SlideShare une entreprise Scribd logo
1  sur  31
Eran Shlomo, IPP tech lead, Haifa
eran.shlomo@intel.com, eran@dataloop.ai
About me
Haifa IoT Ignition lab and IPP(Intel ingenuity partnership program) tech lead.
Intel Perceptual computing.
Compute, cloud and embedded expert.
Maker and Entrepreneur
Focus on Data science and Machine learning in recent years
Soon to work on dataloop.ai
Agenda
What is deep learning
Why now ?
Different network topologies and their usage
The tools race
The processors (HW) race
Buzzwords alignment attempt
AI
Machine
learning
Supervised
learning
Deep
learning
Machine reasoning
Automated tasks
Train based on data
Neural networks
input
logic
output
input
output
logic
Deep learning – basic anatomy
Data driven
Training a model
Input, output and hidden neurons
Input layer Hidden layer(s) Output layer
Deep learning Many hidden (deep) layers
The essence of deeplearning
Xi YiWij(1) Wij(2)
W11(1)
X1
Y1
W11(2)
𝑌 = 𝑓 𝑋 = 𝑊𝑋+b
Deep network is essentially a function
we train to detect some pattern
b (bias) is omitted in this drawing
Data is becoming the fuel behind new SW,
BK (Intel CEO) – “Data is the new Oil”
Why now ?
28.2
25.8
16.4
11.7
6.7
3.57 2.99
2010 2011 2012 2013 2014 2015 2016
ILSVRC top 5 error
ILSVRC top 5 error
8 Layers 22 152
Alexnet
Shallow Ensemble
Data
Neural networks – Background and inspiration
It is pretty common to compare neural networks to how our brain works:
• Coupled well with the term AI
• Has some sense in it, as many different researches show. Yet we are a bit long from really understanding
how the brain works.
𝑘=0
𝑛
𝑊𝑋
W1
W2
W3
X1
X2
X3
𝑓(𝑥)
Network topologies
• There are many network topologies
• The basic principles apply:
• Supervised
• hidden units
• backpropagation training is common to most
• Training on data generates model, later to be used to inference on unseen data:
• Minimize a cost function
Some basic intuition
Model have capacity  Number of parameters.
Generally HW (compute/mem) limits the capacity
From the Paper: AN ANALYSIS OF DEEP NEURAL NETWORK MODELS
FOR PRACTICAL APPLICATIONS
More
compute
& Data
Higher
accuracy
Bigger
model
Model fit scenarios
0
20
40
60
80
100
120
140
160
0 5 10 15
Good model
0
20
40
60
80
100
120
140
160
0 5 10 15
Underfit/High
bias
0
20
40
60
80
100
120
140
160
0 5 10 15
Overfit/ High
Variance
Training model  Bias/Variance “games”
We can look at our model error as follows:
noise
model
error
Total
Error
Our error usually comes from combination of these two, These are all equivalent:
• High variance=modeling noise=not enough data=model too big=overfit
• High bias =model too simple=underfit
Basic network types
Fully connected networks
A very basic/generic network, Full nodes
connectivity
Used as a building block in more complex
topologies
High level task: Maps features into classes
Convolutional neural networks
On very simple images fully connected networks work pretty well with images
converted into vectors, but:
• Simple images (~10x10) works well, bigger images (~100x100) don’t:
• Too much data(parameters) is needed in order to train FC networks that way, not
practical. 100x100 image 10K pixel, 2 layer FC network 100M parameters.
Entering convolutional neural networks:
• Encodes special dependency, kind of Wight sharing
• Two main parts:
• Conv/Subsample acts as feature generators
• FC maps feature ensemble into classes
Recurrent neural networks
In general neural networks works well on bounded
areas, AKA the data collected to train.
In order to predict time series data (like stocks, ...) we
need time factor.
RNNs:
• Neurons as self connected
• Backpropegated through time.
• Each time stamp is now considered a laeyer.
• Issue: We need deep network  Many layers 
Vanishing gradient problem
Long Short Term memory networks
Solves the vanishing gradient problem, Long
memory by default
Contains gates that act as decision points
Usually LSTMs are proffered over RNN , more
compute is needed per timestamp but overall
accuracy is better.
Tools
Assembly C (compiler) C++(OOP) JAVA(managed)
Python (run
time)
Where we are in technology timeline perspective
Model
protos
High level
(e.g.
keras)
???? ???? ????
The programming language
Science Data science and deep learning are very close friends.
All are frontend languages with performant backend language (C++)
3 main languages:
My personal take … :
Python is the leading language:
• Free
• Won the deep learning community
• Most of the new tools / frameworks are python friendly.
• Production friendly
• Easy low level binding
Frameworks
Big frameworks supported by environment
Caffee
TensorFlow
MXNet
Keras
Torch
CNTK
Theano
Good comparision reference : https://github.com/zer0n/deepframeworks
Nnet
MXNet
Darch
deepnet
H2O
Neural networks toolbox
The big data/Cloud arena
All major cloud providers have ML services, deep learning model development
included.
Many other dedicated cloud services , some already acquired by tier 1 providers:
• Nervana
• Databricks
• Turi (GraphLab)
• H2O
• ..
The HW arena
Currently NVIDIA rules
Market top level segmentation:
• Training – building the
model, Data center
• Inference – Running the
model, also edge/client
In the short term intel is
positioned to take significant
inference market share (SW
moves only, existing x86 hw).
The (rough) deep learning compute math
• We have model capacity
• We have chip capacity
• Throughput = chip capacity/model capacity
But the story have few twists, It turns out that:
• Models can work well with low precision parameters
• A lot of sparse areas
• Memory plays significant role as well
New compute architectures wave is coming
Handle 16,8,4,2,1 bit
networks
Expect 100-300x
effective compute boost
Memory paths
adjustments
The race to the AI silicon has kicked off
Everybody is playing: Startups, Technology companies (Verticals), Corporations
Segments of the game:
• ASIC VS FPGA
• Edge VS cloud
• Inference VS training
• Network Generic VS network specific
• Models Arch/Eco-system
Deep learning @ Intel
the AI era – New A
group
Academia
Development
Training and programs
A lot of HW/SW activity, The public
ones 
Knights Mill
Intel FPGA SDK
eran.shlomo@intel.com,
eran@dataloop.ai

