SPLADE for PyTerrier¶

pyterrier-splade is a PyTerrier extension that provides SPLADE learned sparse indexing and retrieval. SPLADE encodes queries and documents into sparse bags of weighted BERT WordPiece tokens, which can be indexed and searched using standard sparse retrieval backends such as Terrier and PISA. It was introduced in the following article:

Citation

Formal et al. SPLADE: Sparse Lexical and Expansion Model for First Stage Ranking. SIGIR 2021. [link]

@inproceedings{DBLP:conf/sigir/FormalPC21,
  author       = {Thibault Formal and
                  Benjamin Piwowarski and
                  St{\'{e}}phane Clinchant},
  editor       = {Fernando Diaz and
                  Chirag Shah and
                  Torsten Suel and
                  Pablo Castells and
                  Rosie Jones and
                  Tetsuya Sakai},
  title        = {{SPLADE:} Sparse Lexical and Expansion Model for First Stage Ranking},
  booktitle    = {{SIGIR} '21: The 44th International {ACM} {SIGIR} Conference on Research
                  and Development in Information Retrieval, Virtual Event, Canada, July
                  11-15, 2021},
  pages        = {2288--2292},
  publisher    = {{ACM}},
  year         = {2021},
  url          = {https://doi.org/10.1145/3404835.3463098},
  doi          = {10.1145/3404835.3463098},
  timestamp    = {Sun, 02 Nov 2025 21:27:20 +0100},
  biburl       = {https://dblp.org/rec/conf/sigir/FormalPC21.bib},
  bibsource    = {dblp computer science bibliography, https://dblp.org}
}

You can install this extension with pip:

pip install pyterrier-splade

Contents

Demo¶

We have a demo of PyTerrier SPLADE at https://huggingface.co/spaces/terrierteam/splade

Credits¶

Craig Macdonald

Sean MacAvaney

Nicola Tonellotto