Sampling via re-noise similar to consistency model?

Thanks for the great work and code release! I have one question: assuming we have only one segment for simplicity, in the paper you mentioned that Euler solver is used to sample from the model. But since it is partially trained with the consistency objective, can we adopt the sampling method in consistency model (adding noise to predicted x_0)? Thanks!