metadata
license: apache-2.0
Pretrained models for our paper (https://arxiv.org/abs/2210.08431)
@inproceedings{wu-etal-2022-modeling,
title = "Modeling Context With Linear Attention for Scalable Document-Level Translation",
author = "Zhaofeng Wu and Hao Peng and Nikolaos Pappas and Noah A. Smith",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2022",
month = dec,
year = "2022",
publisher = "Association for Computational Linguistics",
}
Please see the "Files and versions" tab for the models. You can find our IWSLT models and our OpenSubtitles models that are early-stopped based on BLEU and consistency scores, respectively. The c part in the checkpoint name refers to the number of context sentences used; it is the same as the sliding window size (the L in our paper) minus one.