AfroXLMR-Social

AfroXLMR-Social a social domain specialized model for 19 African Languages.

Pre-training corpus

AfriSocial corpus Available at: https://huggingface.co/datasets/Tadesse/AfriSocial

Languages

• Amharic (amh)
• Somali (som)
• Afrikaans (afr)
• Hausa (hau)
• Igbo (ibo)
• Yorùbá (yor)
• Kinyarwanda (kin)
• Tigrinya (tir)
• Oromo (orm)
• Twi (twi)
• Nigerian Pidgin (pcm)
• Algerian Arabic (arq)
• Moroccan Arabic (ary)
• Mozambican Portug. (ptMZ)
• Swahili (swa)
• Makhuwa (vmw)
• Xitsonga (tso)
• Xhosa (xho)
• Zulu (zul)

Acknowledgment

BibTeX entry and citation info.

@misc{belay2025afroxlmrsocialadaptingpretrainedlanguage,
      title={AfroXLMR-Social: Adapting Pre-trained Language Models for African Languages Social Media Text}, 
      author={Tadesse Destaw Belay and Israel Abebe Azime and Ibrahim Said Ahmad and David Ifeoluwa Adelani and Idris Abdulmumin and Abinew Ali Ayele and Shamsuddeen Hassan Muhammad and Seid Muhie Yimam},
      year={2025},
      eprint={2503.18247},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2503.18247}, 
}