Datasetpublic

mhurhangee/us-patent-descriptions

US Patent Descriptions This dataset contains the descriptions of granted US utility patents, filtered and deduplicated.The original data comes from all granted patents in 2025 up to May 20, available from PatentsView. Splits train: 10,000 rows for model training validation: 2,500 rows for validation test: 2,500 rows for evaluation Columns patent_id: Identifier for the patent; useful for reconciling with other PatentsView datasets… See the full description on the dataset page: https://huggingface.co/datasets/mhurhangee/us-patent-descriptions.

sourceHugging Faceupdated 1y agoView on Hugging Face

0likes91downloads

Dataset Card

US Patent Descriptions

This dataset contains the descriptions of granted US utility patents, filtered and deduplicated. The original data comes from all granted patents in 2025 up to May 20, available from PatentsView.

Splits

—train: 10,000 rows for model training
—validation: 2,500 rows for validation
—test: 2,500 rows for evaluation

Columns

—patent_id: Identifier for the patent; useful for reconciling with other PatentsView datasets
—description_text: Full text of the patent description
—description_length: Character or token length of the description

Notes

—The dataset has been deduplicated to remove exact and near-duplicate descriptions.
—Very long and very short descriptions (outside the 1st and 95th percentile) were removed to reduce outliers.
—Use this dataset for NLP experiments, such as summarization, classification, or sentence scoring.

Usage Example

python

from datasets import load_dataset

ds = load_dataset("mhurhangeee/us-patent-descriptions")
train_data = ds["train"]