Delete train/news_noisy_0001.parquet

Splitted the noisy dataset

Delete old clean dataset

Splitted version of clean dataset

Added clean version on the dataset filtered with lang_id == __label__tat_Cyrl and MinHashLSH (256, 0.85)

Update README.md

Update README.md

Added noisy not filtered data parquet

Upload news_ideal_0001.parquet

Delete train

Update README.md

Create README.md

First version of the dataset

initial commit