voice-gender
augmented-interpretable-models
voice-gender | augmented-interpretable-models | |
---|---|---|
2 | 1 | |
331 | 37 | |
- | - | |
0.0 | 7.4 | |
over 1 year ago | 16 days ago | |
R | Jupyter Notebook | |
- | MIT License |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
voice-gender
- I need help for a project, Trans-voice database or library (vocal training/voice recognition)
-
DIY fix for the "What is Your Voice Gender" web app
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:
augmented-interpretable-models
-
[R] Emb-GAM: an Interpretable and Efficient Predictor using Pre-trained Language Models
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.
What are some alternatives?
Empirical_Study_of_Ensemble_Learning_Methods - Training ensemble machine learning classifiers, with flexible templates for repeated cross-validation and parameter tuning
language-planner - Official Code for "Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agents"
chatgptAutostats - Fit R Stats Models with ChatGPT
shap - A game theoretic approach to explain the output of any machine learning model. [Moved to: https://github.com/shap/shap]
scikit-learn-ts - Powerful machine learning library for Node.js – uses Python's scikit-learn under the hood.
DeepLearning - Contains all my works, references for deep learning
handson-ml - ⛔️ DEPRECATED – See https://github.com/ageron/handson-ml3 instead.
gan-vae-pretrained-pytorch - Pretrained GANs + VAEs + classifiers for MNIST/CIFAR in pytorch.
AutoCog - Automaton & Cognition
imodels - Interpretable ML package 🔍 for concise, transparent, and accurate predictive modeling (sklearn-compatible).
align-transformers - This is an old library. Try pyvene instead!