cardiffnlp/twitter-roberta-base-ner7-latest

This is a RoBERTa-large model trained on 154M tweets until the end of December 2022 and finetuned for topic Name Entity Recognition on the TweetNER7 dataset of SuperTweetEval. The original Twitter-based RoBERTa model can be found here.

Labels

"id2label": { "0": "B-corporation", "1": "B-creative_work", "2": "B-event", "3": "B-group", "4": "B-location", "5": "B-person", "6": "B-product", "7": "I-corporation", "8": "I-creative_work", "9": "I-event", "10": "I-group", "11": "I-location", "12": "I-person", "13": "I-product", "14": "O" }

Example

from transformers import pipeline
text = "Halo Infinite analysis - The only true analysis {{USERNAME}} {{USERNAME}} {{USERNAME}} {{USERNAME}} {{USERNAME}} {{URL}}"

model_name = "cardiffnlp/twitter-roberta-base-ner7-latest"
pipe = pipeline('ner', model=model_name, tokenizer=model_name, aggregation_strategy="simple")
predictions = pipe(text)
predictions
>> [{'entity_group': 'creative_work',
      'score': 0.5278398,
      'word': 'Halo Infinite',
      'start': 0,
      'end': 13}]

Citation Information

Please cite the reference paper if you use this model.

@inproceedings{antypas2023supertweeteval,
  title={SuperTweetEval: A Challenging, Unified and Heterogeneous Benchmark for Social Media NLP Research},
  author={Dimosthenis Antypas and Asahi Ushio and Francesco Barbieri and Leonardo Neves and Kiamehr Rezaee and Luis Espinosa-Anke and Jiaxin Pei and Jose Camacho-Collados},
  booktitle={Findings of the Association for Computational Linguistics: EMNLP 2023},
  year={2023}
}