README.md

metadata

license: apache-2.0
tags:
  - token-classification
datasets:
  - wikiann
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: ner-swedish-wikiann
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: wikiann
          type: wikiann
        metrics:
          - name: Precision
            type: precision
            value: 0.8331921416757433
          - name: Recall
            type: recall
            value: 0.84243586083126
          - name: F1
            type: f1
            value: 0.8377885044416501
          - name: Accuracy
            type: accuracy
            value: 0.91930707459758

ner-swedish-wikiann

This model is a fine-tuned version of nordic-roberta-wiki trained for NER on the wikiann dataset.

eval F1-Score: 83,78

test F1-Score: 83,76

Model Usage

from transformers import AutoTokenizer, AutoModelForTokenClassification

from transformers import pipeline

tokenizer = AutoTokenizer.from_pretrained("birgermoell/ner-swedish-wikiann")

model = AutoModelForTokenClassification.from_pretrained("birgermoell/ner-swedish-wikiann")

nlp = pipeline("ner", model=model, tokenizer=tokenizer)

example = "Jag heter Per och jag jobbar på KTH"

nlp(example)