---
language:
- en
license: mit
tags:
- ner
- named-entity-recognition
- early-modern-english
- historical-texts
- lora
- peft
base_model: Qwen/Qwen3-4B-Instruct-2507
---

# EarlyModernNER Adapters

LoRA adapters for Named Entity Recognition in Early Modern English documents (1500-1800).

## Adapters

| Adapter | Entity Type | Precision | Recall | F1 |
|---------|-------------|-----------|--------|-----|
| `toponym_lora` | Place names | 0.93 | 0.82 | 0.87 |
| `person_lora` | People | 0.93 | 0.69 | 0.80 |
| `organization_lora` | Institutions | 0.93 | 0.46 | 0.62 |
| `commodity_lora` | Trade goods | 0.85 | 0.80 | 0.83 |

## Usage

These adapters are used by the [EarlyModernNER](https://github.com/polayj/earlymodernner) package:

```bash
pip install earlymodernner
python -m earlymodernner --input your_docs/ --output results.jsonl
```

Adapters are automatically downloaded on first use.

## Base Model

All adapters are trained on **Qwen/Qwen3-4B-Instruct-2507** using QLoRA (4-bit quantization).

## License

MIT License

## Author

Jacob Polay, University of Saskatchewan