Speakers: Luo Mai, Jingqing Zhang

Slides:



Advertisements
Similar presentations
A General Distributed Deep Learning Platform
Advertisements

Comparing TensorFlow Deep Learning Performance Using CPUs, GPUs, Local PCs and Cloud Pace University, Research Day, May 5, 2017 John Lawrence, Jonas Malmsten,
TensorFlow The Deep Learning Library You Should Be Using.
Conditional Generative Adversarial Networks
TensorFlow CS 5665 F16 practicum Karun Joseph, A Reference:
TensorFlow– A system for large-scale machine learning
Deep Learning Software: TensorFlow
Reinforcement Learning
Dimensionality Reduction and Principle Components Analysis
Zheng ZHANG 1-st year PhD candidate Group ILES, LIMSI
Early Results of Deep Learning on the Stampede2 Supercomputer
CS 388: Natural Language Processing: LSTM Recurrent Neural Networks
CS 4501: Introduction to Computer Vision Computer Vision + Natural Language Connelly Barnes Some slides from Fei-Fei Li / Andrej Karpathy / Justin Johnson.
Big Data A Quick Review on Analytical Tools
Deep Learning Amin Sobhani.
Automatic Lung Cancer Diagnosis from CT Scans (Week 2)
Convolutional Neural Fabrics by Shreyas Saxena, Jakob Verbeek
Spark Presentation.
Deep Learning Libraries
Ajita Rattani and Reza Derakhshani,
Inception and Residual Architecture in Deep Convolutional Networks
Synthesis of X-ray Projections via Deep Learning
CNN Demo LIU Pengpeng.
Neural networks (3) Regularization Autoencoder
A VERY Brief Introduction to Convolutional Neural Network using TensorFlow 李 弘
Comparison Between Deep Learning Packages
Efficient Deep Model for Monocular Road Segmentation
Shunyuan Zhang Nikhil Malik
Reinforcement learning with unsupervised auxiliary tasks
AI in Cyber-security: Examples of Algorithms & Techniques
Project Implementation for ITCS4122
Adversarially Tuned Scene Generation
Introduction to Deep Learning for neuronal data analyses
Digital Transformation Asia 2018 – CALL FOR SPEAKERS
Neural network systems
Deep Learning Packages
Early Results of Deep Learning on the Stampede2 Supercomputer
An open-source software library for Machine Intelligence
Image Processing Platform
MXNet Internals Cyrus M. Vahid, Principal Solutions Architect,
SAS Deep Learning: From Toolkit to Fast Model Prototyping
Scheduled Accomplishments
A Proposal Defense On Deep Residual Network For Face Recognition Presented By SAGAR MISHRA MECE
Neural Networks Geoff Hulten.
Vinit Shah, Joseph Picone and Iyad Obeid
Lip movement Synthesis from Text
Object Tracking: Comparison of
Sun Yat-sen University
Mark McKelvin EE249 Embedded System Design December 03, 2002
Neural networks (3) Regularization Autoencoder
Martin Schrimpf & Jon Gauthier MIT BCS Peer Lectures
Advances in Deep Audio and Audio-Visual Processing
Inception-v4, Inception-ResNet and the Impact of
Course Recap and What’s Next?
TensorFlow: A System for Large-Scale Machine Learning
Deep Learning Authors: Yann LeCun, Yoshua Bengio, Geoffrey Hinton
Course Summary Joseph E. Gonzalez
The Updated experiment based on LSTM
Deep Learning Libraries
CS295: Modern Systems: Application Case Study Neural Network Accelerator Sang-Woo Jun Spring 2019 Many slides adapted from Hyoukjun Kwon‘s Gatech “Designing.
Recent Advances in Neural Architecture Search
Cloud-DNN: An Open Framework for Mapping DNN Models to Cloud FPGAs
NFV and SD-WAN Multi vendor deployment
What's New in eCognition 9
Debasis Bhattacharya, JD, DBA University of Hawaii Maui College
Search-Based Approaches to Accelerate Deep Learning
CRCV REU 2019 Kara Schatz.
CRCV REU 2019 Aaron Honculada.
Morteza Kheirkhah University College London
Shengcong Chen, Changxing Ding, Minfeng Liu 2018
Presentation transcript:

Speakers: Luo Mai, Jingqing Zhang TensorLayer: A Versatile Deep Learning Library for Developers and Scientists Speakers: Luo Mai, Jingqing Zhang

Deep Learning Frameworks Key reasons for success Simplicity & Performance Large production adoption Large community ~ 115000 GitHub stars ~ 22000 GitHub stars ~ 15000 GitHub stars ~ 15000 GitHub stars

Wrapper libraries that bridge the gap Abstraction Mismatch Deep learning high-level elements: neural networks stacked with layers TensorFlow low-level API: dataflow graph, placeholder, session, queue runner, devices … Wrapper libraries that bridge the gap TFLearn

TensorLayer – Deep Learning Toolbox Design Training Deployment Reference networks Abstract layers Diverse learning functions Customizing layers Data augmentation Training control Distributed training Embedded platforms Model management

Comparison with Existing Libraries TensorLayer Keras TF-Learn Abstraction Modest High Flexibility Excellent Limited Portability No Yes Performance Good Deployment TensorLayer is a unique choice for its simplicity, flexibility and performance.