Contenu connexe

Tendances

Deep Learning Made Easy with Deep Features
Deep Learning Made Easy with Deep FeaturesDeep Learning Made Easy with Deep Features
Deep Learning Made Easy with Deep Features
Turi, Inc.
 

Tendances (20)

On-device machine learning: TensorFlow on Android
On-device machine learning: TensorFlow on AndroidOn-device machine learning: TensorFlow on Android
On-device machine learning: TensorFlow on Android
 
Recent developments in Deep Learning
Recent developments in Deep LearningRecent developments in Deep Learning
Recent developments in Deep Learning
 
Deep learning on mobile - 2019 Practitioner's Guide
Deep learning on mobile - 2019 Practitioner's GuideDeep learning on mobile - 2019 Practitioner's Guide
Deep learning on mobile - 2019 Practitioner's Guide
 
Introduction to Deep Learning and neon at Galvanize
Introduction to Deep Learning and neon at GalvanizeIntroduction to Deep Learning and neon at Galvanize
Introduction to Deep Learning and neon at Galvanize
 
Deep learning on mobile
Deep learning on mobileDeep learning on mobile
Deep learning on mobile
 
Smaller and Easier: Machine Learning on Embedded Things
Smaller and Easier: Machine Learning on Embedded ThingsSmaller and Easier: Machine Learning on Embedded Things
Smaller and Easier: Machine Learning on Embedded Things
 
AWS re:Invent 2016: Using MXNet for Recommendation Modeling at Scale (MAC306)
AWS re:Invent 2016: Using MXNet for Recommendation Modeling at Scale (MAC306)AWS re:Invent 2016: Using MXNet for Recommendation Modeling at Scale (MAC306)
AWS re:Invent 2016: Using MXNet for Recommendation Modeling at Scale (MAC306)
 
Amazon Deep Learning
Amazon Deep LearningAmazon Deep Learning
Amazon Deep Learning
 
GDG-Shanghai 2017 TensorFlow Summit Recap
GDG-Shanghai 2017 TensorFlow Summit RecapGDG-Shanghai 2017 TensorFlow Summit Recap
GDG-Shanghai 2017 TensorFlow Summit Recap
 
Diving into Deep Learning (Silicon Valley Code Camp 2017)
Diving into Deep Learning (Silicon Valley Code Camp 2017)Diving into Deep Learning (Silicon Valley Code Camp 2017)
Diving into Deep Learning (Silicon Valley Code Camp 2017)
 
