---
license: mit
language:
- pt
library_name: transformers.js
pipeline_tag: token-classification
base_model: Luciano/bertimbau-large-lener_br
tags:
- onnx
- transformers.js
- ner
- token-classification
- legal
- portuguese
- bertimbau
- lener_br
datasets:
- peluz/lener_br
---

# bertimbau-large-lener_br-onnx

ONNX conversion of [`Luciano/bertimbau-large-lener_br`](https://huggingface.co/Luciano/bertimbau-large-lener_br) — Brazilian **legal** Named Entity Recognition (LeNER-Br, 6 classes) on top of BERTimbau-large — for [Transformers.js](https://github.com/huggingface/transformers.js) (v3+) and ONNX Runtime.

- **Conversion pipeline (reproducible):** [github.com/rchuluc/bertimbau-large-lener_br-onnx](https://github.com/rchuluc/bertimbau-large-lener_br-onnx)
- **Encoder weights:** unchanged from upstream. **License:** MIT.

## Files

| file | dtype | size |
|---|---|---|
| `onnx/model.onnx` | fp32 | 1.33 GB |
| `onnx/model_quantized.onnx` | int8 (QUInt8 dynamic) | 335 MB |

## Verified parity (vs PyTorch reference, 10 PT-BR legal sentences)

| metric | fp32 | q8 |
|---|---|---|
| entity parity (Optimum/ORT, `aggregation_strategy="simple"`) | **26/26 (100%)** | **25/26 (96.2%)** |
| token-level argmax parity (PyTorch↔ONNX) | **196/196 (100%)** | **193/196 (98.5%)** |
| max \|Δlogit\| vs PyTorch | 2.1e-5 | — |

## Usage

### Python (Optimum / ONNX Runtime) — recommended

```python
from optimum.onnxruntime import ORTModelForTokenClassification
from transformers import AutoTokenizer, pipeline

model = ORTModelForTokenClassification.from_pretrained(
    "rchuluc/bertimbau-large-lener_br-onnx",
    subfolder="onnx",
    file_name="model_quantized.onnx",  # or "model.onnx"
)
tok = AutoTokenizer.from_pretrained("rchuluc/bertimbau-large-lener_br-onnx")

pipe = pipeline("ner", model=model, tokenizer=tok, aggregation_strategy="simple")
print(pipe("Conforme o art. 5º da Constituição Federal, todos são iguais perante a lei."))
```

### Transformers.js

```js
import { pipeline } from '@huggingface/transformers';

const ner = await pipeline(
  'token-classification',
  'rchuluc/bertimbau-large-lener_br-onnx',
  { dtype: 'q8' }, // or 'fp32'
);
const out = await ner('O Supremo Tribunal Federal julgou a ação em Brasília.', {
  ignore_labels: ['O'],
});
// Aggregate B-/I- + WordPiece (##) yourself.
```

> ⚠️ **For legal text, prefer the Python/Optimum path.** Transformers.js v3's
> `TokenClassificationPipeline` drops `[UNK]` tokens and lacks `aggregation_strategy`.
> Legal entities contain out-of-vocab ordinals/symbols (`5º`, `nº` → `[UNK]`), so spans
> like `art. 5º da Constituição Federal` lose pieces in JS. The ONNX model labels those
> tokens correctly (see token-level parity); the loss is in the JS pipeline, not the model.

## Classes (LeNER-Br, 6 types)

`ORGANIZACAO`, `PESSOA`, `TEMPO`, `LOCAL`, `LEGISLACAO`, `JURISPRUDENCIA`

## Attribution

- Original: [Luciano/bertimbau-large-lener_br](https://huggingface.co/Luciano/bertimbau-large-lener_br), fine-tuned on [`peluz/lener_br`](https://huggingface.co/datasets/peluz/lener_br).
- Base: [neuralmind/bert-large-portuguese-cased](https://huggingface.co/neuralmind/bert-large-portuguese-cased) (BERTimbau).
- Dataset: LeNER-Br (Luz de Araujo et al., PROPOR 2018).

Not affiliated with the original authors. Cite the original work in any publication.