Problem of compilation of model

Keras
,
1

Hi

I got a problem when i want to compile my nlp model using keras on google colab.

here is my code :

model = Sequential()
model.add(Embedding(input_dim=len(tokenizer.word_index) + 1, output_dim=128, input_length=max_length))
model.add(LSTM(128, return_sequences=True))
model.add(LSTM(64)) # Dernière LSTM sans return_sequences
model.add(Dropout(0.2))
model.add(Dense(82, activation=‘relu’))
model.add(Dense(7, activation=‘softmax’))
model.compile(optimizer=optimizer, loss=‘categorical_crossentropy’, metrics=[‘accuracy’])
model.summary()

but the output is :

image
image1133×501 36.1 KB

could you please assit me and tell me what is the error ?

Thanks in advance

3

Hi @Nicolas_Orefice, This happens when input shape is not defined. In the first layer

you have not defined the input_shape so model.summary does not provide the output shape.

If you provide the input shape

model = Sequential()
model.add(Embedding(input_dim=786, output_dim=128, input_length=64,input_shape=(64,)))
model.add(LSTM(128, return_sequences=True))
model.add(LSTM(64)) # Dernière LSTM sans return_sequences
model.add(Dropout(0.2))
model.add(Dense(82, activation='relu'))
model.add(Dense(7, activation='softmax'))
model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])

then the model summary will generate the output shape based on the input shape

model.summary()
Model: "sequential_2"
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓
┃ Layer (type)                         ┃ Output Shape                ┃         Param # ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩
│ embedding_2 (Embedding)              │ (None, 64, 128)             │         100,608 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ lstm_2 (LSTM)                        │ (None, 64, 128)             │         131,584 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ lstm_3 (LSTM)                        │ (None, 64)                  │          49,408 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ dropout_1 (Dropout)                  │ (None, 64)                  │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ dense_2 (Dense)                      │ (None, 82)                  │           5,330 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ dense_3 (Dense)                      │ (None, 7)                   │             581 │
└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘
 Total params: 287,511 (1.10 MB)
 Trainable params: 287,511 (1.10 MB)
 Non-trainable params: 0 (0.00 B)

Thank You.