ViCO: A Training Strategy towards Semantic Aware Dynamic High-Resolution
Paper
•
2510.12793
•
Published
•
4
InternVL3.5-Flash
InternVL3.5-Flash is a fast variant of InternVL3.5 using semantic aware dynamic high-resolution strategy.
-
-
OpenGVLab/InternVL3_5-241B-A28B-Flash
Image-Text-to-Text • 242B • Updated • 25 • 7 -
OpenGVLab/InternVL3_5-38B-Flash
Image-Text-to-Text • 40B • Updated • 168 • 6 -
OpenGVLab/InternVL3_5-30B-A3B-Flash
Image-Text-to-Text • 31B • Updated • 109 • 7 -
OpenGVLab/InternVL3_5-14B-Flash
Image-Text-to-Text • 15B • Updated • 5.79k • 6 -
OpenGVLab/InternVL3_5-8B-Flash
Image-Text-to-Text • 9B • Updated • 724 • 4 -
OpenGVLab/InternVL3_5-4B-Flash
Image-Text-to-Text • 5B • Updated • 477 • 3 -
OpenGVLab/InternVL3_5-2B-Flash
Image-Text-to-Text • Updated • 210 • 5 -
OpenGVLab/InternVL3_5-1B-Flash
Image-Text-to-Text • 1B • Updated • 2.96k • 7