Miguel Varela Ramos
Miguel Varela Ramos
Feel free to make a PR to the repository
right now the behaviour should be very similar to dropout, but it has some subtle changes. Particularly, at the moment the mask is broadcasted to all the channels, which dropout...
looks fine to me. you can make a PR with that code and add a few unit tests (check the current ones). I'll happily review it for you.
The gamma issue is a minor thing but I can have a look at it. The channels share the same mask in the paper.
Sure, that is easily fixable Expect it soon Edit: you can also do a PR if you want
> Hi, > Any updates on this? > Best I haven't had much free time to deal with this, but I will review and accept merge requests
> To solve this issue, you could have a look at [this folk(only for DropBlock2D)](https://github.com/Eliza-and-black/dropblock) I would encourage you to do a pull request
It could be a pytorch version incompatibility. Which version are you using?
Feel free to add a PR for that, should be simple
I don't have time to update the PyPI repo at this moment. You can install the library with python 3.8 if you use: `pip install git+https://github.com/miguelvr/dropblock.git#egg=dropblock`