К содержимому
learnspaceYOUR NEXT CHAPTER
ПРОСТРАНСТВО ОБУЧЕНИЯ
ГлавнаяКаталог курсовМоё обучениеCoursera

Знания без границ

Учитесь у лучших университетов и компаний мира.

Открыть Coursera
Интеграция
Пространство университета
Моё пространствоСтраница курса
↵
ЯЛичный кабинетСтудент
© 2026 LearnSpaceКаждый день — возможность узнать больше.Помощь
Generative AI Part 2 · LearnSpace
Назад в каталог
courseraАнализ данных

Generative AI Part 2

Курс от Northeastern University
Средний≈ 21.7 чАнглийский
О курсеНавыкиПрограммаПреподаватели

О курсе

Introduces the theoretical foundations and advanced concepts of neural networks, generative models, transformers, and large language models. Students will explore how these AI systems create new data, process information, and learn through feedback, while analyzing their applications across various fields. The course emphasizes key principles in model building, optimization, and real-world generative AI use cases.

Навыки, которые вы освоите

Generative Model ArchitecturesGenerative Adversarial Networks (GANs)Probability DistributionGenerative AIModel EvaluationLarge Language ModelingEmbeddingsAutoencodersNatural Language ProcessingSampling (Statistics)Model OptimizationStatistical MethodsModel Training

Программа курса

7 модулей · 255 учебных материалов

01Transformer-Based Language Models and Pre-Training28 материалов

Getting Started

Course IntroductionЧтениеMeet Your FacultyЧтениеSyllabus - Generative AI Part 2ЧтениеRecommended Prior KnowledgeЧтение

Учитесь у экспертов

Ramin Mohammadi

Преподаватель курса

Generative AI Part 2
В каталоге вашей программы

Инвестируйте в себя

Новые знания — в удобное для вас время.

Начать на Coursera

Обучение откроется на Coursera
в новой вкладке

Обучение на Coursera

≈ 21.7 ч

7 модулей

Язык: Английский

Часть программы вашего университета
Academic IntegrityЧтение

Lesson 1: Transformers and Pretraining

Module OverviewЧтениеPre-TrainingВидеоTransformers for NLPЧтениеPre-Trained Word EmbeddingsЧтениеPre-Training Whole ModelsЧтениеModule 8- Assess Your Learning 1Задание

Lesson 2: Reconstruction and Pretraining

Reconstructing the InputЧтениеPre-Training Through Language ModelingЧтение

Lesson 3: BERT

BERT & TuningВидеоFine-Tuning BERTЧтениеFine-Tuning In-DepthЧтениеPre-Training DecodersЧтениеModule 8- Assess Your Learning 2Задание

Lesson 4: Pre-Training Strategies for NLP

GPT and RAGВидеоGenerative Pretrained TransformerЧтениеPrompt EngineeringВидеоScaling LawsЧтениеScaling EfficiencyЧтениеPre-Training Encoder/DecodersЧтениеSpan CorruptionЧтениеScaling Law & Transfer LearningВидеоModule 8- Assess Your Learning 3Задание

Module Wrap-Up

Module Wrap-UpЧтение
02Variational Autoencoders and Deep Latent Variable Models23 материалов

Lesson 1: Deep Latent Variable Models

Module OverviewЧтениеProbability, Density, Mass FunctionВидеоDeep Latent Variable ModelsЧтениеMixture of GaussiansЧтениеVAE IntroductionВидеоVariational Autoencoder (VAE)ЧтениеModule 9- Assess Your Learning 1Задание

Lesson 2: Sampling and Optimization in Probabilistic Models

Sampling & Monte Carlo OptimizationВидеоDiscrete and Continuous SpaceЧтениеNaïve Monte CarloЧтениеImportance SamplingЧтениеEvidence Lower Bound (ELBO) Part 1ВидеоELBO Deep DiveЧтение

Lesson 3: Variational Autoencoders (VAEs)

Return to Variational AutoencodersЧтениеVariational Autoencoders in DepthВидеоVariational ApproximationЧтениеVariational Autoencoder ContinuedЧтениеReparameterization TrickЧтениеAmortization in VAEЧтение

Module Wrap-Up

Module Wrap-UpЧтение
03Normalizing Flows37 материалов

Lesson 1: Normalizing Flow Part 1

Module OverviewЧтениеNormalizing Flow Part 1ВидеоIntroduction to Normalizing FlowЧтение1D IntroductionВидео1D Normalizing FlowЧтениеMeasuring ProbabilityЧтениеChange of Variables ExplainedВидеоChange of Variables FormulaЧтениеModule 10- Assess Your Learning 1Задание

