Course Outline
Foundations of Machine Learning and Recurrent Neural Networks (RNN)
- Overview of NN and RNN
- Backpropagation
- Long Short-Term Memory (LSTM)
Introduction to TensorFlow
- Creating, initializing, saving, and restoring TensorFlow variables
- Feeding, reading, and preloading data in TensorFlow
- Utilizing TensorFlow infrastructure for large-scale model training
- Visualizing and evaluating models using TensorBoard
Core TensorFlow Mechanics
- Tutorial Files
- Data Preparation
- Download procedures
- Inputs and Placeholders
- Graph Construction
- Inference
- Loss Calculation
- Training Processes
- Model Training
- Understanding the Graph
- Managing the Session
- Implementing the Training Loop
- Model Evaluation
- Constructing the Evaluation Graph
- Generating Eval Output
Advanced Applications
- Threading and Queues
- Distributed TensorFlow
- Documentation and Model Sharing
- Customizing Data Readers
- Utilizing GPUs¹
- Manipulating TensorFlow Model Files
TensorFlow Serving
- Introduction
- Basic Serving Tutorial
- Advanced Serving Tutorial
- Serving an Inception Model
Convolutional Neural Networks
- Overview
- Goals
- Tutorial Highlights
- Model Architecture
- Code Organization
- CIFAR-10 Model
- Model Inputs
- Prediction Mechanisms
- Training Processes
- Launching and Training the Model
- Evaluating Model Performance
- Multi-GPU Training
- Assigning Variables and Operations to Devices
- Launching and Training Across Multiple GPU Cards
Deep Learning for MNIST
- Setup
- Loading MNIST Data
- Starting the TensorFlow InteractiveSession
- Building a Softmax Regression Model
- Using Placeholders
- Defining Variables
- Predicted Class and Cost Function
- Training the Model
- Evaluating the Model
- Constructing a Multilayer Convolutional Network
- Weight Initialization
- Convolution and Pooling Layers
- First Convolutional Layer
- Second Convolutional Layer
- Densely Connected Layer
- Readout Layer
- Training and Evaluating the Model
Image Recognition
- Inception-v3
- C++ Implementation
- Java Implementation
¹ GPU-related topics are not included in remote course offerings. These can be covered in classroom-based sessions only if agreed upon in advance, and provided that both the instructor and all participants have laptops equipped with supported NVIDIA GPUs running 64-bit Linux (hardware not supplied by NobleProg). NobleProg cannot guarantee the availability of instructors with the necessary hardware.
Requirements
- Python
Testimonials (1)
Very updated approach or CPI (tensor flow, era, learn) to do machine learning.
