Error with shape in LSTM when i try to implement an RNN - Task1a

[AthanJohn]

Hi,
I have a problem with my implementation. I try to implement a RNN and the parameters which i use are shown below.

So, i get this error: Input 0 is incompatible with layer lstm_1: expected ndim=3, found ndim=4.

As you can see at the image below, I reshaped my X_train and X_validation array, but this error still exists.

Can you help me?

PS I use DCASE 2018 implementation, sorry about that. But I think it doesn't matter about this error.

Thanks in advance!

-Athanasiou Giannos