ner-simple 0.1.0

A minimal package exposing a simple NER runner using Hugging Face transformers pipeline.

ner_simple

A minimal package that exposes a single function run_ner(text, model=...) to run Named Entity Recognition using Hugging Face transformers pipeline.

Files

• ner_simple/ner.py - contains run_ner function.
• requirements.txt - basic dependencies.

Usage

pip install -r requirements.txt
# or install transformers and torch manually

from ner_simple import run_ner

text = "Barack Obama was born in Hawaii and was the 44th President of the United States."
entities = run_ner(text)
print(entities)

The default model is dbmdz/bert-large-cased-finetuned-conll03-english which is fine-tuned for CoNLL-03 NER tasks.