LoLo
@stephanenicolas Get! I will try to fix the orders problem :) Thx for help!
@stephanenicolas hah I will mention you once I update the PR. Good luck for your new task :)
@stephanenicolas green now :D plz have a check :)
@stephanenicolas btw, I sent a PR some time ago and it was red as well haha. But I have read the CI test report and the diff is just the...
@stephanenicolas Ya I know the TDD principle, but what I am confused is that I need to add some codes not really elegant if I want to make the CI...
Thx~ I read the source code yesterday and realize that I can use `Dart.inject(Object o, Bundle source)` to implement my need. But there is still something I am confused. If...
Sure. I will write a sample to reproduce this bug. The issue could be close. :) #138
Hi! Thanks for sharing the failure case! I will try to reproduce the result using your code. Do you know how much resource it needs for training?
Thanks for more details. In this case, I guess AdaBound is a little bit sensitive on RCAN model, and a `final_lr` of 0.1 is too large. You may try some...
1e-4 might be too small ... If I understand correctly, the only difference between `mean` and `sum` is a scale of N (the count of samples in a step). If...