Ziheng Qin

Preventing Zero-Shot Transfer Degradation in Continual Learning of Vision-Language Models

2023 · arXiv (Cornell University)

Continual learning (CL) can help pre-trained vision-language models efficiently adapt to new or under-trained data distributions without re-training. Nevertheless, during the continual training of the Contrastive Language-Image Pre-training (CLIP) model, we observe that the model's …