Overview
The “Generative AI” course provides a comprehensive understanding of generative artificial intelligence models and techniques. Participants will learn about the underlying principles of generative models, such as generative adversarial networks (GANs) and variational autoencoders (VAEs), and explore various applications of generative AI in image synthesis, text generation, music composition, and more.
Objectives
At the end of Genrative AI course, participants will be able to
- Gain a deep understanding of generative AI models and their applications. | Learn about popular generative models like GANs and VAEs. | Explore techniques for image synthesis, text generation, and music composition using generative AI. | Understand evaluation metrics and approaches for assessing generative models. | Apply generative AI techniques in real-world projects and creative endeavors. | Gain a deep understanding of generative AI models and their applications.
- Learn about popular generative models like GANs and VAEs.
- Explore techniques for image synthesis, text generation, and music composition using generative AI.
- Understand evaluation metrics and approaches for assessing generative models.
- Apply generative AI techniques in real-world projects and creative endeavors.
Prerequisites
Basic knowledge of machine learning concepts and algorithms. Familiarity with programming languages such as Python. Understanding of neural networks and deep learning principles. Prior experience with popular deep learning frameworks like TensorFlow or PyTorch is beneficial but not mandatory.
Course Outline
- Machine Learning and Industry use casesML use in semiconductor IndustryML use in Chip Design
- Unsupervised Learning & Cluster AnalysisSupervised vs unsupervised learningNeed of Cluster AnalysisK- Means clustering algorithmThe theory behind cluster AnalysisBuilding and interpreting clusters
- Decision TreesSegmentationEntropyInformation gainBuilding Decision TreesValidation of TreesPruning the treesFine tuning the treesPrediction using TreesCustomer retention case study
- Random ForestEnsemble LearningEnsemble ModelsBaggingBoosting
- ML use in semiconductor Industry
- ML use in Chip Design
- Supervised vs unsupervised learning
- Need of Cluster Analysis
- K- Means clustering algorithm
- The theory behind cluster Analysis
- Building and interpreting clusters
- Segmentation
- Entropy
- Information gain
- Building Decision Trees
- Validation of Trees
- Pruning the trees
- Fine tuning the trees
- Prediction using Trees
- Customer retention case study
- Ensemble Learning
- Ensemble Models
- Bagging
- Boosting
- Introduction to Deep LearningOverview of Natural Language Processingo Machine learning methodso Deep learning methods.
- TensorflowIntroduction to TensorFlowInstallation & SetupTensorFlow Basic SyntaxTensorFlow GraphsVariables and Placeholders
- TensorFlow Graph & TF DataBuilding the GRAPHRunning a sessionData augmentationTensor flow optimization
- KerasWhat is kerasBuilding models with KerasUnderstanding various features of Keras
- Artificial Neural NetworksANN IntuitionBuilding an ANNHomework Challenge – Should we say goodbye to that customer?Evaluating, Improving and Tuning the ANNHomework Challenge – Put me one step down on the podium
- Overview of Natural Language Processingo Machine learning methodso Deep learning methods.
- Introduction to TensorFlow
- Installation & Setup
- TensorFlow Basic Syntax
- TensorFlow Graphs
- Variables and Placeholders
- Building the GRAPH
- Running a session
- Data augmentation
- Tensor flow optimization
- What is keras
- Building models with Keras
- Understanding various features of Keras
- ANN Intuition
- Building an ANN
- Homework Challenge – Should we say goodbye to that customer?
- Evaluating, Improving and Tuning the ANN
- Homework Challenge – Put me one step down on the podium
- Convolutional Neural NetworksCNN IntuitionHomework – What’s that pet ?Evaluating, Improving and Tuning the CNN
- Recurrent Neural NetworksRNN IntuitionBuilding an RNNEvaluating, Improving and Tuning the RNN
- CNN Intuition
- Homework – What’s that pet ?
