Text-to-Image Generation on PartiPrompts (test)
18.42FIDMuse (Model-B)
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Muse (Model-B)Model type=Early-stopped Muse model, Base-model iterations=20, Super-resolution model iterations=3, Training dataset=WebLI2023.11 | 18.42 | 17.19 | 0.951 | 6.9 | |
| Muse (Model-A)Model type=Full Muse model, Base-model iterations=24, Super-resolution model iterations=8, Training dataset=WebLI2023.11 | 21.4 | 20.16 | 0.721 | 92.5 |