Deploy a model with custom prediction routine on Google AI Platform.
Use a model deployed on Google AI Platform for inference.
In this 2-hour long project-based course, you will learn how to deploy, and use a model on Google’s AI Platform. Normally, any model trained with the TensorFlow framework is quite easy to deploy, and you can simply upload a Saved Model on Google Storage, and create an AI Platform model with it. But, in practice, we may not always use TensorFlow. Fortunately, the AI Platform allows for custom prediction routines as well and that’s what we are going to focus on. Instead of converting a Keras model to a TensorFlow Saved Model, we will use the h5 ﬁle as is. Additionally, since we will be working with image data, we will use this opportunity to look at encoding and decoding of byte data into string for data transmission and then encoding of the received data in our custom prediction routine on the AI Platform before using it with our model.
This course runs on Coursera’s hands-on project platform called Rhyme. On Rhyme, you do projects in a hands-on manner in your browser. You will get instant access to pre-configured cloud desktops containing all of the software and data you need for the project. Everything is already set up directly in your Internet browser so you can just focus on learning. For this project, you’ll get instant access to a cloud desktop with (e.g. Python, Jupyter, and Tensorflow) pre-installed.
In order to be successful in this project, you should be familiar with Python programming, Google Cloud Platform.
– You will be able to access the cloud desktop 5 times. However, you will be able to access instructions videos as many times as you want.
– This course works best for learners who are based in the North America region. We’re currently working on providing the same experience in other regions.
Custom Prediction Routine on Google AI Platform
In this 2-hour long project-based course, you will learn how to deploy and use a model on Google’s AI Platform. AI Platform allows for custom prediction routines in addition to TensorFlow, XGBoost and Sci-kit models, and that’s what we are going to focus on. Instead of converting a Keras model to a TensorFlow Saved Model, we will use the h5 ﬁle as is.
Notebook Instance and Model Artifact
Testing the Model
Custom Prediction Class
Deploying the Model