0

I am having an issue of where I am obtaining an error to do with the input shapes for the CNN model.

The code for the model

text_input = Input(shape = (max_len,), name = 'text_input')
numerical_input = Input(shape = (train_num_scaled.shape[1],), name = 'numerical_input')

embedding = Embedding(input_dim = vocab_size, input_length = max_len, output_dim = 100, 
                      input_shape = (40,), weights = [glove_embedding_matrix], 
trainable = False)(text_input)

con = Conv1D(filters = 128, kernel_size = 3, activation = 'relu')(embedding)

text_feature = GlobalMaxPooling1D()(con)

numerical_features = Dense(units = 64, activation = 'relu')(numerical_input)

concat = Concatenate()([text_feature, numerical_features])

fully_con = Dense(units = 64, activation = 'relu')(concat)

output = Dense(units = 1, activation = 'sigmoid')(fully_con)

cnn_model = Model(inputs = [text_input, numerical_input], outputs = output)

cnn_model.summary() 

cnn_model.compile(optimizer = 'adam', loss = 'binary_crossentropy', metrics = 
['accuracy'])

The output for the model indicates some sort of issue with the shape of the inputs - suggesting that I am using a Sequential model which isn't the case. The user warning is:

C:\Users---\Python\Lib\site-packages\keras\src\layers\core\embedding.py:90: UserWarning: Argument input_length is deprecated. Just remove it. warnings.warn( C:\Users---\Python\Lib\site-packages\keras\src\layers\core\embedding.py:93: UserWarning: Do not pass an input_shape/input_dim argument to a layer. When using Sequential models, prefer using an Input(shape) object as the first layer in the model instead. super().init(**kwargs)

Output for model summary

I then fit the model where again a user error comes up also suggesting an issue with the structure/shape of the inputs but it still runs all the epochs.

The code:

history = cnn_model.fit( x = [train_pad_tweets, train_num_scaled], 
                       y = train_target, 
                       validation_data = ([val_pad_tweets, val_num_scaled], 
val_target),
                       epochs = 10, batch_size = 32)
#while this works technically - seems to be a warning about the expected and recieved 
structure

The error:

C:\Users----\Python\Lib\site-packages\keras\src\models\functional.py:225: UserWarning: The structure of inputs doesn't match the expected structure: ['text_input', 'numerical_input']. Received: the structure of inputs=('', '') warnings.warn(

I cannot figure out why there is a potential issue with the shapes of inputs or is this is causing an actual potential issue.

I have attempted putting the history fit in a dictory style format however that is leading to more error again around the shape of the data

Improve this question
1
  • The first warning simply says that you should remove the parameter input_length. And it does not say that you are using a sequential model, it just tells you what you should do if you were using a sequential model. Since you are not using one you can ignore it. The second warning looks as if you are passing empty strings to the model, so you should check your inputs. Commented Dec 11, 2024 at 14:20
Related questions
1811
Random string generation with upper case letters and digits
1409
How can I read a text file into a string variable and strip newlines?
0
Concatenate BERT and CNN
Load 1 more related questions Show fewer related questions

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.