Moody Lyrics


            
            

                
Did the model get the mood right?

                
                
            

        

        

            This app analyzes song lyrics to predict their mood. Enter a song title and artist, and get mood predictions along with the lyrics. Deployed using Hugging Face Spaces. The model may be inaccurate in some cases.
            


            Learn more.
            

                
How it Works:

                
This web app utilizes a fine-tuned BERT model to analyze the mood of song lyrics. Upon entering a song title and artist, the app retrieves the song's lyrics from Genius, processes them through the BERT model, and predicts the mood. The model's accuracy is 93% on its test set.

                
Technology:

                
    
                    
  • Model: BERT (Bidirectional Encoder Representations from Transformers) fine-tuned on song lyrics to classify moods.
    • 
                    
  • Framework: PyTorch and Hugging Face Transformers library for model training and prediction.
    • 
                    
  • Frontend: HTML, CSS, and JavaScript.
    • 
                    
  • Backend: Flask.
    • 
                    
  • Deployment: The web app is hosted and deployed using Hugging Face Spaces, which provides a simple interface for serving machine learning models.
    • 
                

                
For the source code, visit: GitHub Repository