TurkEmbed: Turkish Embedding Model on NLI & STS Tasks

TL;DR

TurkEmbed is a new Turkish language embedding model that significantly improves performance on NLI and STS tasks by utilizing diverse datasets and advanced training techniques, surpassing previous models in accuracy and efficiency.

Contribution

The paper introduces TurkEmbed, a novel Turkish embedding model that outperforms existing models on NLI and STS tasks through innovative training methods and dataset utilization.

Findings

TurkEmbed achieves 1-4% improvement over state-of-the-art models.

The model demonstrates robust performance on Turkish NLI and STS benchmarks.

TurkEmbed offers faster encoding suitable for resource-constrained environments.

Abstract

This paper introduces TurkEmbed, a novel Turkish language embedding model designed to outperform existing models, particularly in Natural Language Inference (NLI) and Semantic Textual Similarity (STS) tasks. Current Turkish embedding models often rely on machine-translated datasets, potentially limiting their accuracy and semantic understanding. TurkEmbed utilizes a combination of diverse datasets and advanced training techniques, including matryoshka representation learning, to achieve more robust and accurate embeddings. This approach enables the model to adapt to various resource-constrained environments, offering faster encoding capabilities. Our evaluation on the Turkish STS-b-TR dataset, using Pearson and Spearman correlation metrics, demonstrates significant improvements in semantic similarity tasks. Furthermore, TurkEmbed surpasses the current state-of-the-art model, Emrecan, on All-NLI-TR and STS-b-TR benchmarks, achieving a 1-4\% improvement. TurkEmbed promises to enhance the Turkish NLP ecosystem by providing a more nuanced understanding of language and facilitating advancements in downstream applications.