DL5, W4 Assignment. ValueError: The first argument to `Layer.call` must always be passed

Course Q&A Deep Learning Specialization Sequence Models
1

image
image1224×853 63.2 KB

I am not able to understand, why i am getting the error.
Would anyone help in resolving or throw any light on it?

2

You have not provided the correct arguments to the self.mha() call.
You have to provide data for the query, value, and key. Fortunately, since this is self-attention, you use ‘x’ for all three.
And you do not need the training argument.

1 Like
3

I did the same thing, but i am still getting the same error

4

Please share your full error.

5 

ValueError                                Traceback (most recent call last)
<ipython-input-22-00617004b1af> in <module>
      1 # UNIT TEST
----> 2 EncoderLayer_test(EncoderLayer)

~/work/W4A1/public_tests.py in EncoderLayer_test(target)
     84     encoder_layer1 = target(4, 2, 8)
     85     tf.random.set_seed(10)
---> 86     encoded = encoder_layer1(q, True, np.array([[1, 0, 1]]))
     87 
     88     assert tf.is_tensor(encoded), "Wrong type. Output must be a tensor"

/opt/conda/lib/python3.7/site-packages/tensorflow/python/keras/engine/base_layer.py in __call__(self, *args, **kwargs)
   1010         with autocast_variable.enable_auto_cast_variables(
   1011             self._compute_dtype_object):
-> 1012           outputs = call_fn(inputs, *args, **kwargs)
   1013 
   1014         if self._activity_regularizer:

<ipython-input-21-b5dd6fb08ad4> in call(self, x, training, mask)
     52         # apply dropout layer to ffn output during training (~1 line)
     53         # use `training=training`
---> 54         ffn_output = self.dropout_ffn(training=training)
     55 
     56         # apply layer normalization on sum of the output from multi-head attention (skip connection) and ffn output to get the

/opt/conda/lib/python3.7/site-packages/tensorflow/python/keras/engine/base_layer.py in __call__(self, *args, **kwargs)
    940     #   not to any other argument.
    941     # - setting the SavedModel saving spec.
--> 942     inputs, args, kwargs = self._split_out_first_arg(args, kwargs)
    943     input_list = nest.flatten(inputs)
    944 

/opt/conda/lib/python3.7/site-packages/tensorflow/python/keras/engine/base_layer.py in _split_out_first_arg(self, args, kwargs)
   3046     else:
   3047       raise ValueError(
-> 3048           'The first argument to `Layer.call` must always be passed.')
   3049     return inputs, args, kwargs
   3050 

ValueError: The first argument to `Layer.call` must always be passed.
6

You also need to specify the input to the dropout. For dropout, input is “the output of the self.ffn”

7

thankyou ,it worked

8

Hello Deepak, by any chance have you change the code written in the notebook ??

Because without the code the cell looks like this

START CODE HERE

    # calculate self-attention using mha(~1 line).
    # Dropout is added by Keras automatically if the dropout parameter is non-zero during training
    self_mha_output = # Self attention (batch_size, input_seq_len, fully_connected_dim)
    
    # skip connection
    # apply layer normalization on sum of the input and the attention output to get the  
    # output of the multi-head attention layer (~1 line)
    skip_x_attention =   # (batch_size, input_seq_len, fully_connected_dim)

    # pass the output of the multi-head attention layer through a ffn (~1 line)
    ffn_output =  # (batch_size, input_seq_len, fully_connected_dim)
    
    # apply dropout layer to ffn output during training (~1 line)
    # use `training=training` 
    ffn_output = 
    
    # apply layer normalization on sum of the output from multi-head attention (skip connection) and ffn output to get the
    # output of the encoder layer (~1 line)
    encoder_layer_out =  # (batch_size, input_seq_len, embedding_dim)
    # END CODE HERE

Also your code is incorrect for attn_output which in upgraded lab is mentioned as self_mha_output. (for the write code refer to the ungraded lab: you will find the answer. Also TMosh gave you the hint for the first line of code.

I sincerely suggest you to upgrade the lab if you have not done any changes to the given cell or the notebook.

Hope it helps

Regards
DP