Full Stack Deep Learning Bootcamp

Hands-on program for developers familiar with the basics of deep learning

Why

There are great online courses on how to train deep learning models. Some start with theory, some start with code. But training the model is just one part of shipping a complete deep learning project. Our course is aimed at people who already know the basics, and want experience with the full stack.

Content

Attendees will learn and get first-hand experience with best practices of all components of a deep learning project:

Formulating the problem and estimating project cost
Finding, cleaning, labeling, and augmenting data
Picking the right framework and compute infrastructure
Finding the right model architecture by running experiments in parallel
Troubleshooting training and ensuring reproducibility
Deploying the model at scale
Keeping up with fast-moving research areas such as meta-learning

Project

Attendees will spend 30% of the time on developing and deploying a state-of-the-art computer vision and natural language processing system.

Use the most important building blocks of deep learning: convnets, sequence models, and attention
Receive feedback and support along the way
Learn best practices for running experiments
Deploy the final model for widespread use

Certification

Those on the job market can choose to take a certification exam, aimed to ensure that you are ready for deep learning engineer technical interviews. The exam covers pre-requisites as well as course content, and can be taken on your own time.

Instructors

Pieter Abbeel

Professor at UC Berkeley, President of Covariant.AI

Sergey Karayev

Director of AI at Gradescope

Josh Tobin

Research Scientist at OpenAI

Guest Lecturers

Andrej Karpathy

Director of AI at Tesla
Previously Research Scientist at OpenAI

Jai Ranganathan

Director of Product at Uber
Previously Senior Director of Product at Cloudera, Senior Product Manager at VMWare

Yangqing Jia

Director of AI Infrastructure at Facebook
Developed the Caffe framework at UC Berkeley, helped develop Tensorflow at Google

Lukas Biewald

Founder of Weights & Biases
Previously founded Figure Eight (Crowdflower)

Schedule

Any time within 4 weeks of the bootcamp, attendees can choose to take the certification exam, aimed at showing requisite knowledge for the position of deep learning engineer. A study guide will be part of the program materials.

Pre-requisites

At least one year experience programming in Python.
At least one deep learning course taken online or in higher-ed.
Experience with code versioning, Unix environment, and software engineering.

Logistics

The bootcamp will be held Friday-Sunday August 3-5, 2018 on UC Berkeley campus. The three-day bootcamp cost is $2450, with a discount for students.

At most 150 people will be admitted to the program. We will carefully read every application. Admission will be on a rolling basis, so applying earlier will improve your chances.

Update (July 10): The event is officially full. Feel free to fill out the application to be first in line regarding any future offerings! Or simply sign up for opportunities at https://fullstackdeeplearning.com.