Jina Embeddings: A Novel Set of High-Performance Sentence Embedding Models

TL;DR

Jina Embeddings introduces high-performance sentence models optimized for semantic understanding, developed through meticulous data preparation and evaluated on extensive benchmarks, including a novel negation-aware dataset.

Contribution

The paper presents a new set of sentence embedding models with improved semantic capture, emphasizing data quality and introducing a negation-aware dataset for better linguistic understanding.

Findings

High performance on the MTEB benchmark

Effective handling of negated statements

Public availability of negation dataset

Abstract

Jina Embeddings constitutes a set of high-performance sentence embedding models adept at translating textual inputs into numerical representations, capturing the semantics of the text. These models excel in applications like dense retrieval and semantic textual similarity. This paper details the development of Jina Embeddings, starting with the creation of high-quality pairwise and triplet datasets. It underlines the crucial role of data cleaning in dataset preparation, offers in-depth insights into the model training process, and concludes with a comprehensive performance evaluation using the Massive Text Embedding Benchmark (MTEB). Furthermore, to increase the model's awareness of grammatical negation, we construct a novel training and evaluation dataset of negated and non-negated statements, which we make publicly available to the community.