Open Kacper-W-Kozdon opened 7 months ago
It probably works, it's in the training right now- it took quite a bit to figure out the syntax differences between torch's backward()
and ivy's execute_with_grad()
. The descriptions will be added soon. I can't tell right now if it works correctly.
The demo notebook is set to run on kaggle. In case you want to run it outside of kaggle, there are some extra steps in the notebook to help you do it.