Jimin Jun
https://github.com/tensorflow/tensor2tensor/blob/5623deb79cfcd28f8f8c5463b58b5bd76a81fd0d/tensor2tensor/layers/area_attention.py#L415 In typical dot product attention, logit which is the input matrix of softmax supposed to be divided by square rooted temperature like the equation below.  However, in this...
In dc_crn.py inputs and labels are random tensor. How could I process sample wav files? Is it right to change line 305 at dc_crn.py into some kind of wav file...
First of all, your work is so helpful. So I wanna thank first. Recently, I re-tried to run your code from beginning. But there is too many errors due to...
Recently, I'm trying to study your paper and codes, but I couldn't run shell file with following readme.txt. So I try to run with python step by step with reconstructing...
https://github.com/habla-liaa/ser-with-w2v2/blob/c9be8a9bc8c6c6969838e3e77c2fc8af10e27136/configs/models/dienen_classifier.yaml#L2 Even if I changed gpu number 0 into 2 or 3, still GPU 0 is allocated.  How could I change it?
https://github.com/habla-liaa/ser-with-w2v2/blob/c9be8a9bc8c6c6969838e3e77c2fc8af10e27136/tasks/normalize.py#L22 #ValueError: operands could not be broadcast together with shapes (97,768) (68,768) shape doesn't match while updating normalization. I guess padding might be needed. How could I solve this problem?
https://github.com/habla-liaa/ser-with-w2v2/blob/c9be8a9bc8c6c6969838e3e77c2fc8af10e27136/tasks/feature_extractors.py#L161 **activation** is dictionary. so it indexing this way does not work (activations[-1]) so I changed some into line below features = list(activations.values())[-1].cpu().detach().numpy() you might change indexing making into list...
Hi, I am impressed with your work. What model architecture did you use?
does MovieCLIP dataset contain audio in provided video clip?
 다 {}_Windows 밖에 없는데 리눅스는 어떻게 만드는 건가요? 리눅스 환경에서 어떻게 하나요