Repack parquet into 24 MiB row groups so the dataset viewer can read it

Dataset card written from measured statistics

Upload README.md with huggingface_hub

Upload data/test-00000-of-00001-ac70afe74ea0c516.parquet with huggingface_hub

Upload data/train-00001-of-00002-62025f5a6bb59e25.parquet with huggingface_hub

Upload data/train-00000-of-00002-ffd05059166d8ef1.parquet with huggingface_hub

initial commit