Could you please share the results and performances of your implementation on some popular datasets? For example, the total training time, the hardware requirement, and the FID in cifar10, imagenet64, and the LSUN dataset. Does the performance meet the performance reported in the original paper?
The repo only contains code for the consistency models algorithm. To perform a performance comparison, you would have to implement the neural network that the authors used.
Nice implementation!
Could you please share the results and performances of your implementation on some popular datasets? For example, the total training time, the hardware requirement, and the FID in cifar10, imagenet64, and the LSUN dataset. Does the performance meet the performance reported in the original paper?