8 projects
surgicalplan
SurgicalPLAN is a Python package for predicting postoperative risks from clinical notes using language models. It provides training and inference workflows for fine-tuned models, semi-supervised methods, and multi-task prediction of multiple surgical outcomes. The package is intended for clinical research and educational use, notably for the American College of Surgeons.
tipeft
Tabular-Infused Parameter Efficient Finetuning (tipeft)
changedotorgscraper
Scrapes online petition data from change.org
restep
Reason-based RAG for Early-Stage Developers for Community-based Prospecting (RE-STEP)
ctscams
Continous Time-Series Clustering And Merging towards Sentiments (CT-SCAMS) is a Python package ddesigned to ensure robust generalization across time stamps for sentiment analysis in policy-related texts.
AutoThemeGenerator
Performing thematic analysis with OpenAI's GPT-4 models
ConTextMining
Complementing topic models with few-shot in-context learning to generate interpretable topics
