Docs

How to use

Browse models and datasets by department, domain, category (task) and access; sort by downloads or size.
Search: Chinese medical terms are expanded to English; any keyword may match; ranked over full card text.
Details: model and dataset pages show the authors' original cards; dataset pages also state provenance and the original release.
Request: call hub_request(entry_id, reason). Staff download manually; after requesting, pause and report.
Agents open any page with hub_browse(url); links in a page can be passed as url.

Pull only: agents obtain models from the hub; models trained on private hospital data are never uploaded.
Reviewed: every entry is reviewed by staff (license, compliance) before agents can see it.
Pinned: each entry is pinned to an exact commit, so a download always matches what was reviewed.
Snapshots: each experiment round fixes one snapshot so all agents see the same resources.
Manual downloads: after requesting, agents pause; staff download, then agents resume.
Dataset provenance: Provenance: official = uploaded by the dataset authors/institution; mirror = unmodified third-party copy; processed = third-party subset, slices or format conversion (may differ from the original). Licenses are as declared by the repository; staff check the original license of third-party copies before download.
Not included: third-party copies of credentialed data that require a separate data use agreement (e.g. MIMIC, VinDr, eICU), and copies with unclear origin or license.