CQTdiff

CQTdiff copied to clipboard

Hi! Thank you for your great project. I have learned a lot from both your paper and your codes. I found there is no detailed description of the dataset preparation....

YuXU-Jouuuuuu

reconstruction guidance mismatch with the paper

3

``` norm=torch.linalg.norm(y-den_rec,dim=dim,` ord=2) rec_grads=torch.autograd.grad(outputs=norm, inputs=x) rec_grads=rec_grads[0] normguide=torch.linalg.norm(rec_grads)/x.shape[-1]**0.5 #normalize scaling s=self.xi/(normguide*t_i+1e-6) #optionally apply a treshold to the gradients if self.treshold_on_grads>0: #pply tresholding to the gradients. It is a dirty trick but...

XZWY

Hi, I have been trying to test out this method, specifically in the Google Colab Notebook ([https://colab.research.google.com/github/eloimoliner/CQTdiff/blob/main/notebooks/demo.ipynb](https://colab.research.google.com/github/eloimoliner/CQTdiff/blob/main/notebooks/demo.ipynb)). However, it has errors. Everything was (seemingly) ok up to 'Unconditional Synthesis', which...

ArloMichael