The Dataset Viewer has been disabled on this dataset.
FDABench-File
Raw multimodal corpus distributed alongside the FDABench benchmark.
Contains 1,418 PDFs, 246 images, 108 audio clips, 81 video clips, plus transcripts and a manifest, organized into 50 domain categories.
Use huggingface_hub.snapshot_download("FDAbench2026/FDABench-File", repo_type="dataset") to fetch the full tree, or
hf_hub_download(...) for individual files.
- Downloads last month
- 91