Datasets:
Tasks:
Image-Text-to-Text
Modalities:
Image
Formats:
imagefolder
Languages:
English
Size:
< 1K
ArXiv:
License:
File size: 639 Bytes
b96feb3 | 1 2 3 4 5 6 7 8 | # Attribution
Visual Scaffolding code: Copyright 2026 Visual Scaffolding contributors. Distributed under Apache License 2.0.
The pretrained models are derived from Qwen3-VL-4B-Thinking by the Qwen team. Backbone source: https://huggingface.co/Qwen/Qwen3-VL-4B-Thinking. Backbone license: Apache License 2.0. The release adds learned visual retrieval and evidence adaptation modules and includes task-adapted model weights.
Benchmark artwork includes OpenMoji and Twemoji. Their authors, source URLs, licenses, and transformations are identified in DATA_LICENSE.md. Their graphics licenses remain applicable to the incorporated artwork.
|