Skip to main content

Saving the Model

Summary​

  • Saving a model

    • SavedModel format

      model.save('<path>')
      • HDF5 format
      model.save('<path>.h5')

  • Loading a model

tf.keras.models.load_model('<path>')

Content​

Saving and restoring the models​

model.save("save1")
restored_model = tf.keras.models.load_model("save1")

y_predict = restored_model.predict(X_test)

import matplotlib.pyplot as plt

plt.scatter(X_test, y_test, c="g", label="Expected")
plt.scatter(X_test, y_predict, c="red", label="Actual")