Zero Shot Vanilla Binary Bert

🚀 Zero-shot Vanilla Binary BERT

A BERT model for zero-shot text classification introduced in the Findings of ACL'23 Paper

This is a BERT model designed for zero-shot text classification. It was introduced in the Findings of ACL'23 Paper Label Agnostic Pre-training for Zero-shot Text Classification by Christopher Clarke, Yuzhao Heng, Yiping Kang, Krisztian Flautner, Lingjia Tang and Jason Mars. The code for training and evaluating this model can be found here.

✨ Features

• Intended for zero-shot text classification.
• Trained under the binary classification framework as a baseline with the aspect-normalized UTCD dataset.
• Finetuned from the well - known bert-base-uncased model.

📦 Installation

Install our python package:

pip install zeroshot-classifier

💻 Usage Examples

Basic Usage

>>> from zeroshot_classifier.models import BinaryBertCrossEncoder
>>> model = BinaryBertCrossEncoder(model_name='claritylab/zero-shot-vanilla-binary-bert')

>>> text = "I'd like to have this track onto my Classical Relaxations playlist."
>>> labels = [
>>>     'Add To Playlist', 'Book Restaurant', 'Get Weather', 'Play Music', 'Rate Book', 'Search Creative Work',
>>>     'Search Screening Event'
>>> ]

>>> query = [[text, lb] for lb in labels]
>>> logits = model.predict(query, apply_softmax=True)
>>> print(logits)

[[1.1909954e-04 9.9988091e-01]
 [9.9997509e-01 2.4927122e-05]
 [9.9997497e-01 2.5082643e-05]
 [2.4483365e-04 9.9975520e-01]
 [9.9996781e-01 3.2211588e-05]
 [9.9985993e-01 1.4002046e-04]
 [9.9976152e-01 2.3845369e-04]]

📚 Documentation

Model description

This model is intended for zero-shot text classification. It was trained under the binary classification framework as a baseline with the aspect-normalized UTCD dataset.

• Finetuned from model: bert-base-uncased

📄 License

This project is licensed under the MIT license.