metadata

tags:
  - spacy
  - token-classification
language:
  - en
model-index:
  - name: en_subref_ner
    results:
      - task:
          name: NER
          type: token-classification
        metrics:
          - name: NER Precision
            type: precision
            value: 0.9758551308
          - name: NER Recall
            type: recall
            value: 0.9837728195
          - name: NER F Score
            type: f_score
            value: 0.9797979798

Description

This model is designed to be used in conjunction with the en_torah_ner model. See the README there for how to integrate them.

The model takes citations as input and tags the parts of the citation as entities. This is very useful for parsing the citation.

Technical details

Feature	Description
Name	`en_subref_ner`
Version	`1.0.0`
spaCy	`>=3.4.1,<3.5.0`
Default Pipeline	`tok2vec`, `ner`
Components	`tok2vec`, `ner`
Vectors	218765 keys, 218765 unique vectors (50 dimensions)
Sources	n/a
License	GPLv3
Author	Sefaria

Label Scheme

View label scheme (7 labels for 1 components)

Component	Labels
`ner`	`DH`, `dir-ibid`, `ibid`, `non-cts`, `number`, `range-symbol`, `title`

Accuracy

Type	Score
`ENTS_F`	97.98
`ENTS_P`	97.59
`ENTS_R`	98.38
`TOK2VEC_LOSS`	5193.13
`NER_LOSS`	1103.44