Deep Learning as a Cat/Dog Detector
Deep Learning as a Cat/Dog DetectorDeep Learning as a Cat/Dog Detector
Deep Learning as a Cat/Dog Detector
 
Scalable Deep Learning Using Apache MXNet
Scalable Deep Learning Using Apache MXNetScalable Deep Learning Using Apache MXNet
Scalable Deep Learning Using Apache MXNet
 
Deep Learning on Qubole Data Platform
Deep Learning on Qubole Data PlatformDeep Learning on Qubole Data Platform
Deep Learning on Qubole Data Platform
 
Why is Deep learning hot right now? and How can we apply it on each day job?
Why is Deep learning hot right now? and How can we apply it on each day job?Why is Deep learning hot right now? and How can we apply it on each day job?
Why is Deep learning hot right now? and How can we apply it on each day job?
 
(BDT311) Deep Learning: Going Beyond Machine Learning
(BDT311) Deep Learning: Going Beyond Machine Learning(BDT311) Deep Learning: Going Beyond Machine Learning
(BDT311) Deep Learning: Going Beyond Machine Learning
 
Deep Learning Made Easy with Deep Features
Deep Learning Made Easy with Deep FeaturesDeep Learning Made Easy with Deep Features
Deep Learning Made Easy with Deep Features
 
"New Dataflow Architecture for Machine Learning," a Presentation from Wave Co...
"New Dataflow Architecture for Machine Learning," a Presentation from Wave Co..."New Dataflow Architecture for Machine Learning," a Presentation from Wave Co...
"New Dataflow Architecture for Machine Learning," a Presentation from Wave Co...
 
Deep Learning for Robotics
Deep Learning for RoboticsDeep Learning for Robotics
Deep Learning for Robotics
 
Language translation with Deep Learning (RNN) with TensorFlow
Language translation with Deep Learning (RNN) with TensorFlowLanguage translation with Deep Learning (RNN) with TensorFlow
Language translation with Deep Learning (RNN) with TensorFlow
 
Josh Patterson, Advisor, Skymind – Deep learning for Industry at MLconf ATL 2016
Josh Patterson, Advisor, Skymind – Deep learning for Industry at MLconf ATL 2016Josh Patterson, Advisor, Skymind – Deep learning for Industry at MLconf ATL 2016
Josh Patterson, Advisor, Skymind – Deep learning for Industry at MLconf ATL 2016
 

Similaire à The deep learning tour - Q1 2017

Learn to Build an App to Find Similar Images using Deep Learning- Piotr Teterwak
Learn to Build an App to Find Similar Images using Deep Learning- Piotr TeterwakLearn to Build an App to Find Similar Images using Deep Learning- Piotr Teterwak
Learn to Build an App to Find Similar Images using Deep Learning- Piotr Teterwak
PyData
 
TECHNICAL OVERVIEW NVIDIA DEEP LEARNING PLATFORM Giant Leaps in Performance ...
TECHNICAL OVERVIEW NVIDIA DEEP  LEARNING PLATFORM Giant Leaps in Performance ...TECHNICAL OVERVIEW NVIDIA DEEP  LEARNING PLATFORM Giant Leaps in Performance ...
TECHNICAL OVERVIEW NVIDIA DEEP LEARNING PLATFORM Giant Leaps in Performance ...
Willy Marroquin (WillyDevNET)
 

Similaire à The deep learning tour - Q1 2017 (20)

Deep learning with TensorFlow
Deep learning with TensorFlowDeep learning with TensorFlow
Deep learning with TensorFlow
 
Learn to Build an App to Find Similar Images using Deep Learning- Piotr Teterwak
Learn to Build an App to Find Similar Images using Deep Learning- Piotr TeterwakLearn to Build an App to Find Similar Images using Deep Learning- Piotr Teterwak
Learn to Build an App to Find Similar Images using Deep Learning- Piotr Teterwak
 
Open source ai_technical_trend
Open source ai_technical_trendOpen source ai_technical_trend
Open source ai_technical_trend
 
Big Data Analytics (ML, DL, AI) hands-on
Big Data Analytics (ML, DL, AI) hands-onBig Data Analytics (ML, DL, AI) hands-on
Big Data Analytics (ML, DL, AI) hands-on
 
Synthetic dialogue generation with Deep Learning
Synthetic dialogue generation with Deep LearningSynthetic dialogue generation with Deep Learning
Synthetic dialogue generation with Deep Learning
 
