chunked/nltk_chunking/model_doc_table-transformer/chunk_1.txt

The authors introduce a new dataset, PubTables-1M, to benchmark progress in table extraction from unstructured documents,
as well as table structure recognition and functional analysis.