Abstract
Dataset distillation is a process aimed at condensing datasets while preserving essential characteristics. In the text domain, prevailing methods typically generate distilled data as embedding vectors, which are not human-readable. This approach simplifies optimization but limits the transferability of distilled data across different model architectures. To address this limitation, we introduce a model-agnostic, data-efficient method that leverages Language Model (LM) embeddings. Compared to parameter-efficient methods such as LORA, our approach achieves comparable performance with significantly faster processing times. We evaluate our methodology through classification tasks on datasets like IMDB and AG-News, demonstrating performance that is on par with or exceeds previous model-dependent techniques. By utilizing LM embeddings, our method offers enhanced flexibility and improved transferability, expanding the range of potential applications.- Anthology ID:
- 2024.findings-emnlp.733
- Volume:
- Findings of the Association for Computational Linguistics: EMNLP 2024
- Month:
- November
- Year:
- 2024
- Address:
- Miami, Florida, USA
- Editors:
- Yaser Al-Onaizan, Mohit Bansal, Yun-Nung Chen
- Venue:
- Findings
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 12557–12569
- Language:
- URL:
- https://aclanthology.org/2024.findings-emnlp.733
- DOI:
- 10.18653/v1/2024.findings-emnlp.733
- Cite (ACL):
- Yefan Tao, Luyang Kong, Andrey Kan, and Laurent Callot. 2024. Textual Dataset Distillation via Language Model Embedding. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 12557–12569, Miami, Florida, USA. Association for Computational Linguistics.
- Cite (Informal):
- Textual Dataset Distillation via Language Model Embedding (Tao et al., Findings 2024)
- PDF:
- https://preview.aclanthology.org/dois-2013-emnlp/2024.findings-emnlp.733.pdf