LEIA/LEIA-LM-base

This is a BERTweet-base model that has been further pre-trained with preferential masking of emotion words for 100k steps on about 6.3M Vent posts.

This model is meant to be fine-tuned on labeled data or used as feature extractor for downstream tasks.

Citation

Please cite the following paper if you find the model useful for your work:

@article{aroyehun2023leia,
  title={LEIA: Linguistic Embeddings for the Identification of Affect},
  author={Aroyehun, Segun Taofeek and Malik, Lukas and Metzler, Hannah and Haimerl, Nikolas and Di Natale, Anna and Garcia, David},
  journal={EPJ Data Science},
  volume={12},
  year={2023},
  publisher={Springer}
}