Lesson 2: Geometry Refresher

Geometry InfoЧтениеDeterminants and VolumesЧтениеModule 10- Assess Your Learning 2Задание

Lesson 3: Normalizing Flow Part 2

Introduction to Forward and Inverse MappingВидеоForward and Inverse MappingЧтениеLearningЧтениеGeneral Use CaseЧтение2D Example: Deep Neural NetworkВидеоForward Mapping With a Deep Neural NetworkЧтение

Lesson 4: Requirements for Flow Models and Network Layers

Flow Model RequirementsЧтениеTriangular JacobianЧтение

Lesson 5: Invertible Network Layers

Linear FlowsВидеоOverview and Linear FlowsЧтениеElementwise & Other Types of FlowsВидеоElementwise FlowsЧтениеCoupling FlowsЧтение

Lesson 6: Nonlinear Independent Components Estimation (NICE)

Introduction to NICEЧтениеReal-NVP: Non-Volume Preserving Extension of NICEЧтениеInterpolation in Latent Space With Real-NVPЧтение

Lesson 7: Normalizing Flow Part 3

Autoregressive FlowsЧтениеContinuous Autoregressive Models as Flow ModelsЧтениеInverse Autoregressive FlowsЧтениеSummary of Normalizing FlowsВидеоApplications of Normalizing FlowsЧтениеModule 10- Assess Your Learning 4Задание

Module Wrap-Up

Module Wrap-UpЧтение
04Generative Adversarial Networks34 материалов

Lesson 1: Introduction to GANs

Module OverviewЧтениеRefresherЧтениеTowards Likelihood-Free LearningЧтениеLikelihood-Free LearningЧтениеGenerative Modeling and Two-Sample TestsЧтениеDiscrimination as a SignalЧтениеModule 11- Assess Your Learning 1Задание

Lesson 2: GANs

OverviewЧтениеGenerator vs. Discriminator DiagramЧтениеTraining Objective for DiscriminatorЧтениеInterpretationЧтениеLoss FunctionsЧтениеTraining AlgorithmЧтениеKey ObservationsЧтение

Lesson 3: Challenges in Training GANs

IntroductionЧтениеOptimization Challenges in GANsЧтениеMode CollapseЧтениеBeyond KL and Jenson-Shannon DivergenceЧтениеModule 11- Assess Your Learning 3Задание

Lesson 4: f-GAN

f-divergencesЧтениеWhat is Lower Semicontinuity?ЧтениеExamples of f-divergences and TrainingЧтениеToward Variational Divergence MinimizationЧтениеf-GAN Variational Divergence MinimizationЧтениеModule 11- Assess Your Learning 4Задание

Lesson 5: WGAN

Wasserstein (Earth Mover) DistanceЧтениеDiscrete DistributionsЧтениеWasserstein Distance for Continuous DistributionsЧтениеInferring Latent Representations in GANsЧтениеModule 11- Assess Your Learning 5Задание

Module Wrap-Up

Module Wrap-UpЧтение
05Energy-Based Models and Score-Based Models39 материалов

Lesson 1: Energy-Based Models

Module OverviewЧтениеBackgroundЧтениеParameterizing Probability Distribution: DefinitionЧтениеParameterizing Probability Distributions: SolutionЧтениеEnergy-Based ModelsЧтениеPros and Cons of Energy Based ModelsЧтениеModule 12- Assess Your Learning 1Задание

Lesson 2: Applications of Energy-Based Models

ExamplesЧтениеExamples ContinuedЧтениеComputing the Normalization ConstantЧтение

Lesson 3: Training Intuition and Score Function

IntroductionЧтениеContrastive Divergence AlgorithmЧтениеSampling in Energy-Based ModelsЧтениеScore FunctionЧтениеScore MatchingЧтениеModule 12- Assess Your Learning 2Задание

Lesson 4: Score-Based Models

Score-Based Models IntroductionЧтениеBackgroundЧтение

Lesson 5: Score Matching

Denoising Score Matching Part 1: IntroductionЧтениеDenoising Score Matching Part 2: Defining the ObjectiveЧтениеDenoising Score Matching Part 3: Gradient ExpansionЧтениеGradient DerivationЧтениеIntuitionЧтениеWhy Denoising Works in Score MatchingЧтение

Lesson 6: Score-Based Models Continued

Data Generation with Score-Based ModelsЧтениеPitfalls With Score-Based ModelsЧтениеSolution to PitfallsЧтениеModule 12- Assess Your Learning 4Задание

Lesson 7: Noise Conditional Score-Based Models

Introduction to NCSBMЧтениеAnnealed Langevin DynamicsЧтениеTraining Noise Conditional Score NetworksЧтениеChoosing Noise ScalesЧтениеChoosing the Weighting FunctionЧтениеModule 12- Assess Your Learning 5Задание

