tandevllc/cve_dataset
TanDev CVE Dataset (NVD SlimPajama Corpus) A SlimPajama‑style corpus of CVE entries derived from the NIST NVD (CVE 2.0) data feeds (2002→present). Each row is a cleaned, single‑document text representation of a CVE with structured metadata for CVSS, CWE(s), timestamps, and a canonical NVD link—ready for pretraining/continued‑pretraining, RAG, retrieval/evaluation, and downstream classifiers. Token count: ~108.2M tokens. License: "TanDev Proprietary License — All Rights Reserved"… See the full description on the dataset page: https://huggingface.co/datasets/tandevllc/cve_dataset.
This repository is gated, so its file contents are only served once you have accepted the publisher's terms at Hugging Face. Open it at the source above.