Image-to-Text Retrieval on Flickr30k (TR@1)
0.893TR@1TuneCLIP
Evaluation Results
| Method | Links | |
|---|---|---|
| TuneCLIPBase Model=SigLIP ViT-B/162026.01 | 0.893 | |
| BaselineBase Model=SigLIP ViT-B/162026.01 | 0.891 | |
| TuneCLIPBase Model=OpenAI ViT-B/162026.01 | 0.8659 | |
| BaselineBase Model=LAION ViT-B/322026.01 | 0.841 | |
| TuneCLIPBase Model=OpenAI ViT-B/322026.01 | 0.833 | |
| TuneCLIPBase Model=LAION ViT-B/322026.01 | 0.832 | |
| BaselineBase Model=OpenAI ViT-B/162026.01 | 0.822 | |
| OpenCLIPBase Model=LAION ViT-B/322026.01 | 0.813 | |
| BaselineBase Model=OpenAI ViT-B/322026.01 | 0.789 | |
| OpenCLIPBase Model=OpenAI ViT-B/162026.01 | 0.789 | |
| FastCLIPBase Model=LAION ViT-B/322026.01 | 0.7699 | |
| OpenCLIPBase Model=OpenAI ViT-B/322026.01 | 0.761 | |
| FastCLIPBase Model=OpenAI ViT-B/162026.01 | 0.74 | |
| OpenCLIPBase Model=SigLIP ViT-B/162026.01 | 0.701 | |
| FastCLIPBase Model=OpenAI ViT-B/322026.01 | 0.6819 | |
| FastCLIPBase Model=SigLIP ViT-B/162026.01 | 0.6179 |