pinialt/

cefr-texts-10languages

	Gold Standard CEFR Validation Dataset




	
	Dataset Summary

This dataset is a high-quality synthetic validation set designed to evaluate models on CEFR (Common European Framework of Reference for Languages) Level Classification. The dataset was generated using OpenAI's GPT-4o-mini. It contains approximately 3,000 examples balanced across 10 languages and 6 proficiency levels.

	Dataset Structure




	
	Data Fields

Each entry in the dataset consists of the… See the full description on the dataset page: https://huggingface.co/datasets/pinialt/cefr-texts-10languages.