I am following a YouTube tutorial on transfer learning with TensorFlow and TensorFlow Hub. I am trying to use the MobileNet V2 model as a feature extractor in a keras.Sequential model. However, I encounter a ValueError stating that only instances of keras.Layer can be added to a Sequential model, despite using hub.KerasLayer. The relevant part of my code is:

import tensorflow_hub as hub
from tensorflow import keras

feature_extractor_model = "https://tfhub.dev/google/tf2-preview/mobilenet_v2/feature_vector/4"
model_without_top_layer = hub.KerasLayer(
    feature_extractor_model,
    input_shape=(224, 224, 3),
    trainable=False
)

num_of_flowers = 5

model = keras.Sequential([
    model_without_top_layer,
    keras.layers.Dense(num_of_flowers)
])

model.summary()

I noticed the code runs when i comment model_without_top_layer out. I’m not sure what’s going on