Tevatron: An Efficient and Flexible Toolkit for Dense Retrieval

TL;DR

Tevatron is a versatile, efficient, and user-friendly toolkit for dense retrieval that supports various datasets, models, and hardware platforms, facilitating research and development in information retrieval and question answering.

Contribution

The paper introduces Tevatron, a standardized, flexible, and efficient dense retrieval toolkit that simplifies implementation and experimentation across different datasets, models, and hardware.

Findings

Tevatron achieves high efficiency in dense retrieval tasks.

It demonstrates strong performance across multiple IR and QA datasets.

The toolkit's flexible design enables easy adaptation to various models and hardware.

Abstract

Recent rapid advancements in deep pre-trained language models and the introductions of large datasets have powered research in embedding-based dense retrieval. While several good research papers have emerged, many of them come with their own software stacks. These stacks are typically optimized for some particular research goals instead of efficiency or code structure. In this paper, we present Tevatron, a dense retrieval toolkit optimized for efficiency, flexibility, and code simplicity. Tevatron provides a standardized pipeline for dense retrieval including text processing, model training, corpus/query encoding, and search. This paper presents an overview of Tevatron and demonstrates its effectiveness and efficiency across several IR and QA data sets. We also show how Tevatron's flexible design enables easy generalization across datasets, model architectures, and accelerator platforms(GPU/TPU). We believe Tevatron can serve as an effective software foundation for dense retrieval system research including design, modeling, and optimization.