Text-to-Image Generation on BarbieCore
170.25FIDStelLA
Evaluation results
| Method | Links | ||
|---|---|---|---|
| StelLABase Model=SD 1.5, Trainable Parameters (%)=0.0932025.10 | 170.25 | 30.15 | |
| LoRABase Model=SD 2.0, Trainable Parameters (%)=0.0962025.10 | 171.68 | 30.72 | |
| StelLABase Model=SD 2.0, Trainable Parameters (%)=0.0962025.10 | 171.83 | 30.92 | |
| DoRABase Model=SD 1.5, Trainable Parameters (%)=0.1042025.10 | 175.04 | 30.36 | |
| LoRABase Model=SD 1.5, Trainable Parameters (%)=0.0932025.10 | 175.48 | 30.31 | |
| DoRABase Model=SD 2.0, Trainable Parameters (%)=0.1072025.10 | 176.16 | 30.87 | |
| w/o finetuneBase Model=SD 1.5, Trainable Parameters (%)=-2025.10 | 208.11 | 30.84 | |
| w/o finetuneBase Model=SD 2.0, Trainable Parameters (%)=-2025.10 | 210.45 | 31.12 | |
| PiSSABase Model=SD 1.5, Trainable Parameters (%)=0.0932025.10 | 299.49 | 17.32 | |
| PiSSABase Model=SD 2.0, Trainable Parameters (%)=0.0962025.10 | 315.64 | 16.86 |