AWS re:Invent 2016: Bringing Deep Learning to the Cloud with Amazon EC2 (CMP314)
AWS re:Invent 2016: Bringing Deep Learning to the Cloud with Amazon EC2 (CMP314)AWS re:Invent 2016: Bringing Deep Learning to the Cloud with Amazon EC2 (CMP314)
AWS re:Invent 2016: Bringing Deep Learning to the Cloud with Amazon EC2 (CMP314)
 
A Platform for Accelerating Machine Learning Applications
 A Platform for Accelerating Machine Learning Applications A Platform for Accelerating Machine Learning Applications
A Platform for Accelerating Machine Learning Applications
 
MXNet Workshop
MXNet WorkshopMXNet Workshop
MXNet Workshop
 
Austin,TX Meetup presentation tensorflow final oct 26 2017
Austin,TX Meetup presentation tensorflow final oct 26 2017Austin,TX Meetup presentation tensorflow final oct 26 2017
Austin,TX Meetup presentation tensorflow final oct 26 2017
 
TECHNICAL OVERVIEW NVIDIA DEEP LEARNING PLATFORM Giant Leaps in Performance ...
TECHNICAL OVERVIEW NVIDIA DEEP  LEARNING PLATFORM Giant Leaps in Performance ...TECHNICAL OVERVIEW NVIDIA DEEP  LEARNING PLATFORM Giant Leaps in Performance ...
TECHNICAL OVERVIEW NVIDIA DEEP LEARNING PLATFORM Giant Leaps in Performance ...
 
Faster deep learning solutions from training to inference - Amitai Armon & Ni...
Faster deep learning solutions from training to inference - Amitai Armon & Ni...Faster deep learning solutions from training to inference - Amitai Armon & Ni...
Faster deep learning solutions from training to inference - Amitai Armon & Ni...
 
Track2 02. machine intelligence at google scale google, kaz sato, staff devel...
Track2 02. machine intelligence at google scale google, kaz sato, staff devel...Track2 02. machine intelligence at google scale google, kaz sato, staff devel...
Track2 02. machine intelligence at google scale google, kaz sato, staff devel...
 
AI Deep Learning - CF Machine Learning
AI Deep Learning - CF Machine LearningAI Deep Learning - CF Machine Learning
AI Deep Learning - CF Machine Learning
 
Artificial Intelligence and Deep Learning in Azure, CNTK and Tensorflow
Artificial Intelligence and Deep Learning in Azure, CNTK and TensorflowArtificial Intelligence and Deep Learning in Azure, CNTK and Tensorflow
Artificial Intelligence and Deep Learning in Azure, CNTK and Tensorflow
 
Big Data & Artificial Intelligence
Big Data & Artificial IntelligenceBig Data & Artificial Intelligence
Big Data & Artificial Intelligence
 
Accelerating TensorFlow with RDMA for high-performance deep learning
Accelerating TensorFlow with RDMA for high-performance deep learningAccelerating TensorFlow with RDMA for high-performance deep learning
Accelerating TensorFlow with RDMA for high-performance deep learning
 
Machine learning the next revolution or just another hype
Machine learning   the next revolution or just another hypeMachine learning   the next revolution or just another hype
Machine learning the next revolution or just another hype
 
Dog Breed Classification using PyTorch on Azure Machine Learning
Dog Breed Classification using PyTorch on Azure Machine LearningDog Breed Classification using PyTorch on Azure Machine Learning
Dog Breed Classification using PyTorch on Azure Machine Learning
 
Deep Learning Demystified
Deep Learning DemystifiedDeep Learning Demystified
Deep Learning Demystified
 
AIoT: Intelligence on Microcontroller
AIoT: Intelligence on MicrocontrollerAIoT: Intelligence on Microcontroller
AIoT: Intelligence on Microcontroller
 

Plus de Eran Shlomo

Plus de Eran Shlomo (7)

Practical deep learning for computer vision
Practical deep learning for computer visionPractical deep learning for computer vision
Practical deep learning for computer vision
 
Deep learning from scratch
Deep learning from scratch Deep learning from scratch
Deep learning from scratch
 
Industrial internet of things
Industrial internet of thingsIndustrial internet of things
Industrial internet of things
 
PyCourse - Self driving python course
PyCourse - Self driving python coursePyCourse - Self driving python course
PyCourse - Self driving python course
 
Imagine. Capture. Create. Interact
Imagine. Capture.Create. InteractImagine. Capture.Create. Interact
Imagine. Capture. Create. Interact
 
Python - The Good, The Bad and The ugly
Python - The Good, The Bad and The ugly Python - The Good, The Bad and The ugly
Python - The Good, The Bad and The ugly
 
Internet of things - 2016 trends.
Internet of things - 2016 trends. Internet of things - 2016 trends.
Internet of things - 2016 trends.
 

