LightX2V
LightX2V copied to clipboard
How can a 4-step distillation model be used to train other downstream tasks?
Based on the original code using the official wan-i2v code, what modifications are needed?