Independent publisher

Dong Yang

yangd05

Models in Library0

Datasets in Library1

Models on Hugging Face—

Followers3

Compare the published evidence

This profile connects the 1 model and dataset record shown below to their upstream cards and sources. The library counts describe SAVRN's coverage, rather than this publisher's entire output.

Published tasks: Visual question answering.

Compare model specifications, source dates, and hardware estimates on the individual records. Publisher identity verification is separate from a model's quality, deployment readiness, or suitability for your work.

Datasets

Dataset · Visual question answering

SurgVLM-Video-Benchmark

Dong Yang

Held-out video eval set for SurgVLM-style surgical video tasks. Each QA row is paired with an MP4 clip (the listed frames encoded in order). - 36 tasks, 200 QA rows each = 7,200 rows - 6,512 MP4 clips (rows can share a clip) videopath is relative to this repo root, e.g. media/GraSP/GraSPCASE041002476-002485n10.mp4. Clip identity is mediauid (SHA-256 of {imagepaths, clipdurationsec, fps}). Multiple QA rows can share one clip. If you hit Xet Runtime Error / Task Panic, retry with HFHUBDISABLEXET=1.

Access requested at publisher 1K<n<10K

View dataset