T\"urk\c{c}e Dil Modellerinin Performans Kar\c{s}{\i}la\c{s}t{\i}rmas{\i} Performance Comparison of Turkish Language Models

TL;DR

This paper compares seven Turkish language models on their contextual learning and question-answering abilities, highlighting the effectiveness of continued pretraining for multilingual models and analyzing in-context learning performance.

Contribution

It provides the first comprehensive performance comparison of open-source Turkish language models using new datasets and evaluation methods.

Findings

Continued pretraining improves question-answering in Turkish models.

In-context learning performance is not strongly related to question-answering success.

Multilingual models benefit from additional pretraining before fine-tuning.

Abstract

The developments that language models have provided in fulfilling almost all kinds of tasks have attracted the attention of not only researchers but also the society and have enabled them to become products. There are commercially successful language models available. However, users may prefer open-source language models due to cost, data privacy, or regulations. Yet, despite the increasing number of these models, there is no comprehensive comparison of their performance for Turkish. This study aims to fill this gap in the literature. A comparison is made among seven selected language models based on their contextual learning and question-answering abilities. Turkish datasets for contextual learning and question-answering were prepared, and both automatic and human evaluations were conducted. The results show that for question-answering, continuing pretraining before fine-tuning with instructional datasets is more successful in adapting multilingual models to Turkish and that in-context learning performances do not much related to question-answering performances.