xlm-roberta-large-sag

XLM-RoBERTa-large-sag

Model description

This is a model based on the XLM-RoBERTa large topology (provided by Facebook, see original paper) with additional training on two sets of medicine-domain texts:

• about 250.000 text reviews on medicines (1000-tokens-long in average) collected from the site irecommend.ru;
• the raw part of the RuDReC corpus (about 1.4 million texts, see paper).

The XLM-RoBERTa-large calculations for one epoch on this data were performed using one Nvidia Tesla v100 and the Huggingface Transformers library.

BibTeX entry and citation info

If you have found our results helpful in your work, feel free to cite our publication as:

@article{sboev2021analysis,
  title={An analysis of full-size Russian complexly NER labelled corpus of Internet user reviews on the drugs based on deep learning and language neural nets},
  author={Sboev, Alexander and Sboeva, Sanna and Moloshnikov, Ivan and Gryaznov, Artem and Rybka, Roman and Naumov, Alexander and Selivanov, Anton and Rylkov, Gleb and Ilyin, Viacheslav},
  journal={arXiv preprint arXiv:2105.00059},
  year={2021}
}