sekarkrishna/legalbert-int8

LegalBERT INT8 — ONNX Quantized

ONNX INT8 quantized version of nlpaueb/legal-bert-base-uncased for efficient legal text embeddings.

Model Details

What is this?

This is a quantized ONNX export of LEGAL-BERT, a BERT model pre-trained on 12GB of diverse legal text including EU legislation, UK legislation, European Court of Justice cases, and US court cases and contracts. The INT8 quantization reduces model size and improves inference speed while maintaining high accuracy for legal domain embeddings.

Use Cases

• Legal document search and retrieval
• Contract analysis
• Case law similarity
• Legal text classification
• Regulatory compliance document matching

Files

• model_quantized.onnx — INT8 quantized ONNX model
• tokenizer.json — Fast tokenizer
• vocab.txt — Vocabulary file
• config.json — Model configuration

Usage with JustEmbed

from justembed import Embedder

embedder = Embedder("legalbert-int8")
vectors = embedder.embed(["breach of fiduciary duty"])

Usage with ONNX Runtime

import onnxruntime as ort
from transformers import AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained(".")
session = ort.InferenceSession("model_quantized.onnx")

inputs = tokenizer("breach of fiduciary duty", return_tensors="np")
outputs = session.run(None, dict(inputs))

Quantization Details

• Method: Dynamic INT8 quantization via ONNX Runtime
• Source: Original PyTorch weights converted to ONNX, then quantized
• Speed: ~2-3x faster inference than FP32
• Size: ~4x smaller than FP32

License

This model is a derivative work of nlpaueb/legal-bert-base-uncased.

The original model is licensed under Creative Commons Attribution-ShareAlike 4.0 International (CC BY-SA 4.0). This quantized version is distributed under the same license as required by the ShareAlike clause. See the LICENSE file for details.

Citation

@inproceedings{chalkidis2020legal,
  title={LEGAL-BERT: The Muppets straight out of Law School},
  author={Chalkidis, Ilias and Fergadiotis, Manos and Malakasiotis, Prodromos and Aletras, Nikolaos and Androutsopoulos, Ion},
  booktitle={Findings of EMNLP},
  year={2020}
}

Acknowledgments

• Original model by AUEB NLP Group
• Quantization and packaging by JustEmbed