NVIDIA DLI Workshopdoe.cusat.ac.in/Nvidia_DLI.pdf · 2017-10-07 · NVIDIA DLI Workshop DEEP...
1
NVIDIA DLI Workshop DEEP LEARNING INSTITUTE Brought to you by 12-13 October 2017 The NVIDIA Deep Learning Institute (DLI) offers hands-on training for developers, data scientists, and researchers looking to solve the world’s most challenging problems with deep learning. The industrial revolution gave us the machine power and unprecedented growth but the need for a brain behind the wheels has never gone away. Ever since we have been designing automated systems, artificial intelligence has been the final frontier. In technical partnership with NVIDIA Deep Learning Institute, it gives us immense pleasure to invite you to take a one day instructor led DLI class as well as a GPU programming workshop on the succeeding day as per the details below. Attendees will receive a certificate of completion. Through self-paced online labs and instructor-led workshops, DLI provides training on the latest techniques for designing, training, and deploying neural networks across a variety of application domains. Explore widely used open-source frameworks as well as NVIDIA’s latest GPU-accelerated deep learning platforms. Deep Learning Demystified and Applied Deep Learning Image Classification with DIGITS Frameworks: Caffe Architect a Deep Neural Network to run on a GPU Manage the process of data preparation, model definition, model training and troubleshooting Use validation data to test and try different strategies for improving model performance On completion of this lab, you will be able to use NVIDIA DIGITS to architect, train, evaluate and enhance the accuracy of Convolutional Neural Networks (CNNs) on your own image classification application Lab Session Object Detection with DIGITS Frameworks: NVIDIA DIGITS Lab Session Explores three approaches to identify a specific feature within an image Neural Network Deployment with DIGITS and TensorRT Lab Session Frameworks: Caffe Make various optimizations in the inference process Understand the role of batch size in inference performance Explore inference for a variety of different DNN architectures trained in other DLI labs On completion of this lab, you will be able to execute a full Deep Learning workflow: from loading data, to training a neural network, to deploying that trained network to production Lecture Session Greetings from Department of Electronics, CUSAT Introduction to Heterogeneous Computing GPU Programming Architecture: CUDA Parallelism using Threads Flynns Taxonomy: GPU Programming with SPMD GPU Memory Hierarchy GPU Architecture ( PCI - NVLINK, Kepler and Pascal ) Related Programming Model: OpenACC + CUDA Hands-On Exercise GPU Accelerated Application, Libraries and Case studies In-House Application Development and Collaboration opportunities Identify application/s to port/optimize on CUDA architecture in collaboration with Nvidia Deep Learning Bootcamp-12 th October 2017 Arun A. Balakrishnan / Mithun Haridas T. P. Department of Electronics CUSAT In association with For more information, please contact: Prerequisite: Basic Knowledge of C/C++, Computer Architecture Basics of GPU Programming-13 th October 2017 Department of Electronics CUSAT For further details, visit: Register at: doe.cusat.ac.in/news.php#nvidia https://developer.nvidia.com/dli/onlinelabs Read more at: Signal processing, control systems, robotics, machine vision, communication systems, every facet of modern electronics is in the phase of an imminent revolution, nothing going to be untouched. Artificial intelligence, led by the recent breakthroughs in machine learning especially the Deep learning algorithms, leveraged by the exponential growth in computational power and massive amounts of data has helped to cross the knee point over to the steeps. As a technology pioneer in this domain, NVIDIA GPUs have democratized supercomputing to desktop scales and even to SoCs. Already, organizations are using deep learning to transform moonshots into real results. Building intelligent machines that can perceive the world as we do, understand our language, and navigate around on their own has remained a dream over the last few decades and now we are in the brink of the next giant leap in the human history. Register @ doe.cusat.ac.in Venue: Auditorium, Department of Electronics, CUSAT Time: 9.00 am - 5.00 pm Each approach is measured in relation to three metrics: model training time, model accuracy and speed of detection during deployment On completion of this lab, you will understand the merits of each approach and learn how to detect objects using neural networks trained on NVIDIA DIGITS on real-world datasets +919496346370/+919447096888 [email protected] / [email protected] Register @ Industry-Academia Collaboration Programme Together Further As a part of Industry-Academia Collaboration Programme DOE - CUSAT Together Further As a part of Industry-Academia Collaboration Programme DOE - CUSAT
Transcript of NVIDIA DLI Workshopdoe.cusat.ac.in/Nvidia_DLI.pdf · 2017-10-07 · NVIDIA DLI Workshop DEEP...
