Closed ShenZheng2000 closed 1 month ago
How much time and how many GPUs are needed to train the model?
For video generation, training time is similar to MagicDrive-t. For reconstruction optimization, as discussed in Appendix A of our paper, it takes about 30min on a single V100.
How much time and how many GPUs are needed to train the model?