Skip to page content
SAVRN
Search Plan a project

Dataset

Net-VQA

by Zhang yan8818/Net-VQA

Net-VQA is a visual question answering benchmark for evaluating multimodal models on network topology understanding and configuration reasoning.

Rows—
Configurations—
Size2.5 MB
Licensecc-by-4.0
AccessAccess requested at publisher
Monthly Downloads—

Dataset Card

By Zhang, published under cc-by-4.0, revision 0159e6a9c54a.

Net-VQA is a visual question answering benchmark for evaluating multimodal models on network topology understanding and configuration reasoning. The complete benchmark contains 100 enterprise network scenarios and 500 questions covering three tasks: network topology parsing, high-level configuration understanding, and configuration strategy generation. Each scenario includes a topology diagram and configuration intent, with questions and reference answers annotated and validated by networking experts.

The current release provides the scenarios approved for public disclosure. A small number of topology scenarios remain unavailable because their diagrams contain commercially sensitive information and require company approval before release. These scenarios will be added to the dataset once approval is obtained.

Details

Repository
yan8818/Net-VQA
Publisher
Zhang
Task category
Not stated by the source
Tags
Not stated by the source
Size category
Not stated by the source
Languages
Not stated by the source
Revision
0159e6a9c54a3c8d2b26273ae01f6d06d6c63540
Last updated
2026-09-28

Files

89 files, 2.5 MB in total.

Data1 file · 145.4 KB
Documentation1 file · 849 B
Other85 files · 2.3 MB
Repository2 files · 2.5 KB
Every file
FileTypeSizeSHA-256
Net-VQA.jsonData145.4 KB—
README.mdDocumentation849 B—
images/fig_dc_1.pngOther9.9 KB—
images/fig_dc_100.pngOther62.7 KB—
images/fig_dc_101.pngOther82.5 KB—
images/fig_dc_102.pngOther56.1 KB—
images/fig_dc_103.pngOther52.5 KB—
images/fig_dc_104.pngOther50.0 KB—
images/fig_dc_105.pngOther75.8 KB—
images/fig_dc_106.pngOther46.4 KB—
images/fig_dc_12.pngOther6.2 KB—
images/fig_dc_15.pngOther10.8 KB—
images/fig_dc_17.pngOther39.5 KB—
images/fig_dc_19.pngOther17.0 KB—
images/fig_dc_2.pngOther21.3 KB—
images/fig_dc_23.pngOther16.8 KB—
images/fig_dc_24.pngOther19.3 KB—
images/fig_dc_25.pngOther20.1 KB—
images/fig_dc_27.pngOther11.3 KB—
images/fig_dc_28.pngOther24.5 KB—
images/fig_dc_29.pngOther39.9 KB—
images/fig_dc_3.pngOther18.9 KB—
images/fig_dc_30.pngOther39.5 KB—
images/fig_dc_32.pngOther13.1 KB—
images/fig_dc_33.pngOther4.0 KB—
images/fig_dc_34.pngOther6.9 KB—
images/fig_dc_35.pngOther12.3 KB—
images/fig_dc_36.pngOther7.5 KB—
images/fig_dc_37.pngOther4.9 KB—
images/fig_dc_38.pngOther18.1 KB—
images/fig_dc_39.pngOther23.1 KB—
images/fig_dc_4.pngOther15.2 KB—
images/fig_dc_40.pngOther18.9 KB—
images/fig_dc_41.pngOther10.8 KB—
images/fig_dc_42.pngOther9.8 KB—
images/fig_dc_43.pngOther10.5 KB—
images/fig_dc_46.pngOther11.0 KB—
images/fig_dc_47.pngOther17.8 KB—
images/fig_dc_48.pngOther19.9 KB—
images/fig_dc_5.pngOther16.8 KB—
images/fig_dc_50.pngOther18.7 KB—
images/fig_dc_51.pngOther11.1 KB—
images/fig_dc_52.pngOther40.7 KB—
images/fig_dc_53.pngOther30.4 KB—
images/fig_dc_54.pngOther39.7 KB—
images/fig_dc_55.pngOther40.7 KB—
images/fig_dc_56.pngOther31.9 KB—
images/fig_dc_57.pngOther22.5 KB—
images/fig_dc_58.pngOther4.9 KB—
images/fig_dc_59.pngOther36.4 KB—
images/fig_dc_60.pngOther19.5 KB—
images/fig_dc_61.pngOther11.3 KB—
images/fig_dc_62.pngOther12.5 KB—
images/fig_dc_63.pngOther22.6 KB—
images/fig_dc_64.pngOther17.3 KB—
images/fig_dc_65.pngOther18.9 KB—
images/fig_dc_66.pngOther21.3 KB—
images/fig_dc_67.pngOther20.8 KB—
images/fig_dc_68.pngOther21.6 KB—
images/fig_dc_69.pngOther19.8 KB—
images/fig_dc_7.pngOther23.1 KB—
images/fig_dc_70.pngOther19.6 KB—
images/fig_dc_71.pngOther10.5 KB—
images/fig_dc_72.pngOther21.5 KB—
images/fig_dc_73.pngOther39.9 KB—
images/fig_dc_74.pngOther15.5 KB—
images/fig_dc_75.pngOther17.9 KB—
images/fig_dc_76.pngOther22.8 KB—
images/fig_dc_78.pngOther23.9 KB—
images/fig_dc_79.pngOther18.9 KB—
images/fig_dc_8.pngOther10.4 KB—
images/fig_dc_81.pngOther19.1 KB—
images/fig_dc_84.pngOther7.1 KB—
images/fig_dc_86.pngOther23.8 KB—
images/fig_dc_87.pngOther37.0 KB—
images/fig_dc_88.pngOther36.7 KB—
images/fig_dc_89.pngOther34.2 KB—
images/fig_dc_90.pngOther55.0 KB—
images/fig_dc_91.pngOther40.1 KB—
images/fig_dc_92.pngOther50.1 KB—
images/fig_dc_93.pngOther30.9 KB—
images/fig_dc_94.pngOther49.0 KB—
images/fig_dc_95.pngOther49.2 KB—
images/fig_dc_96.pngOther57.3 KB—
images/fig_dc_97.pngOther63.2 KB—
images/fig_dc_98.pngOther56.6 KB—
images/fig_dc_99.pngOther65.4 KB—
.gitattributesRepository2.5 KB—
images/.gitkeepRepository—

License and Download

License
cc-by-4.0
Access
Access requested at publisher
Request access from Zhang

Zhang grants access through its official repository on Hugging Face. Read the license.