Description
In this course, you will learn about GPT-3, writing prompts in the playground, and converting them into an API using AWS infrastructure. You will also see how to build a full-stack text paraphraser app using no-code tools like BubblePrerequisites: Strong Python programming skills with basic knowledge of AI are assumed. Previous experience with Amazon Web Services (AWS) is helpful. We will begin by understanding GPT-3 and its various parameters that you can modify to get the desired output. We will create a prompt to do sentence paraphrasing using few-shot examples. We will extend it to see how to parallely call the GPT-3 API to paraphrase multiple sentences at once to build general text paraphraser. Later we will see how we can convert the GPT-3 prompt in the playground to code in Google colab. We will further convert the code in Google Colab to API using AWS Lambda serverless infrastructure. We will use AWS SAM (serverless application model) to make it easy to code and deploy our APIs easily. Finally, we will create a UI for our prompts using Bubble no-code tool, integrating the APIs we created earlier. We will also see how to store user runs in database and throttle usage using Bubble
