deep-learning-from-scratch-3 icon indicating copy to clipboard operation
deep-learning-from-scratch-3 copied to clipboard

Published 20 hours ago • verified publisher icon oreilly-japan

Implementation ideas of backpropagation of concat operator

Open owenliang opened this issue 1 year ago • 3 comments

This book is the best deep learning book I have ever seen. With your help, I also wrote a framework with personal characteristics.

At the end of the study of the whole book, I have one more technical detail that I would like to ask you about, which is the frequently used "concat" operator. Its backpropagation involves the reverse splitting of gradient, which I really can't figure out. , can you provide some ideas or add this method to dezero, thank you.

owenliang avatar Nov 19 '23 12:11 owenliang

Thank you for reading this book. I would love to think about that question ("concat"), but I'm busy writing my next book right now.

I would like to take some time when I finish writing my next book. Or maybe someone else can help you (I announced it on Twitter).

koki0702 avatar Nov 20 '23 00:11 koki0702

I implemented it like this. https://github.com/laksjdjf/dezero-diffusion/blob/a223c7e2bb06e149ff0a8b0714fcc88fb38b08b7/modules/unet.py#L10-L40

laksjdjf avatar Nov 20 '23 01:11 laksjdjf

I implemented it like this. https://github.com/laksjdjf/dezero-diffusion/blob/a223c7e2bb06e149ff0a8b0714fcc88fb38b08b7/modules/unet.py#L10-L40

3Q，slice helps a lot! Wish this issue also helps others.

owenliang avatar Nov 25 '23 02:11 owenliang