deep learning demystified - appg · 2018-03-20 · 2 about me - deep learning solution architect @...
TRANSCRIPT
APPG AI
DEEP LEARNING DEMYSTIFIED
2
ABOUT ME
- Deep Learning Solution Architect @ NVIDIA - Supporting delivery of AI / Deep Learning solutions
- 10 years expereince deliverying Machine Learning of all scale (from embedded, mobile to Big Data)
- My past experience:
- Capgemini: https://goo.gl/MzgGbq
- Jaguar Land Rover Research: https://goo.gl/ar7LuU
Adam Grzywaczewski – [email protected]
3
NEURAL NETWORKS
4
NEURAL NETWORKS ARE NOT NEWAnd are disapintingly simple as an algorithm
5
NEURAL NETWORKS ARE NOT NEWAnd have historically never worked
6
WHY NOW?
7
LIFE AFTER MOORE’S LAW40 Years of Microprocessor Trend Data
1980 1990 2000 2010 2020
102
103
104
105
106
107
Single-threaded perf
1.5X per year
1.1X per yearTransistors
(thousands)
GPU-Computing perf
1.5X per year 1000X
By 2025
8
2016 – Baidu Deep Speech 2Superhuman Voice Recognition
2015 – Microsoft ResNetSuperhuman Image Recognition
2017 – Google Neural Machine TranslationNear Human Language Translation
100 ExaFLOPS8700 Million Parameters
20 ExaFLOPS300 Million Parameters
7 ExaFLOPS60 Million Parameters
For both compute and data
NEURAL NETWORK ARE GREEDY
9
100 EXFLOPS = 2 YEARS ON A DUAL SOCKET CPU SERVER
10
ONCE A FAILED MACHINE LEARNING EXPERIMENT
11
WHAT IS POSSIBLE TODAY?Overwhelming majority of our day to day activities
Based on a presentation from Andrew Ng
“If a typical person can do a mental task with less than one second of thought, we can probably automate it using AI either now or in the near future.”
Andrew Ng , Founder of Google Brain
12
EMPIRICAL EVIDENCE
13
14
MAKING IMPOSSIBLE PROBLEMS EXPENSIVE
Hestness, J., Narang, S., Ardalani, N., Diamos, G., Jun, H., Kianinejad, H., ... & Zhou, Y. (2017). Deep Learning Scaling is Predictable, Empirically. arXiv preprint arXiv:1712.00409.
With enough data we can find an approximation of every problem
• Translation
• Language Models
• Character Language Models
• Image Classification
• Attention Speech Models
15
MORE THAN JUSTIFIABLE EXCITEMENT
16
FUNDAMENTAL CHANGE TO THE ECONOMYImpact
17
AI IS THE NEW ELECTRICITY
“Just as electricity transformed almost everything 100 years ago, today I actually have a hard time thinking of an industry that I don’t think AI will transform in the
next several years, …”
Andrew Ng , Founder of Google Brain
Affecting every aspect of our lives
18
CALL TO ACTION
19
ENSURE VISIBILITY
It will affect all aspects of our lives so all departments
(they need skills to make unbiased decisions)
Entire UK Government needs to understand the technology
20
SUPPORT EDUCATION
It is possible to teach logic, software development AI and robotics in primary schools
(more important than ever)
At all levels
21
SUPPORT RESEARCH
Just a single research department of a large internet company has 10x more research infrastructure than the largest UK AI cluster.
There is no escaping from large datasets and as a consequence large compute.
Need to compete with large US and Chinese organisation
22
LAB OVERVIEW
23
DEEP LEARNING INSTITUTEDLI Mission
Helping people solve challenging problems using AI and deep learning.
• Developers, data scientists and engineers
• Self-driving cars, healthcare and robotics
• Training, optimizing, and deploying deep neural networks
24
HOW IT WORKS
25
HOW IT WORKS
26
HOW IT WORKS
27
HOW IT WORKS
28
LAUNCHING THE LAB
29
NAVIGATING TO QWIKLABS
1. Navigate to: https://nvlabs.qwiklab.com
1. Login or create a new account
30
ACCESSING LAB ENVIRONMENT
3. Select the event specific In-Session Class in the upper left
3. Click the “Image Classification with DIGITS” Class from the list
31
LAUNCHING THE LAB ENVIRONMENT
5. Click on the Select button to launch the lab environment
• After a short wait, lab Connection information will be shown
• Please ask Lab Assistants for help!
32
LAUNCHING THE LAB ENVIRONMENT
6. Click on the Start Lab button
33
LAUNCHING THE LAB ENVIRONMENT
You should see that the lab environment is “launching” towards the upper-right corner
34
CONNECTING TO THE LAB ENVIRONMENT
7. Click on “here” to access your lab environment / Jupyter notebook
35
CONNECTING TO THE LAB ENVIRONMENT
You should see your “Image Classification with DIGITS” Jupyternotebook
36
JUPYTER NOTEBOOK
1. Place your cursor in the code
2. Click the “run cell” button
2. Confirm you receive the same result
37
STARTING DIGITS
Instruction in Jupyter notebook will link you to DIGITS
38
ACCESSING DIGITS
• Will be prompted to enter a username to access DIGITS
• Can enter any username
• Use lower case letters