Fine-tuned CoNLL-2003 NER weights for lucid.models.bert_base_token_cls.
dslim/bert-base-NER — a cased BERT-Base (vocab 28 996) with a
9-way BIO tag head (O, B/I-PER, B/I-ORG, B/I-LOC, B/I-MISC). Tokenize
with the cased BERTTokenizer (do_lower_case=False).
Attributes
CONLL2003WeightEntryCoNLL-2003 NER fine-tuned checkpoint (F1 91.3), sourced from
transformers/dslim/bert-base-NER.DEFAULTWeightEntryAlias for
CONLL2003.Notes
Reference: Devlin, Chang, Lee, Toutanova, "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding", NAACL 2019 (arXiv:1810.04805). Fine-tuned on CoNLL-2003 NER (dslim/bert-base-NER).
Examples
>>> from lucid.models import bert_base_token_cls
>>> model = bert_base_token_cls(pretrained=True).eval()