I tried to create model in tensorflow 2.x using functional API, but got LSTM layers incompatible error

Question

Error reads:

Input 0 of layer lstm_28 is incompatible with the layer: expected ndim=3, found ndim=4. Full shape received: [None, None, 15, 12]

In LSTM layer, input tf.nn.embedding_lookup(embedding, neighbor) have shape =(15,12) and one None is for batch size, how it comes the size of [None, None, 15,12]? How to deal with this error? Below is the dummy model that I created.

    def create_model(embedding, embedding_dim, samp_size):
        
        
        node = Input(shape=(None,), dtype=tf.int64)
        neighbor = Input(shape=(None, samp_size), dtype=tf.int64)
        label = Input(shape=(None,), dtype=tf.int64)
        
        cell = LSTMCell(embedding_dim,)
        _,h,c = LSTM(embedding_size, return_sequences=True, return_state=True)(tf.nn.embedding_lookup(embedding, neighbor))
        predict_info = tf.squeeze(Dense(1, activation='relu'))(h)
        
        return h
    
    
    
    node_size = 1000
    embedding_dim = 12
    sampling_size = 15
    embedding = tf.random.uniform([node_size, embedding_dim])
    
    model = create_model (embedding, embedding_dim, sampling_size)

Answer 1

when using Keras functional API, do not include the None for batch dimension. For example, if your input is of dimension (batch_size, image_w, image_h, image_channels) do it like this:

inp = tf.keras.Input(shape=(IMG_W, IMG_H, IMG_CH))