Text-guided Video Editing on WebVid-10M (val)
94.9Frame Consistency (CLIP Score)SimDA
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| SimDATraining Runtime [min]=2.5, Inference Runtime [min]=0.4, Training Memory [Gib]=28.6, Inference Memory [Gib]=8.8, Tuned Params [Mb]=24.92023.08 | 94.9 | 68.8 | 31.9 | 60.5 | |
| Tune-A-VideoTraining Runtime [min]=9.3, Inference Runtime [min]=0.8, Training Memory [Gib]=31.3, Inference Memory [Gib]=11.4, Tuned Params [Mb]=74.42023.08 | 94.1 | 31.2 | 31.8 | 39.5 |