-
Notifications
You must be signed in to change notification settings - Fork 19.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Dropout in ConvLSTM Cell #20063
Comments
This code snippet was commented in the fix here 2cae421 to fix Conv LSTM correctness |
Hi @sachinprasadhs, can you be a little more clear with this explanation? What does "correctness" mean here, exactly? I've been looking at the base RNN layer here: https://github.com/keras-team/keras/blob/master/keras/src/layers/rnn/rnn.py. Is the dropout now done under the hood? In the Dropout Test here (https://github.com/keras-team/keras/blob/master/keras/src/layers/rnn/dropout_rnn_cell_test.py), those two calls aren't commented out. |
Implemented dropout and recurrent dropout in `ConvLSTMCell` with the same approach as in Keras 2. Fixes keras-team#20063
1 similar comment
Thank you for the report. This was an oversight. It is now fixed. |
Really quick question here.
In the ConvLSTMCell here, the dropout code has been commented out.
recurrent_dropout
also doesn't seem to do anything, except it is used in theDropoutRNNCell
mixin.Is this a bug or am I missing something?
The text was updated successfully, but these errors were encountered: