gpt-bert-babylm-base

GPT-BERT (BabyLM 100M)

Submission to the BabyLM challenge 2024 trained on Baby-cosmo-fine-100M.

The training scripts are published here: https://github.com/ltgoslo/gpt-bert

@inproceedings{charpentier-samuel-2024-bert,
    title = "{BERT} or {GPT}: why not both?",
    author = "Charpentier, Lucas Georges Gabriel  and
      Samuel, David",
    editor = "Hu, Michael Y.  and
      Mueller, Aaron  and
      Ross, Candace  and
      Williams, Adina  and
      Linzen, Tal  and
      Zhuang, Chengxu  and
      Choshen, Leshem  and
      Cotterell, Ryan  and
      Warstadt, Alex  and
      Wilcox, Ethan Gotlieb",
    booktitle = "The 2nd BabyLM Challenge at the 28th Conference on Computational Natural Language Learning",
    month = nov,
    year = "2024",
    address = "Miami, FL, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2024.conll-babylm.24/",
    pages = "262--283",
}