voice-gender VS augmented-interpretable-models

Hey y'all. I discovered the aforementioned app (What is Your Voice Gender? ) a while back and thought it was super cool. Recently I've been ramping up my voice work and went to use this app again, but any file I try to input results in an error (see a photo here) that makes it unusable. Recalling this was a personal project for the creator, I went searching for the project code in Github and found it here. I realized that despite the web app showing an error, it was kinda easy to go into RStudio myself and get the predictions without an error. I saw another user in the comments of a recent post experiencing this same issue, so I figured I'd just write out the steps to my DIY fix that worked, in case anybody else is having issues and would like to try this fix:

Deep learning models have achieved impressive prediction performance but often sacrifice interpretability, a critical consideration in high-stakes domains such as healthcare or policymaking. In contrast, generalized additive models (GAMs) can maintain interpretability but often suffer from poor prediction performance due to their inability to effectively capture feature interactions. In this work, we aim to bridge this gap by using pre-trained neural language models to extract embeddings for each input before learning a linear model in the embedding space. The final model (which we call Emb-GAM) is a transparent, linear function of its input features and feature interactions. Leveraging the language model allows Emb-GAM to learn far fewer linear coefficients, model larger interactions, and generalize well to novel inputs (e.g. unseen ngrams in text). Across a variety of NLP datasets, Emb-GAM achieves strong prediction performance without sacrificing interpretability. All code is made available on Github.