使用 tf.saved_model 保存和加载 model 时会发生什么变化

Question

I was following the official TensorFlow quick start tutorial and I wanted to save and load the model, but the following code raises on the final line

ValueError: Could not find matching function to call loaded from the SavedModel. Got:
  Positional arguments (3 total):
    * Tensor("inputs:0", shape=(1, 28, 28), dtype=float64)
    * False
    * None
  Keyword arguments: {}

with declaring a keras sequential model and saving like so:

print(model(x_train[:1]))
tf.saved_model.save(model, export_path)
loaded_model = tf.saved_model.load(export_path)
print(loaded_model(x_train[:1]))

Answer 1

I was able to work around the problem by using model.save(export_path) and tf.keras.load_model(export_path) respectively. The relevant documentation can be found here .