Gaurav Mittal

In the paper, it says that the 80-dimensional Mel-scale filter bank features are used or 200 dimensional log-magnitude spectrum is used but only for audio reconstruction. What does this mean?

I am unable to locate the file being pointed to "data/spec_scp/train/dataset.cfg". Can you please let me know where it is? Thanks

Hi I was wondering if it could be possible for you to make available any pretrained model for the given code. I am interested in using the feature space of...

This merge will allow visualizing for gray scale (single channel) images based model such as MNIST.

How to download the files from golf.txt?

6

Can you please give an example of full URL for downloading files from the golf.txt mentioned on your website? http://data.csail.mit.edu/videogan/golf.txt

Computation Requirement to train CogVideo

1

Hi, First of all, great work in developing CogVideo. Could you please provide information on how many GPUs and how much duration it took to train the model? Thanks Gaurav

support for non-human faces

1

Hi, First of all, thanks for sharing the implementation of your amazing work. I was wondering if it supports non-human faces? Thanks

Please create a separate branch to support tensorflow-1.0 as previous version are used by many.

Is the KL divergence computed correctly?

1

Hi @MissT157 and I are experimenting with this code and we are finding the implementation of KL divergence a bit awry. Can you please revisit and confirm whether it's been...

Rename punctuate to punctuate_and_cut to reflect the correct function name from Sentencify class.