Text-to-Image Generation on Anime Domain Dataset
38.2932FIDD-OPSD
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| D-OPSDModel Architecture=FLUX.2-klein, Fine-tuning Protocol=Full-finetuning2026.05 | 38.2932 | 0.0902 | 0.6476 | 3.8328 | 3.5511 | |
| D-OPSDModel Architecture=Z-Image-Turbo, Fine-tuning Protocol=Full-finetuning2026.05 | 40.4938 | 0.1088 | 0.6419 | 3.8438 | 3.6195 | |
| Base ModelModel Architecture=FLUX.2-klein, Fine-tuning Protocol=None2026.05 | 45.2335 | 0.121 | 0.6894 | 3.7626 | 3.4465 | |
| Base-ModelModel Architecture=Z-Image-Turbo, Fine-tuning Protocol=None2026.05 | 48.6858 | 0.1274 | 0.7036 | 3.7624 | 3.5218 | |
| PSOModel Architecture=FLUX.2-klein, Fine-tuning Protocol=Full-finetuning2026.05 | 82.1438 | 0.2185 | 0.6511 | 2.8066 | 2.5952 | |
| Vanilla SFTModel Architecture=Z-Image-Turbo, Fine-tuning Protocol=Full-finetuning2026.05 | 82.2036 | 0.1896 | 0.6787 | 2.6121 | 2.4852 | |
| PSOModel Architecture=Z-Image-Turbo, Fine-tuning Protocol=Full-finetuning2026.05 | 88.4343 | 0.1716 | 0.653 | 2.8653 | 2.6424 | |
| Vanilla SFTModel Architecture=FLUX.2-klein, Fine-tuning Protocol=Full-finetuning2026.05 | 98.7893 | 0.2565 | 0.6567 | 2.5038 | 2.4121 |