TopGuNN: Fast NLP Training Data Augmentation using Large Corpora
Rebecca Iglesias-Flores, Megha Mishra, Ajay Patel, Akanksha Malhotra, Reno Kriz, Martha Palmer, Chris Callison-Burch
Abstract
Acquiring training data for natural language processing systems can be expensive and time-consuming. Given a few training examples crafted by experts, large corpora can be mined for thousands of semantically similar examples that provide useful variability to improve model generalization. We present TopGuNN, a fast contextualized k-NN retrieval system that can efficiently index and search over contextual embeddings generated from large corpora. TopGuNN is demonstrated for a training data augmentation use case over the Gigaword corpus. Using approximate k-NN and an efficient architecture, TopGuNN performs queries over an embedding space of 4.63TB (approximately 1.5B embeddings) in less than a day.- Anthology ID:
- 2021.dash-1.14
- Volume:
- Proceedings of the Second Workshop on Data Science with Human in the Loop: Language Advances
- Month:
- June
- Year:
- 2021
- Address:
- Online
- Editors:
- Eduard Dragut, Yunyao Li, Lucian Popa, Slobodan Vucetic
- Venue:
- DaSH
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 86–101
- Language:
- URL:
- https://aclanthology.org/2021.dash-1.14
- DOI:
- 10.18653/v1/2021.dash-1.14
- Cite (ACL):
- Rebecca Iglesias-Flores, Megha Mishra, Ajay Patel, Akanksha Malhotra, Reno Kriz, Martha Palmer, and Chris Callison-Burch. 2021. TopGuNN: Fast NLP Training Data Augmentation using Large Corpora. In Proceedings of the Second Workshop on Data Science with Human in the Loop: Language Advances, pages 86–101, Online. Association for Computational Linguistics.
- Cite (Informal):
- TopGuNN: Fast NLP Training Data Augmentation using Large Corpora (Iglesias-Flores et al., DaSH 2021)
- PDF:
- https://preview.aclanthology.org/ingest-2024-clasp/2021.dash-1.14.pdf
- Code
- penn-topgunn/topgunn