README.md

---
language: en
tags: [ner, token-classification, criminal-justice, alfred-ai]
license: mit
base_model: dslim/bert-base-NER
---
# Alfred NER

Fine-tuned from `dslim/bert-base-NER` for criminal justice entity extraction.

## Labels
PER · ORG · LOC · DATE · VEH · WEAPON · CASE · ALIAS

## Usage
```python
from transformers import pipeline
ner = pipeline("ner", model="sammmmmmm25/alfred-ner", aggregation_strategy="simple")
ner("Suspect Kevin Truro fled in a white Honda Accord on Route 9.")
```

## Training
- Base model: `dslim/bert-base-NER`
- Epochs: 10 | Batch size: 16 | fp16: True
- Validation F1: 0.8646