Pranay Manocha

Hi. Thanks for replying. I see that but the problem we are facing is that of converting our dataset (which is the original NYT into your pickled data). There seems...

Hi, Thanks for the comments. The repo indeed doesn't support tf2.0. I currently don't have the bandwidth to convert the repo to being, tf2.0 compatible but I guess tf2.0 has...

Hi, Could you try using CDPAM with batch_size = 1 (i.e just evaluating one pair of clips)? I am guessing that you are able to load the model and copy...

Can you try loading the model on cpu to see if that works? Just making sure that the pip module is setup properly. I didn't benchmark my memory requirements for...

The issue here is the length... The maximum I have tested is a few seconds (as most problems in SE and vocoding deal with small length ~2-3seconds long). Can you...

Hi Adrien, Thanks for the comments. I am not too familiar with pytorch, and so your comment above makes complete sense and that is what I was trying to refer...

Hi Adrien, I was reviewing the codebase and have one more observation - models.py line 71. `diff.shape[0]/diff.shape[1]` . Shouldn't it be `diff.shape[1]/diff.shape[2]` ? The idea is to take the mean...