joaopauloschuler
joaopauloschuler
@Kryuski , Have you got this error while running one of the examples that come with this API?
@mikerabat , thank you for bringing to my attention that random is not thread safe! In my own experimentation, random feeding gives better results than a predictable feeding. I can...
@JeanYvesJonet, sorry for taking so long to reply. I'll try to reproduce this problem and let you know.
Hello @StepHaze , Applying NN for playing games is certainly possible. There are currently 234 papers at https://paperswithcode.com/task/atari-games about this subject. I do not have any source code example with...
I would have a look at https://paperswithcode.com/task/game-of-chess
The forward pass of categorical cross-entropy is implemented via `TNNetSoftMax`. There is an example at: https://github.com/joaopauloschuler/neural-api/blob/master/examples/SimpleImageClassifier/SimpleImageClassifier.lpr Regarding the backpropagation, I have changed my mind about the best approach to it...
@mikerabat , you are certainly not annoying. Glad to help.
Dear @mikerabat , I'll look into it and FUP.
Many thanks for reporting the bug. The `Enabled` property is controlled by `TNNet.EnableDropouts` and `TNNetDataParallelism.EnableDropouts` so it can be enabled for training and disabled for testing. You are correct at...
Dear @mikerabat , I believe that this bug has been fixed. If you believe otherwise, please feel free reopening it with additional details.