A group of artificial intelligence experts at Iran’s Amirkabir University of Technology has developed an evaluation system for Persian language models, considered the most comprehensive of its kind.
The system aims to equip Iran’s AI arena with an accurate metric for assessing large language models. Saeedeh Momtazi, a member of the Academic Board at the university’s Faculty of Computer Engineering, mentioned that experts from the Part Artificial Intelligence Research Center and the Natural Language Processing Laboratory at Amirkabir University have developed the most powerful evaluation system for Persian language models.
The newly-developed evaluation system includes over 40,000 samples, some translated from reputable global benchmarks into Persian, while others have been created domestically. Momtazi also noted that the evaluation dataset is continuously growing to enhance assessment accuracy.
The model presented in this project is as valuable as top global projects, with a segment of the dataset available to the public as open-source.