Practical Deep Learning on the Cloud, published by Packt.
This is the code repository for Practical Deep Learning on the Cloud [Video], published by Packt. It contains all the supporting project files necessary to work through the video course from start to finish.
Deep learning and machine learning applications are becoming the backbone of many businesses in both technological and traditional companies. Once organizations have achieved their first success in using ML/AI algorithms, the main issue they often face is how to automate and scale up their ML/AI workflows. This course will help you to design, develop, and train deep learning applications faster on the cloud without spending undue time and money. This course will heavily utilize contemporary public cloud services such as AWS Lambda, Step functions, Batch and Fargate. Serverless infrastructures can process thousands of requests in parallel at scale. You will learn how to solve problems that ML and data engineers encounter when training many models in a cost-effective way and building data pipelines to enable high scalability. We walk through some techniques that involve using pre-trained convolutional neural network models to solve computer vision tasks. You'll make a deep learning training pipeline; address issues such as multiple frameworks, parallel training, and cost optimization; and save time by importing a pre-trained convolutional neural network model and using it for your project. By the end of the course, you'll be able to build scalable and maintainable production-ready deep learning applications directly on the cloud.
- Training, exporting, and deploying deep learning models on the cloud (TensorFlow)
- Using pre-trained models for your computer vision task
- Working with cluster infrastructures on AWS (AWS Batch and Fargate)
- Creating deep learning pipeline for training models using AWS Batch
- Creating deep learning pipelines to deploy a model into production with AWS Lambda and AWS Step functions
- Creating a data pipeline using AWS Fargate
This course is ideal for data scientists and machine learning engineers who are familiar with TensorFlow and now want to learn how to organize deep learning production systems on the cloud. The course will cover key services including AWS Sagemaker, AWS Fargate, AWS Batch, AWS Step Functions, AWS Lambda. Some experience with these cloud platforms would be helpful, but isn't necessary.
This course has the following software requirements:
For an optimal experience with hands-on labs and other practical activities, we recommend the following configuration:
● Processor: Quad-core 2GHz+ CPU
● Memory: 16GB
● Storage: 100GB
● Operating system: Windows 10, Linux, Mac
● Browser: Firefox, Chrome
● AWS account
● Python installation with pip (Python 3.x recommended)
● TensorFlow installation
● Node.js installation with npm
● Docker