Mini-o3/VisualProbe_train
Viewer • Updated • 2.9k • 1.3k • 3
This model is the result of our paper Learning to Zoom Efficiently with a Contrastive Curriculum. Concretely, it is the model that we refer as 'Ours+' in the paper. Please use it together with the evaluation code on Github.
Base model
Qwen/Qwen2.5-VL-7B-Instruct