Video Generation on VBench Wan2.1 1.3B extrapolation, 141x448x832 1.0 (test)
64.27AQFull Attention
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Full AttentionModel=Wan2.1 1.3B, Resolution=141x448x832, Kernel=FullAttn, Training Setting=Fine-tuned2026.01 | 64.27 | 92.73 | 63.89 | 61.28 | 92.63 | — | 640.5 | 214.2 | |
| VSAModel=Wan2.1 1.3B, Resolution=141x448x832, Kernel=VSA, Training Setting=Fine-tuned2026.01 | 63.04 | 92.21 | 66.67 | 59.8 | 89.71 | 90 | 184.2 | 112.6 | |
| VMonarchModel=Wan2.1 1.3B, Resolution=141x448x832, Kernel=VMonarch, Training Setting=Fine-tuned2026.01 | 60.45 | 92.25 | 69.44 | 59.92 | 88.86 | 94.4 | 170.2 | 102.7 |