Emojify_V2 Expected values does not match input value

Course Q&A Deep Learning Specialization Sequence Models
1

I am having a hard time figuring out what I got wrong on this workbook. I get the following error. I recognize that I somehow need to revise the LSTM layer, but I am unsure how. I would appreciate guidance.

Layer (type)                    Output Shape         Param #     Connected to                     
==================================================================================================
input_9 (InputLayer)            [(None, 4)]          0                                            
__________________________________________________________________________________________________
embedding_10 (Embedding)        (None, 4, 2)         30          input_9[0][0]                    
__________________________________________________________________________________________________
lstm_10 (LSTM)                  (None, 4, 128)       67072       embedding_10[0][0]               
__________________________________________________________________________________________________
dropout_10 (Dropout)            (None, 4, 128)       0           lstm_10[0][0]                    
__________________________________________________________________________________________________
lstm_11 (LSTM)                  [(None, 128), (None, 131584      dropout_10[0][0]                 
__________________________________________________________________________________________________
dropout_11 (Dropout)            (3, None, 128)       0           lstm_11[0][0]                    
                                                                 lstm_11[0][1]                    
                                                                 lstm_11[0][2]                    
__________________________________________________________________________________________________
dense_5 (Dense)                 (3, None, 5)         645         dropout_11[0][0]                 
__________________________________________________________________________________________________
activation_5 (Activation)       (3, None, 5)         0           dense_5[0][0]                    
==================================================================================================
Total params: 199,331
Trainable params: 199,301
Non-trainable params: 30
__________________________________________________________________________________________________
None
Test failed 
 Expected value 

 ['LSTM', (None, 128), 131584, (None, 4, 128), 'tanh', False] 

 does not match the input value: 

 ['LSTM', [(None, 128), (None, 128), (None, 128)], 131584, (None, 4, 128), 'tanh', False]
2

Here are some hints:

  1. Output shape of the 2nd lstm layer is incorrect. Since the 2nd LSTM layer only returns the output for the last sequence, the output shape should be (None, #units) where None corresponds to batch dimension and #units = 128.
  2. Since dropout only zeros parts of input, the output shape of dropout should match the output shape of the 2nd lstm layer.
  3. Output shape of the Dense layer should be (None, # units in dense layer).
  4. Activation layer should not change the output shape.