A Memory Efficient Baseline for Open Domain Question Answering
arXiv:2012.15156
Abstract
Recently, retrieval systems based on dense representations have led to important improvements in open-domain question answering, and related tasks. While very effective, this approach is also memory intensive, as the dense vectors for the whole knowledge source need to be kept in memory. In this paper, we study how the memory footprint of dense retriever-reader systems can be reduced. We consider three strategies to reduce the index size: dimension reduction, vector quantization and passage filtering. We evaluate our approach on two question answering benchmarks: TriviaQA and NaturalQuestions, showing that it is possible to get competitive systems using less than 6Gb of memory.
References in corpus (7)
- Decoupled Weight Decay Regularization
- Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
- Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
- REALM: Retrieval-Augmented Language Model Pre-Training
- Sparse, Dense, and Attentional Representations for Text Retrieval
- Distilling Knowledge from Reader to Retriever for Question Answering
- AmbigQA: Answering Ambiguous Open-domain Questions
Cited by in corpus (6)
- Pretrained Transformers for Text Ranking: BERT and Beyond
- NeurIPS 2020 EfficientQA Competition: Systems, Analyses and Lessons Learned
- RecJPQ: Training Large-Catalogue Sequential Recommenders
- Efficient Passage Retrieval with Hashing for Open-domain Question Answering
- Highly Parallel Autoregressive Entity Linking with Discriminative Correction
- Designing a Minimal Retrieve-and-Read System for Open-Domain Question Answering