No Parameter Left Behind: How Distillation and Model Size Affect Zero-Shot Retrieval
Guilherme Moraes Rosa(Universidade Estadual de Campinas (UNICAMP)), Rodrigo Nogueira(United Food and Commercial Workers), Luiz Bonifacio, Hugo Abonizio, Marzieh Fadaee, Vitor Jeronymo, Roberto Lotufo(Universidade Estadual de Campinas (UNICAMP))
Cited by 6
Related Papers
BERTimbau: Pretrained BERT Models for Brazilian Portuguese
|Lecture notes in computer science|2020|611
Document Ranking with a Pretrained Sequence-to-Sequence Model
|Unknown|2020|434
Fingerprint Liveness Detection Using Convolutional Neural Networks
|IEEE Transactions on Information Forensics and Security|2016|357
Pyserini: A Python Toolkit for Reproducible Information Retrieval Research with Sparse and Dense Representations
|Unknown|2021|355
Passage Re-ranking with BERT
|arXiv (Cornell University)|2019|346