- Evaluating, Improving and Tuning the CNN
- RNN Intuition
- Building an RNN
- Evaluating, Improving and Tuning the RNN
- Introduction to GANDeriving the MiniMax Loss FunctionAnalysis of Least Squares Loss FunctionGAN Architecture UsedCoding the Generator and Discriminator ArchitectureFunction and GAN EstimatorModifying GAN Framework to Change Dataset and Loss Function
- The generative revolution: coming homeThe present and future of AI is generativeApplications of generative AI
- Deriving the MiniMax Loss Function
- Analysis of Least Squares Loss Function
- GAN Architecture Used
- Coding the Generator and Discriminator Architecture
- Function and GAN Estimator
- Modifying GAN Framework to Change Dataset and Loss Function
- The present and future of AI is generative
- Applications of generative AI
- Benefits and possibilities of Generative AIUnderstanding the battle between generator and discriminatorUnderstanding Cross Entropy in depthUnderstanding the equation to calculate the discriminator lossUnderstanding the equation to calculate the generator loss
- GAN’s – CodingCoding: importing libraries and declaring a visualization functionCoding: hyperparameters and the DataLoaderCoding: the generator classCoding: the discriminator class
- Coding an advanced generative architectureChallenges and issues of the basic GANThe Wasserstein LossThe Gradient PenaltyCoding: setting up libraries and parametersCoding: Login and setup of the Wandb stats libraryCoding: Beginning the generatorCoding: Understanding convolutions
- Generating images from text by combining two advanced architecturesMultimodal generation, an incredible adventureCoding: importing the librariesCoding: helper functions and hyperparametersCoding: Setting up the CLIP modelCoding: Setting up the Generative transformer modelCoding: Setting up the latent space parameters to be optimized
- Understanding the battle between generator and discriminator
- Understanding Cross Entropy in depth
- Understanding the equation to calculate the discriminator loss
- Understanding the equation to calculate the generator loss
- Coding: importing libraries and declaring a visualization function
- Coding: hyperparameters and the DataLoader
- Coding: the generator class
- Coding: the discriminator class
- Challenges and issues of the basic GAN
- The Wasserstein Loss
- The Gradient Penalty
- Coding: setting up libraries and parameters
- Coding: Login and setup of the Wandb stats library
- Coding: Beginning the generator
- Coding: Understanding convolutions
- Multimodal generation, an incredible adventure
- Coding: importing the libraries
- Coding: helper functions and hyperparameters
- Coding: Setting up the CLIP model
- Coding: Setting up the Generative transformer model
- Coding: Setting up the latent space parameters to be optimized
- Sentiment AnalysisUnderstanding sentiment AnalysisSentiment analysis Hands on using twitter Data
- Understanding sentiment Analysis
- Sentiment analysis Hands on using twitter Data
- Introduction to Artificial Intelligence & Machine LearningIntroduction to Machine LearningMachine Learning tools and Techniques
- Introduction to ChatbotsUnderstanding the transactional chatbotsUnderstanding the informational chatbots
- NLTK Introduction & InstallationSentence Splitter & TokenizationStemming & LemmatizationStop work removalsentiParts of Speech (POS) taggingChunking
- Google Dialog flowThe Big PictureIntroducing DialogFlowThe Big PictureSetting Up DialogflowBuilding Blocks of Interaction ModelsCreating Your First AgentExploring Agent SettingsDefault IntentsSmalltalkCustom IntentsSystem Entities and Developer EntitiesDefining Developer EntitiesUser Expressions for Intents
- Configuration of DialogflowConfiguring and Testing the Book Cars IntentConfiguring and Testing the Book Rooms IntentLinear and Non-linear DialogsSection OverviewContextsFollow up IntentsLinear DialogsNon-linear DialogsNon-linear Dialogs continuedFulfilment, Deployment and 3rd Party Integration
- Setting up AgentSection OutlineCheck Weather IntentBasic Setup of Webhook CodeExtracting Parameter Values and Structuring ResponsesCalling the Open Weather Map APIRetrieving Weather Info from Open Weather Map
- BuildingBuilding: Intents and EntitiesBuilding: DialogDeployment and testing
- Advanced chatbot improvementContext VariablesImprovement Tab
- Nuance – AIIntentsEntitiesDialogMix Dash boardAdding IntentsAdding EntitiesLinking EntitiesTrain & TryDialog GreetingIntent MapperIntent ComponentsRecovery MessagesNode TypesGrammarsVariables
- Introduction to Machine Learning
- Machine Learning tools and Techniques
- Understanding the transactional chatbots
- Understanding the informational chatbots
- Sentence Splitter & Tokenization
- Stemming & Lemmatization
- Stop work removal
- sentiParts of Speech (POS) tagging
- Chunking
- The Big Picture
- Introducing DialogFlow
- The Big Picture
- Setting Up Dialogflow
- Building Blocks of Interaction Models
- Creating Your First Agent
- Exploring Agent Settings
- Default Intents
- Smalltalk
- Custom Intents
- System Entities and Developer Entities
- Defining Developer Entities
- User Expressions for Intents
- Configuring and Testing the Book Cars Intent
- Configuring and Testing the Book Rooms Intent
- Linear and Non-linear Dialogs
- Section Overview
- Contexts
- Follow up Intents
- Linear Dialogs
- Non-linear Dialogs
- Non-linear Dialogs continued
- Fulfilment, Deployment and 3rd Party Integration
- Section Outline
- Check Weather Intent
- Basic Setup of Webhook Code
- Extracting Parameter Values and Structuring Responses
- Calling the Open Weather Map API
- Retrieving Weather Info from Open Weather Map
- Building: Intents and Entities
- Building: Dialog
- Deployment and testing
- Context Variables
- Improvement Tab
- Intents
- Entities
- Dialog
- Mix Dash board
- Adding Intents
- Adding Entities
- Linking Entities
- Train & Try
- Dialog Greeting
- Intent Mapper
- Intent Components
- Recovery Messages
- Node Types
- Grammars
- Variables
- Couse conclusion
- Reference books, videos and blogs
- Next steps
- Final Q&A
- Final assessment (optional)
