Headless Language Models: Learning without Predicting with Contrastive Weight Tying
Paper • 2309.08351 • Published • 3
This model is a Pythia-70m architecture trained on OpenWebText-2 using the Contrastive Weight Tying objective, and briefly fine-tuned for language generation on the same dataset.
BibTeX:
@misc{godey2023headless,
title={Headless Language Models: Learning without Predicting with Contrastive Weight Tying},
author={Nathan Godey and Éric de la Clergerie and Benoît Sagot},
year={2023},
eprint={2309.08351},
archivePrefix={arXiv},
primaryClass={cs.CL}
}