[Advanced] Train a model using both tf.keras and tf.estimator.

TensorFlow

You can train models in TensorFlow by using both tf.keras and tf.estimator. Your task is to create and document a Google Colab notebook where you will train a machine learning model using both tf.keras and tf.estimator.

Your notebook should be divided into three parts:

Train a model using the tf.keras API.
Train the same model using the tf.estimator API.
Convert the tf.keras model to an Estimator by using tf.keras.estimator.model_to_estimator().
In your notebook, compare both the APIs.

Task tags

tf
keras
colab
estimators

Students who completed this task

Prem, matthewlee, Panther314, Suryansh5545, g00g1y_5p4, jimmy_page, rpal, Joseph, Kitsunebula, Tom, Rachin, arpand, jedlimlx, george1, malder8, Jake S., generationxcode, msteknoadam, Qwerty71, As1234, MaanavS, anigasan, Ryan, An Onimous Boy, Nikhil Arora, Rick Wierenga, boron, abhaykashyap03, WZHANG, surya_nayar, Sambhav Gupta, JuanC, Ja-sniff/Javismb, Abee, RichieX, Cking100, bkkaggle

Task type

Code
Documentation / Training