model_path = hf_hub_download(repo_id="healing-music-classifier", filename="models/model.joblib")
scaler_path = hf_hub_download(repo_id="healing-music-classifier", filename="models/scaler.joblib")

# Load model and scaler
model = joblib.load(model_path)
scaler = joblib.load(scaler_path)

# Use the model (after feature extraction)
# prediction = model.predict_proba(scaled_features)[0][1]
```

## Web Interface

You can try the model directly through our Streamlit interface at:
https://huggingface.co/spaces/[your-username]/healing-music-classifier

## License

MIT License