LinguDistill: Recovering Linguistic Ability in Vision- Language Models via Selective Cross-Modal Distillation
Open paper
Benchmark
Model
Rank
Results
sentence-completion-on-hellaswag
LinguDistill
–
Accuracy: 39.4
vision-language-understanding-on-mmstar
LinguDistill
–
Accuracy: 34.9
visual-question-answering-vqa-on-realworldqa
LinguDistill
–
Accuracy: 49.0
Rank counts only results with a code link.