Loading a TF DF model?

Research & Models
#1

Hi

I trained a tf df model which I now want to productionalize on cloud. I saved the model with the model.save(“path”) option. However I can’t find any documentation on how to load the model again?

When I try:

model = tf.saved_model.load(“path”)

It loads but it gives me the following errors when I try to summarize the model:

AttributeError: ‘_UserObject’ object has no attribute ‘summary’

Any help would be appreciated

#3

Hi @JP_Swanepoel ,

Could you please try to load using below code and summarize the model.

model=tf.keras.models.load_model(“path”)

model.summary()

As per my understanding tf.saved_model.load() loads the entire saved model graph, but it does not provide access to the summary() or If you want to see the summary of the model after loading it using tf.saved_model.load() , you can try to use tf.keras.utils.plot_model() function to visualize the model graph.

Hope this helps you!

Thanks.