Text-to-Image Generation on COCO 10K
0.3224CLIP ScoreTDM*
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| TDM*Model=Stable Diffusion 3.5 Large, Step=2, #NFE=2, Resolution=1024x1024, Img-free=true2026.04 | 0.3224 | 26.8084 | 5.311 | 22.1424 | 0.9307 | 28.4919 | |
| 1.x-Distill-slowModel=Stable Diffusion 3 Medium, Step=2, #NFE=1.75, Resolution=1024x1024, Img-free=true2026.04 | 0.3208 | 15.7863 | 5.1844 | 22.5161 | 1.1312 | 32.255 | |
| 1.x-Distill-fastModel=Stable Diffusion 3 Medium, Step=2, #NFE=1.67, Resolution=1024x1024, Img-free=true2026.04 | 0.3204 | 16.7179 | 5.1206 | 22.3342 | 1.0673 | 31.685 | |
| Base ModelModel=Stable Diffusion 3.5 Large, Step=28, #NFE=28x2, Resolution=1024x1024, Img-free=false2026.04 | 0.3196 | — | 5.9178 | 22.5994 | 1.0641 | 31.1081 | |
| 1.x-DistillModel=Stable Diffusion 3.5 Large, Step=2, #NFE=1.74, Resolution=1024x1024, Img-free=true2026.04 | 0.3191 | 22.0545 | 5.7976 | 22.7963 | 1.1463 | 32.0059 | |
| Ours-SFDModel=Stable Diffusion 3.5 Large, Step=4, #NFE=4, Resolution=1024x1024, Img-free=true2026.04 | 0.3187 | 17.3588 | 5.9939 | 22.9046 | 1.2011 | 32.902 | |
| TDM*Model=Stable Diffusion 3 Medium, Step=2, #NFE=2, Resolution=1024x1024, Img-free=true2026.04 | 0.3186 | 19.3005 | 5.1441 | 22.4756 | 1.1101 | 31.4725 | |
| Base ModelModel=Stable Diffusion 3 Medium, Step=28, #NFE=28x2, Resolution=1024x1024, Img-free=false2026.04 | 0.3176 | — | 5.6348 | 22.5554 | 1.0429 | 30.7197 | |
| FlashModel=Stable Diffusion 3 Medium, Step=4, #NFE=4, Resolution=1024x1024, Img-free=false2026.04 | 0.3166 | 15.6443 | 5.5485 | 22.3879 | 0.8938 | 29.4835 | |
| TurboModel=Stable Diffusion 3.5 Large, Step=4, #NFE=4, Resolution=1024x1024, Img-free=false2026.04 | 0.3161 | 15.3123 | 6.1308 | 22.7418 | 0.9288 | 30.4127 | |
| Ours-SFDModel=Stable Diffusion 3 Medium, Step=4, #NFE=4, Resolution=1024x1024, Img-free=true2026.04 | 0.3149 | 14.1349 | 5.9197 | 22.8155 | 1.1196 | 32.5337 | |
| TDMModel=Stable Diffusion 3 Medium, Step=4, #NFE=4, Resolution=1024x1024, Img-free=true2026.04 | 0.3128 | 14.6424 | 5.5494 | 22.4681 | 1.0021 | 31.7512 | |
| Hyper-SDModel=Stable Diffusion 3 Medium, Step=4, #NFE=4x2, Resolution=1024x1024, Img-free=false2026.04 | 0.3127 | 15.5475 | 4.9582 | 21.6407 | 0.7543 | 28.7578 | |
| DMD2*Model=Stable Diffusion 3 Medium, Step=4, #NFE=4, Resolution=1024x1024, Img-free=false2026.04 | 0.3122 | 14.7125 | 5.4632 | 22.412 | 0.9981 | 31.0152 | |
| PCMModel=Stable Diffusion 3 Medium, Step=4, #NFE=4, Resolution=1024x1024, Img-free=false2026.04 | 0.3102 | 17.5605 | 5.7743 | 22.069 | 0.6715 | 29.0864 | |
| PCMModel=Stable Diffusion 3 Medium, Step=2, #NFE=2, Resolution=1024x1024, Img-free=false2026.04 | 0.3077 | 41.6561 | 5.1325 | 20.9493 | 0.2011 | 24.8431 |