Text-to-Video Generation on PenguinVideoBenchmark
1.81SpeedupAdaCluster
Evaluation results
| Method | Links | ||
|---|---|---|---|
| AdaClusterModel=Wan2.1-14B, Hardware=NVIDIA H100 GPUs, Resolution=832 × 480, Inference Steps=502026.04 | 1.81 | 404.94 | |
| AdaClusterModel=HunyuanVideo, Hardware=NVIDIA H100 GPUs, Resolution=1280 × 720, Inference Steps=302026.04 | 1.67 | 322.21 | |
| SVG2Model=Wan2.1-14B, Hardware=NVIDIA H100 GPUs, Resolution=832 × 480, Inference Steps=502026.04 | 1.61 | 427.43 | |
| SVG2Model=HunyuanVideo, Hardware=NVIDIA H100 GPUs, Resolution=1280 × 720, Inference Steps=302026.04 | 1.58 | 335.36 |