Get started with Deep Learning using TensorFlow with our new course

Our course, Applied Deep Learning with TensorFlow and Google Cloud AI, is now publicly available! If you would like to know more about end to end machine learning using TensorFlow and Cloud ML Engine, feel free to take the course.

Take a peak on Github —

It’s our ( Haohan Wang and Christian 郑梵力 Ramsey) first course and definitely not our last! We learned so much designing this course and will continue to contribute to the deep learning community. We hope it’s somewhat useful and would love feedback on the course.

What You Will Learn

  • Gain hands-on experience designing, training, and deploying your Deep Learning models with TensorFlow and Keras to handle large volumes of data and complex neural network architectures
  • Get a better understanding of how parallelism and distribution work in TensorFlow and Keras
  • Design and experiment with complex neural network architectures using low-level TensorFlow while also using TensorFlow’s high level APIs and Keras
  • Scale out training and prediction using different distributed techniques such as data parallelism using GPUs on our local machine and in the cloud using Google Cloud ML Engine
  • Develop, train, and deploy models using Google Cloud MLE to production.
  • Deploy your model as a production level API


PREPARATION — Installation and Setup

  • Nvidia Setup
  • Anaconda Setup
  • TensorFlow GPU and Google Cloud
  • Requirements

SECTION I — Deep Learning with Keras

  • 1.1 Keras Introduction
  • 1.2 Review of backends Theano, TensorFlow, and Mxnet
  • 1.3 Design and compile a model
  • 1.4 Keras Model Training, Evaluation and Prediction
  • 1.5 Training with augmentation
  • 1.6 Training Image data on the disk with Transfer Learning and Data augmentation

SECTION II — Scaling Deep Learning using Keras and TensorFlow

  • 2.1 TensorFlow Introduction
  • 2.2 TensorBoard Introduction
  • 2.3 Types of Parallelism in Deep Learning — Synchronous vs Asynchronous
  • 2.4 Distributed Deep Learning with TensorFlow
  • 2.5 Configuring Keras to use TensorFlow for distributed problems

SECTION III — Distributed Deep Learning with Google Cloud MLE

  • 3.1 Representing data in TensorFlow
  • 3.2 Diving into Estimators
  • 3.3 Creating your Data Input Pipeline
  • 3.4 Creating your Estimator
  • 3.5 Packaging your model/trajectory
  • 3.6 Training in the Cloud
  • 3.7 Automated Hyperparameter Tuning
  • 3.9 Deploying your Model to the Cloud for Prediction
  • 3.10 Creating your Machine Learning API



Applied Deep Learning with TensorFlow and Google Cloud AI [Video]


Haohan’s Life Blog —

Christian’s Life Blog —

Authors: Haohan Wang & Christian Fanli Ramsey > dyad x machina

Source: Deep Learning on Medium