Computer Science > Information Retrieval
arXiv:2212.05225 (cs)
[Submitted on 10 Dec 2022 (v1), last revised 11 Dec 2023 (this version, v2)]
Title:LEAD: Liberal Feature-based Distillation for Dense Retrieval
Authors:Hao Sun,Xiao Liu,Yeyun Gong,Anlei Dong,Jingwen Lu,Yan Zhang,Linjun Yang,Rangan Majumder,Nan Duan
View a PDF of the paper titled LEAD: Liberal Feature-based Distillation for Dense Retrieval, by Hao Sun and 8 other authors
View PDFHTML (experimental)Abstract:Knowledge distillation is often used to transfer knowledge from a strong teacher model to a relatively weak student model. Traditional methods include response-based methods and feature-based methods. Response-based methods are widely used but suffer from lower upper limits of performance due to their ignorance of intermediate signals, while feature-based methods have constraints on vocabularies, tokenizers and model architectures. In this paper, we propose a liberal feature-based distillation method (LEAD). LEAD aligns the distribution between the intermediate layers of teacher model and student model, which is effective, extendable, portable and has no requirements on vocabularies, tokenizers, or model architectures. Extensive experiments show the effectiveness of LEAD on widely-used benchmarks, including MS MARCO Passage Ranking, TREC 2019 DL Track, MS MARCO Document Ranking and TREC 2020 DL Track. Our code is available inthis https URL.
| Comments: | Accepted by WSDM 2024 |
| Subjects: | Information Retrieval (cs.IR); Computation and Language (cs.CL) |
| Cite as: | arXiv:2212.05225 [cs.IR] |
| (orarXiv:2212.05225v2 [cs.IR] for this version) | |
| https://doi.org/10.48550/arXiv.2212.05225 arXiv-issued DOI via DataCite |
Submission history
From: Hao Sun [view email][v1] Sat, 10 Dec 2022 06:30:54 UTC (267 KB)
[v2] Mon, 11 Dec 2023 09:41:29 UTC (105 KB)
Full-text links:
Access Paper:
- View PDF
- HTML (experimental)
- TeX Source
- Other Formats
View a PDF of the paper titled LEAD: Liberal Feature-based Distillation for Dense Retrieval, by Hao Sun and 8 other authors
References & Citations
Bookmark
Bibliographic and Citation Tools
Bibliographic Explorer(What is the Explorer?)
Connected Papers(What is Connected Papers?)
Litmaps(What is Litmaps?)
scite Smart Citations(What are Smart Citations?)
Code, Data and Media Associated with this Article
alphaXiv(What is alphaXiv?)
CatalyzeX Code Finder for Papers(What is CatalyzeX?)
DagsHub(What is DagsHub?)
Gotit.pub(What is GotitPub?)
Hugging Face(What is Huggingface?)
Papers with Code(What is Papers with Code?)
ScienceCast(What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower(What are Influence Flowers?)
CORE Recommender(What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community?Learn more about arXivLabs.