How do you save and load a trained Keras model?

To save a trained Keras model, you can use the ‘model.save()‘ method, which stores the model architecture, weights, and optimizer state in a single file format called HDF5.

Here’s an example:

from tensorflow import keras

# Create and train a Keras model
model = keras.models.Sequential([
    keras.layers.Dense(10, input_shape=(20,), activation="relu"),
    keras.layers.Dense(1, activation="sigmoid")
])
model.compile(loss="binary_crossentropy", optimizer="sgd")
model.fit(x_train, y_train, epochs=10)

# Save the trained model
model.save("my_model.h5")

To load a saved model, you can use the ‘keras.models.load_model()‘ method, which loads the architecture, weights, and optimizer state from the HDF5 file.

Here’s an example:

# Load a saved Keras model
model = keras.models.load_model("my_model.h5")

# Make predictions with the loaded model
y_pred = model.predict(x_test)

Note that when you load a model, you should use the same version of Keras and TensorFlow that were used to save the model, otherwise loading the model may fail.