Hello, thank you for your great work! I want to continue finetuning your pretrained model to achieve multiview consistency (similar to Difix3D+). Since you have already distilled it to a one-step model, I am wondering do I still need the vsd regularizer? If needed, can you share the regularizer checkpoint? Looking forward to your reply!
Hello, thank you for your great work! I want to continue finetuning your pretrained model to achieve multiview consistency (similar to Difix3D+). Since you have already distilled it to a one-step model, I am wondering do I still need the vsd regularizer? If needed, can you share the regularizer checkpoint? Looking forward to your reply!