search

Yufang-Liu / clip_hallucination

[EMNLP 2024] Investigating and Mitigating Object Hallucinations in Pretrained Vision-Language (CLIP) Models
5 stars 0 forks source link

About Checkpoints #1

Closed haohaodw closed 2 weeks ago

haohaodw commented 2 weeks ago

Very interesting work. I noticed that you published the checkpoint, and there is a subfolder called CLIP under the huggingface link. Does this CLIP refer to the full fine-tuning CLIP mentioned in your paper, or is it the original CLIP?

Yufang-Liu commented 2 weeks ago

Very interesting work. I noticed that you published the checkpoint, and there is a subfolder called CLIP under the huggingface link. Does this CLIP refer to the full fine-tuning CLIP mentioned in your paper, or is it the original CLIP?

It is the fine-tuned model.

haohaodw commented 2 weeks ago

Thanks for your reply.