Jad Mounayer
Jad Mounayer
Thank you for the comment! I agree, using jnp.full is a better way. I fixed this and followed the instructions in CONTRIBUTING.md.
Hello @vroulet, thank you for your answer! You're right, I should have enclosed a MWE. Below, you can find a typical code written in equinox, where the model is simply...
Hello @vroulet , @randomekek , thank you for your replies. Apologies for the delay. I don't see what you mean @vroulet in the example you proposed. What we want is...
No problem! Thanks for replying.
No problem @vroulet . Thanks for the cooperation!
@TugdualKerjan thanks! I won't be able to work on it anytime soon, so feel free to take the lead :)
Let me share with you what I know. The most poopular Unpooling technique is Max Unpooling which is basically the transpose of a Max pooling layer. When a max pooling...
Hello, It's been a while since this issue was discussed. But I wanted to share the following paper: [Robust differentiable SVD](https://arxiv.org/abs/2104.03821). Which proposes a way of using Taylor's extension to...