This is a version of the bsmock/pubtables-1m dataset, where the XML annotations and JSON tokens have been merged and converted together into HTML table representations using the monoamine11231/pubtables-1m-to-html script.
The HTML table representations are stored with the original name prefix as in the original dataset. The HTML tables include the header (if existant) and the body part, as directed by the table annotation files.… See the full description on the dataset page:
https://huggingface.co/datasets/monoamine11231/pubtables-1m-html.
