This is our fine-tuned mDeBERTa SEA translationese classifier for the "SEACrowd: A Multilingual Multimodal Data Hub and Benchmark Suite for Southeast Asian Languages" paper.
SEACrowd is a collaborative initiative that consolidates a comprehensive resource hub that fills the resource gap by providing standardized corpora in nearly 1,000 Southeast Asian (SEA) languages across three modalities.
To analyze the generation quality of LLMs in SEA languages, we build a text classifier to discriminate between translationese and natural texts. We construct a translationese classification training and testing dataset using 49 and 62 data subsets, respectively, covering approximately 39.9k and 51.5k sentences across 9 SEA languages: English (eng), Indonesian (ind), Khmer (khm), Lao (lao), Burmese (mya), Filipino (fil), Thai (tha), Vietnamese (vie), and Malay (zlm).
Our translationese vs. natural train/test data is available on SEACrowd/sea_translationese_resampled.
To fine-tune the translationese classifier, check out our experiments repository on GitHub. We use a binary label (translationese, i.e., machine-translated or human-translated, or natural, i.e., human-generated) instead of 3 labels (machine-translated, human-translated, human-generated).
This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
To discriminate between translationese and natural texts in 9 SEA languages: English (eng), Indonesian (ind), Khmer (khm), Lao (lao), Burmese (mya), Filipino (fil), Thai (tha), Vietnamese (vie), and Malay (zlm).
The model is developed for detecting whether a text is human-translated, machine-translated, or natural.
The model supports 9 languages: eng, ind, khm, lao, mya, fil, tha, vie, zsm
The label mapping of the model is defined as follows:
{0: 'Human-translated', 1: 'Machine-translated', 2: 'Natural'}
where both 0 and 1 correspond to translationese and 2 is natural.
The model achieves 79.08% accuracy on translationese (combining human-translated and machine-translated) vs natural in our evaluation——averaged across the aforementioned SEA languages.
Users should be aware of the risks that there might be potential error produced by the model.
See our paper for more details.
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model.
tokenizer = AutoTokenizer.from_pretrained('SEACrowd/mdeberta-v3_sea_translationese')
model = AutoModelForSequenceClassification.from_pretrained('SEACrowd/mdeberta-v3_sea_translationese')
inputs = tokenizer('<INPUT_TEXT>', padding='longest', max_length=512, truncation=True)
outputs = model(**inputs)
If you are using any resources from SEACrowd, including datasheets, dataloaders, code, etc., please cite the following publication:
@article{lovenia2024seacrowd,
title={SEACrowd: A Multilingual Multimodal Data Hub and Benchmark Suite for Southeast Asian Languages},
author={Holy Lovenia and Rahmad Mahendra and Salsabil Maulana Akbar and Lester James V. Miranda and Jennifer Santoso and Elyanah Aco and Akhdan Fadhilah and Jonibek Mansurov and Joseph Marvin Imperial and Onno P. Kampman and Joel Ruben Antony Moniz and Muhammad Ravi Shulthan Habibi and Frederikus Hudi and Railey Montalan and Ryan Ignatius and Joanito Agili Lopo and William Nixon and Börje F. Karlsson and James Jaya and Ryandito Diandaru and Yuze Gao and Patrick Amadeus and Bin Wang and Jan Christian Blaise Cruz and Chenxi Whitehouse and Ivan Halim Parmonangan and Maria Khelli and Wenyu Zhang and Lucky Susanto and Reynard Adha Ryanda and Sonny Lazuardi Hermawan and Dan John Velasco and Muhammad Dehan Al Kautsar and Willy Fitra Hendria and Yasmin Moslem and Noah Flynn and Muhammad Farid Adilazuarda and Haochen Li and Johanes Lee and R. Damanhuri and Shuo Sun and Muhammad Reza Qorib and Amirbek Djanibekov and Wei Qi Leong and Quyet V. Do and Niklas Muennighoff and Tanrada Pansuwan and Ilham Firdausi Putra and Yan Xu and Ngee Chia Tai and Ayu Purwarianti and Sebastian Ruder and William Tjhi and Peerat Limkonchotiwat and Alham Fikri Aji and Sedrick Keh and Genta Indra Winata and Ruochen Zhang and Fajri Koto and Zheng-Xin Yong and Samuel Cahyawijaya},
year={2024},
eprint={2406.10118},
journal={arXiv preprint arXiv: 2406.10118}
}