Dernier

Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slide
vu2urc
 

Dernier (20)

AWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of TerraformAWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of Terraform
 
Boost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfBoost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdf
 
HTML Injection Attacks: Impact and Mitigation Strategies
HTML Injection Attacks: Impact and Mitigation StrategiesHTML Injection Attacks: Impact and Mitigation Strategies
HTML Injection Attacks: Impact and Mitigation Strategies
 
[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf
 
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, AdobeApidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
 
2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...
 
Automating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps ScriptAutomating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps Script
 
Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slide
 
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
 
Exploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone ProcessorsExploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone Processors
 
Advantages of Hiring UIUX Design Service Providers for Your Business
Advantages of Hiring UIUX Design Service Providers for Your BusinessAdvantages of Hiring UIUX Design Service Providers for Your Business
Advantages of Hiring UIUX Design Service Providers for Your Business
 
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemkeProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
 
Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024
 
🐬 The future of MySQL is Postgres 🐘
🐬  The future of MySQL is Postgres   🐘🐬  The future of MySQL is Postgres   🐘
🐬 The future of MySQL is Postgres 🐘
 
Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024
 
GenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day PresentationGenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day Presentation
 
Driving Behavioral Change for Information Management through Data-Driven Gree...
Driving Behavioral Change for Information Management through Data-Driven Gree...Driving Behavioral Change for Information Management through Data-Driven Gree...
Driving Behavioral Change for Information Management through Data-Driven Gree...
 
Understanding Discord NSFW Servers A Guide for Responsible Users.pdf
Understanding Discord NSFW Servers A Guide for Responsible Users.pdfUnderstanding Discord NSFW Servers A Guide for Responsible Users.pdf
Understanding Discord NSFW Servers A Guide for Responsible Users.pdf
 
Boost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivityBoost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivity
 
Developing An App To Navigate The Roads of Brazil
Developing An App To Navigate The Roads of BrazilDeveloping An App To Navigate The Roads of Brazil
Developing An App To Navigate The Roads of Brazil
 

