micro1-inc/longextract-bench-50 download history
micro1-inc/longextract-bench-50 is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 665 times (108 in the last 7 days), and 2,216 times in total. It ranks #26,402 among datasets by monthly downloads.
LongExtractBench-50 A 50-document subset of LongExtractBench, a benchmark for schema-driven structured extraction from long, table-heavy PDFs. Each example pairs a real-world document with a target JSON Schema and a human-reconciled ground-truth extraction, so any extraction system —