Fine Grained Named Entity Recognition with Transformer

Last updated on Mar 15, 2021

Fine Grained Named Entity Recognition with Transformer

Parameters 125 Million
File Size 418.33 MB
Training Data OntoNotes 5.0

Training Techniques AdamW
Architecture CRF, Dropout, Layer Normalization, Linear Layer, RoBERTa, Tanh
LR 1e-05
Epochs 20
Batch Size 16
Encoder Type pass_through
SHOW MORE
SHOW LESS
README.md

Summary

Fine-grained NER model

How do I load this model?

from allennlp_models.pretrained import load_predictor
predictor = load_predictor("tagging-fine-grained-transformer-crf-tagger")

Getting predictions

sentence = "Jobs and Wozniak cofounded Apple in 1976."
preds = predictor.predict(sentence)
for word, tag in zip(preds["words"], preds["tags"]):
    print(word, tag)
# prints:
# Jobs U-PERSON
# and O
# Wozniak U-PERSON
# cofounded O
# Apple U-ORG
# in O
# 1976 U-DATE
# . O

You can also get predictions using allennlp command line interface:

echo '{"sentence": "Jobs and Wozniak cofounded Apple in 1976."}' | \
    allennlp predict https://storage.googleapis.com/allennlp-public-models/fgner-transformer.2021-02-11.tar.gz -

How do I train this model?

To train this model you can use allennlp CLI tool and the configuration file fgner_transformer.jsonnet:

allennlp train fgner_transformer.jsonnet -s output_dir

See the AllenNLP Training and prediction guide for more details.

Results

Named Entity Recognition on Ontonotes v5 (English)

Named Entity Recognition
BENCHMARK MODEL METRIC NAME METRIC VALUE GLOBAL RANK
Ontonotes v5 (English) Fine Grained Named Entity Recognition with Transformer F1 88 # 1