metadata
tags:
- spacy
- token-classification
language:
- en
model-index:
- name: en_grantss
results:
- task:
name: NER
type: token-classification
metrics:
- name: NER Precision
type: precision
value: 0.769098972
- name: NER Recall
type: recall
value: 0.6617812852
- name: NER F Score
type: f_score
value: 0.7114156528
Introduction
Three variants of the model is built with Spacy3 for grant applications. A simple named entity recognition custom model from scratch with annotation tool prodi.gy. Github info: https://github.com/RaThorat/ner_model_prodigy The most general model is 'en_grantss'. The model en_ncv is more suitable to extract entities from narrative CV's. The model en_grant is the first model in the series.
| Feature | Description |
|---|---|
| Name | en_grantss |
| Version | 0.0.0 |
| spaCy | >=3.4.3,<3.5.0 |
| Default Pipeline | tok2vec, ner |
| Components | tok2vec, ner |
| Vectors | 0 keys, 0 unique vectors (0 dimensions) |
| Sources | research grant applications |
| License | n/a |
| Author | Rahul Thorat |
Label Scheme
View label scheme (18 labels for 1 components)
| Component | Labels |
|---|---|
ner |
ACTIVITY, DISCIPLINE, EVENT, GPE, JOURNAL, KEYWORD, LICENSE, MEDIUM, METASTD, MONEY, ORG, PERSON, POSITION, PRODUCT, RECOGNITION, REF, REPOSITORY, WEBSITE |
Accuracy
| Type | Score |
|---|---|
ENTS_F |
71.14 |
ENTS_P |
76.91 |
ENTS_R |
66.18 |
TOK2VEC_LOSS |
1412244.09 |
NER_LOSS |
1039417.96 |