Efficient Title Reranker for Fast and Improved Knowledge-Intense NLP
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2023
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866914691438084096 |
|---|---|
| author | Chen, Ziyi Jiang, Jize Zuo, Daqian Tao, Heyi Yang, Jun Wei, Yuxiang |
| author_facet | Chen, Ziyi Jiang, Jize Zuo, Daqian Tao, Heyi Yang, Jun Wei, Yuxiang |
| contents | In recent RAG approaches, rerankers play a pivotal role in refining retrieval accuracy with the ability of revealing logical relations for each pair of query and text. However, existing rerankers are required to repeatedly encode the query and a large number of long retrieved text. This results in high computational costs and limits the number of retrieved text, hindering accuracy. As a remedy of the problem, we introduce the Efficient Title Reranker via Broadcasting Query Encoder, a novel technique for title reranking that achieves a 20x-40x speedup over the vanilla passage reranker. Furthermore, we introduce Sigmoid Trick, a novel loss function customized for title reranking. Combining both techniques, we empirically validated their effectiveness, achieving state-of-the-art results on all four datasets we experimented with from the KILT knowledge benchmark. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2312_12430 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Efficient Title Reranker for Fast and Improved Knowledge-Intense NLP Chen, Ziyi Jiang, Jize Zuo, Daqian Tao, Heyi Yang, Jun Wei, Yuxiang Information Retrieval Artificial Intelligence Computation and Language Machine Learning In recent RAG approaches, rerankers play a pivotal role in refining retrieval accuracy with the ability of revealing logical relations for each pair of query and text. However, existing rerankers are required to repeatedly encode the query and a large number of long retrieved text. This results in high computational costs and limits the number of retrieved text, hindering accuracy. As a remedy of the problem, we introduce the Efficient Title Reranker via Broadcasting Query Encoder, a novel technique for title reranking that achieves a 20x-40x speedup over the vanilla passage reranker. Furthermore, we introduce Sigmoid Trick, a novel loss function customized for title reranking. Combining both techniques, we empirically validated their effectiveness, achieving state-of-the-art results on all four datasets we experimented with from the KILT knowledge benchmark. |
| title | Efficient Title Reranker for Fast and Improved Knowledge-Intense NLP |
| topic | Information Retrieval Artificial Intelligence Computation and Language Machine Learning |
| url | https://arxiv.org/abs/2312.12430 |