Switch-KD: Visual-Switch Knowledge Distillation for Vision-Language Models
Open paper
Benchmark
Model
Rank
Results
object-hallucination-evaluation-on-pope
Switch-KD 0.5B
#15
F1: 87.3
multimodal-reasoning-on-mmmu
Switch-KD 1.5B
#130
Accuracy: 34.9
Rank counts only results with a code link.