---
language:
- sw
license: apache-2.0
datasets:
- masakhaner
pipeline_tag: token-classification
examples: null
widget:
  - text: Joe Bidden ni rais wa marekani.
    example_title: Sentence 1
  - text: Tumefanya mabadiliko muhimu katika sera zetu za faragha na vidakuzi.
    example_title: Sentence 2
  - text: Mtoto anaweza kupoteza muda kabisa.
    example_title: Sentence 3
metrics:
- accuracy
---

# Swahili Named Entity Recognition

- **TUS-NER-sw** is a fine-tuned BERT model that is ready to use for **Named Entity Recognition** and achieves **state-of-the-art performance 😀** 
- Finetuned from model: [eolang/SW-v1](https://huggingface.co/eolang/SW-v1)

## Intended uses & limitations

#### How to use

You can use this model with Transformers *pipeline* for NER.

```python
from transformers import pipeline
from transformers import AutoTokenizer, AutoModelForTokenClassification

tokenizer = AutoTokenizer.from_pretrained("eolang/SW-NER-v1")
model = AutoModelForTokenClassification.from_pretrained("eolang/SW-NER-v1")

nlp = pipeline("ner", model=model, tokenizer=tokenizer)
example = "Tumefanya mabadiliko muhimu katika sera zetu za faragha na vidakuzi"

ner_results = nlp(example)
print(ner_results)
```

## Training data

This model was fine-tuned on the Swahili Version of the [Masakhane Dataset](https://github.com/masakhane-io/masakhane-ner/tree/main/MasakhaNER2.0/data/swa) from the [MasakhaneNER Project](https://github.com/masakhane-io/masakhane-ner).
MasakhaNER is a collection of Named Entity Recognition (NER) datasets for 10 different African languages.
The languages forming this dataset are: Amharic, Hausa, Igbo, Kinyarwanda, Luganda, Luo, Nigerian-Pidgin, Swahili, Wolof, and Yorùbá.


## Training procedure

This model was trained on a single NVIDIA RTX 3090 GPU with recommended hyperparameters from the [original BERT paper](https://arxiv.org/pdf/1810.04805).