NVIDIA DLI WorkshopDEEP LEARNING INSTITUTE
Brought to you by
12-13 October 2017
The NVIDIA Deep Learning Institute (DLI) offers hands-on training for developers, data scientists, and researchers looking to solve the world’s most challenging problems with deep learning.
The industrial revolution gave us the machine power and unprecedented growth but the need for a brain behind the wheels has never gone away. Ever since we have been designing automated systems, artificial intelligence has been the final frontier.
In technical partnership with NVIDIA Deep Learning Institute, it gives us immense pleasure to invite you to take a one day instructor led DLI class as well as a GPU programming workshop on the succeeding day as per the details below. Attendees will receive a certificate of completion.
Through self-paced online labs and instructor-led workshops, DLI provides training on the latest techniques for designing, training, and deploying neural networks across a variety of application domains. Explore widely used open-source frameworks as well as NVIDIA’s latest GPU-accelerated deep learning platforms.
Deep Learning Demystified and Applied Deep Learning
Image Classification with DIGITS Frameworks: Caffe
Architect a Deep Neural Network to run on a GPUManage the process of data preparation, model definition, model training and troubleshootingUse validation data to test and try different strategies for improving model performanceOn completion of this lab, you will be able to use NVIDIA DIGITS to architect, train, evaluate and enhance the accuracy of Convolutional Neural Networks (CNNs) on your own image classification application
Lab Session
Object Detection with DIGITS Frameworks: NVIDIA DIGITSLab Session
Explores three approaches to identify a specific feature within an image
Neural Network Deployment with DIGITS and TensorRT Lab Session Frameworks: Caffe
Make various optimizations in the inference processUnderstand the role of batch size in inference performance
Explore inference for a variety of different DNN architectures trained in other DLI labsOn completion of this lab, you will be able to execute a full Deep Learning workflow: from loading data, to training a neural network, to deploying that trained network to production
Lecture Session
Greetings from Department of Electronics, CUSAT
Introduction to Heterogeneous Computing
GPU Programming Architecture: CUDA Parallelism using Threads
Flynns Taxonomy: GPU Programming with SPMD
GPU Memory Hierarchy
GPU Architecture ( PCI - NVLINK, Kepler and Pascal )
Related Programming Model: OpenACC + CUDA
Hands-On Exercise
GPU Accelerated Application, Libraries and Case studies
In-House Application Development and Collaboration opportunities
Identify application/s to port/optimize on CUDA architecture in collaboration with Nvidia
Deep Learning Bootcamp-12th October 2017
Arun A. Balakrishnan / Mithun Haridas T. P.
Department of Electronics CUSAT
In association with
For more information, please contact:
Prerequisite: Basic Knowledge of C/C++, Computer Architecture
Basics of GPU Programming-13th October 2017
Department of Electronics CUSAT
For further details, visit:
Register at: doe.cusat.ac.in/news.php#nvidia
https://developer.nvidia.com/dli/onlinelabs Read more at:
Signal processing, control systems, robotics, machine vision, communication systems, every facet of modern electronics is in the phase of an imminent revolution, nothing going to be untouched. Artificial intelligence, led by the recent breakthroughs in machine learning especially the Deep learning algorithms, leveraged by the exponential growth in computational power and massive amounts of data has helped to cross the knee point over to the steeps.
As a technology pioneer in this domain, NVIDIA GPUs have democratized supercomputing to desktop scales and even to SoCs. Already, organizations are using deep learning to transform moonshots into real results.
Building intelligent machines that can perceive the world as we do, understand our language, and navigate around on their own has remained a dream over the last few decades and now we are in the brink of the next giant leap in the human history.
Register @
doe.cusat.ac.in
Venue: Auditorium, Department of Electronics, CUSAT Time: 9.00 am - 5.00 pm
Each approach is measured in relation to three metrics: model training time, model accuracy and speed of detection during deployment On completion of this lab, you will understand the merits of each approach and learn how to detect objects using neural networks trained on NVIDIA DIGITS on real-world datasets
+919496346370/[email protected] / [email protected]
Register @
Industry-Academia Collaboration Programme
TogetherFurther
As a part of
Industry-Academia Collaboration ProgrammeDOE - CUSAT
TogetherFurther
As a part of
Industry-Academia Collaboration ProgrammeDOE - CUSAT
