Text-to-Audio Generation on InternalSFX (test)
246.9FDWoosh-Flow-Private
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Woosh-Flow-PrivateParams=337 M, NFEs=≈2×702026.04 | 246.9 | 2.348 | 0.301 | |
| Woosh-DFlow-PrivateParams=337 M, NFEs=42026.04 | 270.9 | 2.651 | 0.2569 | |
| Woosh-Flow-PublicParams=337 M, NFEs=≈2×702026.04 | 282.7 | 2.248 | 0.3752 | |
| Woosh-DFlow-PublicParams=337 M, NFEs=42026.04 | 322.5 | 2.404 | 0.3093 | |
| SAOParams=1057 M, NFEs=2×1002026.04 | 341 | 2.69 | 0.2544 | |
| TangoFluxParams=515 M, NFEs=2×1002026.04 | 378.1 | 2.729 | 0.2249 |