learning_by_grad_by_grad_repro
learning_by_grad_by_grad_repro copied to clipboard
Any idea to scale up to complex networks?
Thank you for the great effort!
- Is the reason of not using nn.Parameters but using Variables can be explained by this post?
- I think relying on Variables is hard to scale up to complex networks (e.g. ResNet). Do you have any idea or suggestions?
Thank you!