In this project, we will show you step-by-step how to use IBM Code Engine to deploy your AI application on IBM Cloud. IBM Code Engine is a fully managed, serverless platform which provides an abstraction for the underlying infrastructure required to deploy your apps and lets you focus on the source code only (such as the Python code). The following picture shows an example of the text-generation model. On the left-hand side, the user provides a text prompt, which the GPT2 model will take as an input and return an output on the right-hand side. You can also access and play with the application here.

sampletext (2).png 149 KB

You will also use Gradio, a Python framework that allows you to build demos or interactive apps of your Machine Learning models, APIs, or Data Science workflows and share them. Gradio lets you quickly generate a user interface for your application with just a few lines of code, which means even without knowledge of HTML or CSS, you can still create a simple UI by calling its powerful Python APIs.

Learning Objectives:

There is no prerequisite for this project. We will teach you all the Python code and the Code Engine CLI commands. We will make sure you understand everything as you go through the steps of the project. If you don’t have a finished AI model, don’t worry. We will show you how to import a GPT2 model pipeline from HuggingFace. We will also teach you some basics of deploying containerized applications.