Skill by nvidia

tao-train-nvdinov2

NVDINOv2 for self-supervised visual representation learning. Trains vision transformers via self-distillation (teacher-student) without labels and produces general-purpose visual features. Use when training, distilling, exporting, or running inference for