The deep learning tour - Q1 2017

  • 1. Eran Shlomo, IPP tech lead, Haifa eran.shlomo@intel.com, eran@dataloop.ai
  • 2. About me Haifa IoT Ignition lab and IPP(Intel ingenuity partnership program) tech lead. Intel Perceptual computing. Compute, cloud and embedded expert. Maker and Entrepreneur Focus on Data science and Machine learning in recent years Soon to work on dataloop.ai
  • 3. Agenda What is deep learning Why now ? Different network topologies and their usage The tools race The processors (HW) race
  • 4. Buzzwords alignment attempt AI Machine learning Supervised learning Deep learning Machine reasoning Automated tasks Train based on data Neural networks input logic output input output logic
  • 5. Deep learning – basic anatomy Data driven Training a model Input, output and hidden neurons Input layer Hidden layer(s) Output layer Deep learning Many hidden (deep) layers
  • 6. The essence of deeplearning Xi YiWij(1) Wij(2) W11(1) X1 Y1 W11(2) 𝑌 = 𝑓 𝑋 = 𝑊𝑋+b Deep network is essentially a function we train to detect some pattern b (bias) is omitted in this drawing Data is becoming the fuel behind new SW, BK (Intel CEO) – “Data is the new Oil”
  • 7. Why now ? 28.2 25.8 16.4 11.7 6.7 3.57 2.99 2010 2011 2012 2013 2014 2015 2016 ILSVRC top 5 error ILSVRC top 5 error 8 Layers 22 152 Alexnet Shallow Ensemble Data
  • 8. Neural networks – Background and inspiration It is pretty common to compare neural networks to how our brain works: • Coupled well with the term AI • Has some sense in it, as many different researches show. Yet we are a bit long from really understanding how the brain works. 𝑘=0 𝑛 𝑊𝑋 W1 W2 W3 X1 X2 X3 𝑓(𝑥)
  • 9. Network topologies • There are many network topologies • The basic principles apply: • Supervised • hidden units • backpropagation training is common to most • Training on data generates model, later to be used to inference on unseen data: • Minimize a cost function
  • 10. Some basic intuition Model have capacity  Number of parameters. Generally HW (compute/mem) limits the capacity From the Paper: AN ANALYSIS OF DEEP NEURAL NETWORK MODELS FOR PRACTICAL APPLICATIONS More compute & Data Higher accuracy Bigger model
  • 11. Model fit scenarios 0 20 40 60 80 100 120 140 160 0 5 10 15 Good model 0 20 40 60 80 100 120 140 160 0 5 10 15 Underfit/High bias 0 20 40 60 80 100 120 140 160 0 5 10 15 Overfit/ High Variance
  • 12. Training model  Bias/Variance “games” We can look at our model error as follows: noise model error Total Error Our error usually comes from combination of these two, These are all equivalent: • High variance=modeling noise=not enough data=model too big=overfit • High bias =model too simple=underfit
  • 14. Fully connected networks A very basic/generic network, Full nodes connectivity Used as a building block in more complex topologies High level task: Maps features into classes
  • 15. Convolutional neural networks On very simple images fully connected networks work pretty well with images converted into vectors, but: • Simple images (~10x10) works well, bigger images (~100x100) don’t: • Too much data(parameters) is needed in order to train FC networks that way, not practical. 100x100 image 10K pixel, 2 layer FC network 100M parameters. Entering convolutional neural networks: • Encodes special dependency, kind of Wight sharing • Two main parts: • Conv/Subsample acts as feature generators • FC maps feature ensemble into classes
  • 16. Recurrent neural networks In general neural networks works well on bounded areas, AKA the data collected to train. In order to predict time series data (like stocks, ...) we need time factor. RNNs: • Neurons as self connected • Backpropegated through time. • Each time stamp is now considered a laeyer. • Issue: We need deep network  Many layers  Vanishing gradient problem
  • 17. Long Short Term memory networks Solves the vanishing gradient problem, Long memory by default Contains gates that act as decision points Usually LSTMs are proffered over RNN , more compute is needed per timestamp but overall accuracy is better.
  • 18. Tools
  • 19. Assembly C (compiler) C++(OOP) JAVA(managed) Python (run time) Where we are in technology timeline perspective Model protos High level (e.g. keras) ???? ???? ????
  • 20. The programming language Science Data science and deep learning are very close friends. All are frontend languages with performant backend language (C++) 3 main languages: My personal take … : Python is the leading language: • Free • Won the deep learning community • Most of the new tools / frameworks are python friendly. • Production friendly • Easy low level binding
  • 21. Frameworks Big frameworks supported by environment Caffee TensorFlow MXNet Keras Torch CNTK Theano Good comparision reference : https://github.com/zer0n/deepframeworks Nnet MXNet Darch deepnet H2O Neural networks toolbox
  • 22. The big data/Cloud arena All major cloud providers have ML services, deep learning model development included. Many other dedicated cloud services , some already acquired by tier 1 providers: • Nervana • Databricks • Turi (GraphLab) • H2O • ..
  • 24. Currently NVIDIA rules Market top level segmentation: • Training – building the model, Data center • Inference – Running the model, also edge/client In the short term intel is positioned to take significant inference market share (SW moves only, existing x86 hw).
  • 25. The (rough) deep learning compute math • We have model capacity • We have chip capacity • Throughput = chip capacity/model capacity But the story have few twists, It turns out that: • Models can work well with low precision parameters • A lot of sparse areas • Memory plays significant role as well
  • 26. New compute architectures wave is coming Handle 16,8,4,2,1 bit networks Expect 100-300x effective compute boost Memory paths adjustments
  • 27. The race to the AI silicon has kicked off Everybody is playing: Startups, Technology companies (Verticals), Corporations Segments of the game: • ASIC VS FPGA • Edge VS cloud • Inference VS training • Network Generic VS network specific • Models Arch/Eco-system
  • 29. the AI era – New A group Academia Development Training and programs
  • 30. A lot of HW/SW activity, The public ones  Knights Mill Intel FPGA SDK

Notes de l'éditeur

  1. Smart Home Industry 4.0 Retail Autonomous cars Robotics Medical FinTech Cognitive computing 5G Wearables