Conditional Generation on Maestro (test)
0.667FADLATENTFT-MLP
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| LATENTFT-MLP2026.04 | 0.667 | 0.809 | 0.967 | 0.553 | 0.085 | |
| LATENTFT-UNet2026.04 | 0.865 | 0.83 | 0.968 | 0.59 | 0.085 | |
| Guidance Gradients2026.04 | 8.588 | 0.53 | 0.795 | 1.483 | 0.116 | |
| ILVR2026.04 | 9.923 | 0.58 | 0.817 | 0.976 | 0.118 | |
| DAC2026.04 | 11.745 | 0.729 | 0.835 | 4.088 | 0.243 | |
| Vampnet2026.04 | 11.914 | — | — | — | — | |
| Spectrogram2026.04 | 14.208 | 0.413 | 0.853 | 1.981 | 0.152 |