English NER (7 classes) — Flukes

Named entity recognition for English text. Predicts seven entity types:

⚠️ Default license: noncommercial use only. This model is released under the Flukes NC 1.0 License. Commercial use — including using this model's predictions in a commercial product or service — requires a separate license. Contact alan.akbik@hu-berlin.de.

---

Use

Install flukes:

pip install flukes

Then:

from flukes.models import Model

# load the named entity recognition (NER) model
model = Model.load("ner")

# the text you would like to tag
text = "Washington bought himself a pair of Nike Vaporflys.\n\nHe wore them at the Berlin Marathon."

# predict tags for the model
doc = model.predict(text)

# print the document with tags
print(doc)

Output:

Document[89]: Washington bought himself a pair of Nike Vaporflys.
              ╰───PER──╯                          ╰───PRODUCT──╯

              He wore them at the Berlin Marathon.
                                  ╰────EVENT────╯

Access Annotations

You can also access all annotations directly:

from flukes.models import Model

# load the named entity recognition (NER) model
model = Model.load("ner")

# the text you would like to tag
text = "Washington bought himself a pair of Nike Vaporflys.\n\nHe wore them at the Berlin Marathon."

# predict tags for the model
doc = model.predict(text)

# iterate over spans and print information
for span in doc.spans:
    print(f"{span.text!r:30}  {span.label}  ({span.score:.2f})")

Output:

'Washington'                    PER  (1.00)
'Nike Vaporflys'                PRODUCT  (1.00)
'Berlin Marathon'               EVENT  (1.00)

Performance

This model scores 96.1 F1 on our internal test sets.

License

Model weights: Flukes Noncommercial License 1.0. Personal, academic, and other noncommercial use permitted. Commercial use requires a separate license — contact alan.akbik@hu-berlin.de.

Flukes library code: Apache 2.0.