Skip to content

Latest commit

 

History

History
25 lines (19 loc) · 1.58 KB

README.md

File metadata and controls

25 lines (19 loc) · 1.58 KB
Raw

AnnotationVocabulary

This is the official repository of the paper Annotation Vocabulary (Might Be) All You Need.

The data and models were developed using a proprietary internal developers package (udev), for which we've refactored a public version (udev_public). This repo is set up with the code intact, but currently, you will need to modify into your own files / file structure. For example, if you want to train a camp model you will have to place the udev_public file in a location that the train.py can see, or use import sys to make the parent folder available. We are planning on releasing some ready to run scripts in the future. Please be patient with us! We plan on releasing everything except the GSM data and weights for now - which is getting redone anyways for peer review submission.

Models, training data

Huggingface

Downstream data

Huggingface

gsm_demo.mp4

If you use any of our models, data, or code, please cite the following paper (and be aware of the GPL-3.0 license)

@article{hallee2024annotation,
      title={Annotation Vocabulary(Might Be) All You Need}, 
      author={Logan Hallee and Niko Rafailidis and Colin Horger and David Hong and Jason P. Gleghorn},
      year={2024},
      eprint={},
      archivePrefix={biorXiv},
      primaryClass={cs.LG}
}