ChamaraVishwajithRajapaksha/

cnn-dailymail-sinhala-continuous-pretrain

	CNN DailyMail Sinhala Continuous Pretraining Dataset




	
	Dataset Description

This dataset is designed for continuous pretraining of Sinhala Small Language Models (SLMs) and Large Language Models (LLMs). The dataset was created by processing the original Sinhala news articles from:

CNN Daily Mail Sinhala Dataset

The article_sinhala field from the original dataset was extracted, cleaned, and concatenated into larger continuous text blocks suitable for language model… See the full description on the dataset page: https://huggingface.co/datasets/ChamaraVishwajithRajapaksha/cnn-dailymail-sinhala-continuous-pretrain.