Flexible API to Build Complex Networks def model(x : tf.Tensor, train : bool) -> tl.Layer: with tf.variable_scope(”model”, reuse=TF.AUTO_REUSE): net = tl.InputLayer(x) net = tl.Conv2d(net, 64, (5, 5), (1, 1), padding='SAME') net = tl.BatchNormLayer(net, is_train=train, act=lambda: x:5*x) net = tl.MaxPool2d(net, (3, 3), (2, 2), pool=tf.nn.max_pool) net = tl.LambdaLayer(net, fn=tf.keras.Dropout) net = tl.DenseLayer(net, 10) return net Hide low-level TF APIs Flexible API Customizable Import Keras and TF-Slim TensorLayer has 100+ layers for computer vision, reinforcement learning and natural language processing ….

Zero-cost Abstraction Titan X Pascal GPU 2017 When training common CNNs, TensorLayer is usually 1.2 – 5x faster than Keras implementations

Scaling Training: Default Approach Synchronized Stochastic Gradient Descent (S-SGD) Extra works to include parameter server in TF programs Requirements from TensorLayer users: Minimal changes to user programs Homogeneous cluster High-performance Parameter Servers Parameters Local gradients GPU workers Optimal resource configuration is hard to determine

Scaling Training: TensorLayer + Horovod S-SGD through peer-to-peer communication Aggregate gradients Broadcast parameters GPU Workers Initialize environment Enable synchronization

Scaling Training: What Next? S-SGD is a weak scaling approach Experiment of training ResNet-32 Training collapse with large batch size Increased batch size does not reduce training time

Scaling Training: What Next? “If the training rate of some models is restricted to small batch sizes, then we will need to find other algorithmic and architectural approaches to their acceleration” – Jeff Dean & David Patterson

Scaling Training: What Next? Imperial College researchers are working towards strong scaling Constantly reduce training time Converge faster TensorLayer adopt timely research results to resolve practical AI problems

December 2018 ~ 4500 Stars ~ 1000 Forks ~ 70 Contributors Impact - Github December 2018 ~ 4500 Stars ~ 1000 Forks ~ 70 Contributors

Impact - Media and Award

Computer Vision OpenPose Plus TensorLayer Real-time pose estimation Multi-person keypoint detection Applications Automated grocery stores Security surveillance Smart home Deep Learning CNN VGG, VGG tiny, MobileNet TensorLayer Flexible architecture Distributed training TensorRT compatibility OpenPose Plus https://github.com/tensorlayer/openpose-plus

https://github.com/tensorlayer/srgan Adversarial Learning Super Resolution Generating realistic super-resolution images from their low-resolution counterparts. Applications Image recovery Image compression Deep Learning SR-GAN ResNet VGG TensorLayer Pretrained model Adversarial training   https://github.com/tensorlayer/srgan

Deep Poincare Map (DPM) for MRI Segmentation Medical Imaging Deep Poincare Map (DPM) for MRI Segmentation Cardiac segmentation in MRI Applications Medical imaging analysis Image segmentation Deep Learning CNN TensorLayer High performance data augmentation Visualization The Deep Poincare Map: A Novel Approach for Left Ventricle Segmentation. Y. Mo, F. Liu et al. MICCAI 2018.

Natural Language Processing Semantic Image Synthesis Image manipulation by natural language Applications Image manipulation Deep Learning VGG GAN LSTM TensorLayer More examples waiting for your exploration! Semantic Image Synthesis via Adversarial Learning. H. Dong, S. Yu et al. ICCV 2017.

Tutorials and Examples Actor-Critic DDPG RNN Distributed Training SRGAN A3C Policy Gradient VAE Inception Word2vec GAN-CLS Text generation Chatbot TRPO Data augmentation Autoencoder Anti-Spam DCGAN DAGGER LSTM CNN VAE-GAN Translation Semantic Image Synthesis Deep Q Network VGG Image-to-image translation Brain tumor segmentation

https://github.com/tensorlayer Luo Mai: Jingqing Zhang: luo.mai11@imperial.ac.uk  jz9215@imperial.ac.uk  https://github.com/tensorlayer