EncartaLabs

Apache SINGA

( Duration: 5 Days )

SINGA is a general distributed deep learning platform for training big deep learning models over large datasets. It is designed with an intuitive programming model based on the layer abstraction. A variety of popular deep learning models are supported, namely feed-forward models including convolutional neural networks (CNN), energy models like restricted Boltzmann machine (RBM), and recurrent neural networks (RNN).

In Apache SINGA training course delegates will learn to apply deep learning algorithms like CNN, RBM, and RNN with Apache SINGA

By attending Apache SINGA workshop, delegates will learn to:

  • Understand SINGA’s structure and deployment mechanisms
  • Be able to carry out installation / production environment / architecture tasks and configuration
  • Be able to assess code quality, perform debugging, monitoring
  • Be able to implement advanced production like training models, embedding terms, building graphs and logging

  • Researchers, Engineers and Developers seeking to utilize Apache SINGA as a deep learning framework.

COURSE AGENDA

1

Introduction

2

Installation

3

Quick Start

  • Training on a Single Node
  • Preparing data and job configuration
  • Training without parallelism
  • Asynchronous parallel training
  • Synchronous parallel training
  • Training in a Cluster
  • Starting Zookeeper
  • Training with GPUS
4

Programming

  • NeuralNet
    • Layer
    • Param
  • TrainOneBatch
  • Updater
5

Distributed Training

6

Data Preparation

  • Data-Loading and fetching data batches
  • Image Tool-Utility model for image augmentation
7

Training Linear Regression

  • Generating Training Data
  • Training via SGD
8

Training Multi-layer Perceptron (MLP)

  • Creating MLP Model
  • Training the model
  • Training convolutional neural network (CNN)
  • Data Preparation
  • Creating CNN Model
  • Initializing the Parameters
  • Setting up the optimizer and tensors
  • Conducting SGD
  • Saving and Loading the model
  • Prediction & Debugging
9

Train Char-Recurrent Neural Network (RNN)

  • Dataset Pre-processing
  • Creating the network
  • Conducting SGD
  • Checkpoint
  • Sample

Encarta Labs Advantage

  • One Stop Corporate Training Solution Providers for over 4,000 Modules on a variety of subjects
  • All courses are delivered by Industry Veterans
  • Get jumpstarted from newbie to production ready in a matter of few days
  • Trained more than 50,000 Corporate executives across the Globe
  • All our trainings are conducted in workshop mode with more focus on hands-on sessions

View our other course offerings by visiting http://encartalabs.com/course-catalogue-all.php

Contact us for delivering this course as a public/open-house workshop/online training for a group of 10+ candidates.

Top