da03/Residual-EBM
da03/Residual-EBM has 26 GitHub stars and 2 forks. GitHot measured 0 stars in 24 hours. Last checked 2026-11-19 12:09 UTC. Last push 2026-11-11.
Progression mesurée des étoiles
- in 24h
- 0
- in 7 days
- Historique en cours
- in 15 days
- Historique en cours
- in 1 month
- Historique en cours
Variations nettes sur des périodes qui se chevauchent, pas des totaux quotidiens. Les mesures absentes restent vides. Les étoiles indiquent l’intérêt, pas la qualité.
À propos
Code for Residual Energy-Based Models for Text Generation in PyTorch.
Apprentissage automatiqueREADME.md
OriginalResidual Energy-Based Models for Text Generation
This repo implements the Residual Energy-Based Model as described in Residual Energy-Based Models for Text Generation. Note that both data and model (generator+discriminator) are different from the original paper: the goal is just to show how the pipeline works, using much less computational resources.
The training and evaluation scripts can be found in REBM.ipynb. We suggest opening it using Google Colab with a GPU instance.
Method and related research
Energy-based text generation: residual models, resampling and diffusion explains the sequence-level energy formulation, noise-contrastive estimation, and importance resampling, with an interactive reweighting example and connections to energy-based diffusion language models.
Citation
@inproceedings{deng2020residual,
title={Residual Energy-Based Models for Text Generation},
author={Deng, Yuntian and Bakhtin, Anton and Ott, Myle and Szlam, Arthur and Ranzato, Marc'Aurelio},
booktitle={International Conference on Learning Representations},
year={2020}
}
@article{bakhtin2021residual,
title={Residual Energy-Based Models for Text},
author={Bakhtin, Anton and Deng, Yuntian and Gross, Sam and Ott, Myle and Ranzato, Marc'Aurelio and Szlam, Arthur},
journal={Journal of Machine Learning Research},
volume={22},
number={40},
pages={1--41},
year={2021}
}