Arda Senocak

@MaigoAkisame Thank you so much for quick reply and the information. I was really curious about the time :) Unfortunately I need frames for my implementation so have to wait...

Also note that some paths that are listed in train_videos.txt are not actually exist in folders after unzipping. For example; videos2/1/0/9/0/5/9/1/6/10610905916.mp4

@eborboihuc surprisingly you have very similar values. In my case, values are also differs a lot even thou I used the same file with same settings...But seems like your workaround...

@eborboihuc When i test both library with the voice.mp3 (example sound file in torch-audio), I get very similar values. And the dimension difference between two libraries is 576 for sr=22050....

@eborboihuc yes I can confirm this on my side for voice.mp3 file. However when I try different sound files (.mp3 extension), such as 02 - "Canon" (in D-Major), Pachebel from...

@gyglim I also can't believe that 3 nets can make 10~12% difference... Actually when I run the original caffe implementation on my own machine (but note that batch size is...