Module Wrap-Up

Module Wrap-UpЧтение
06Diffusion Models47 материалов

Lesson 1: Model Families

Module OverviewЧтениеIntroductionЧтениеModel Families ContinuedЧтение

Lesson 2: Diffusion Models

DefinitionЧтениеDiffusion ProcessЧтениеDistribution of Each TermЧтениеDiffusion KernelЧтениеMarginal DistributionsЧтениеConditional DistributionЧтениеBackward Diffusion Process-DecoderЧтениеModule 13- Assess Your Learning 1Задание

Lesson 3: Training

Encoder / DecoderЧтениеLoss FunctionЧтениеGaussian Distribution and Its MeanЧтение

Lesson 4: Diffusion Models as Score-Based Models

Diffusion Models as Score-Based ModelsЧтениеDecoder ParameterizationЧтениеLoss FunctionЧтениеTraining and InferenceЧтениеModule 13- Assess Your Learning 2Задание

Lesson 5: Architectures for the Denoiser

U-Net ArchitectureЧтениеInfinite Noise Levels Score-Based ModelingЧтениеPerturbing Data With Stochastic ProcessesЧтениеStochastic Differential Equations (SDEs)ЧтениеTypes of SDEs and Noise EvolutionЧтениеReverse Stochastic ProcessЧтение

Lesson 6: Score-Based Generative Modeling via SDEs

Time-Dependent Score-Based ModelЧтениеTraining ObjectiveЧтениеReverse-Time SDEЧтениеEuler-Maruyama Approximation and SummaryЧтениеModule 13- Assess Your Learning 4Задание

Lesson 7: Reverse Process: Sampling From the Reverse SDE

Where Does the Time Step Come From?ЧтениеStep-by-Step Sampling: Euler-Maruyama MethodЧтение

Lesson 8: Predictor-Corrector Sampling Methods

Predictor-Corrector Sampling MethodsЧтениеCombined Predictor-Corrector SamplingЧтение

Lesson 9: Probability Flow Ordinary Differential Equations (ODEs)

Probability Flow ODEЧтениеLikelihood ComputationЧтениеPractical Considerations and ConclusionЧтениеModule 13- Assess Your Learning 5Задание

Lesson 10: Latent Diffusion Models

Intro to Latent Diffusion ModelsЧтениеConditional GenerationЧтениеImproving Image QualityЧтениеControl the Generation ProcessЧтениеExamplesЧтениеModule 13- Assess Your Learning 6Задание

Module Wrap-Up

Module Wrap-UpЧтение
07Annealed Importance Sampling and Model Evaluation47 материалов

Lesson 1: Annealed Importance Sampling

Module OverviewЧтениеOverview of AISЧтениеExample: AIS With a Gaussian DistributionЧтениеIntermediate Step (t = 1)ЧтениеIntermediate Step (t = 2)ЧтениеFinal Steps (t = 8)ЧтениеModule 14- Assess Your Learning 1Задание

Lesson 2: AIS Example: Uniform Distributions (t = 8)

SetupЧтениеStep-By Step Solution for t = 1ЧтениеApplications and TakeawaysЧтениеModule 14- Assess Your Learning 2Задание

Lesson 3: Normalization of Probability Distributions

Normalization of Probability Density FunctionsЧтениеExamples of Normalizing ConstantsЧтениеSteps to Normalize p(z)ЧтениеWrapping Up Probability DistributionsЧтениеModule 14- Assess Your Learning 3Задание

Lesson 4: Evaluation Metrics Overview

Model Family RecapЧтениеModel Families ContinuedЧтениеDistances of Probability DistributionsЧтениеEvaluating Generative ModelsЧтениеWhat is the Task That You Care About?ЧтениеEvaluationЧтениеModule 14- Assess Your Learning 4

Lesson 5: Sample Quality Evaluation

Kernel Density Estimation (KDE)ЧтениеLatent Variables & Sample QualityЧтениеHYPE: Human Eye Perceptual EvaluationЧтениеInception ScoresЧтениеSharpnessЧтениеDiversityЧтениеInception Scores Finalized

Lesson 6: Distance-Based Evaluation Methods

Frechet Inception Distance (FID)ЧтениеKernel Inception Distance (KID)ЧтениеFID vs. KIDЧтениеModule 14- Assess Your Learning 6Задание

Lesson 7: Specialized Application

Evaluating Sample Quality for Text-to-Image ModelsЧтениеEvaluating Latent RepresentationsЧтениеClusteringЧтениеLossy Compression or ReconstructionЧтениеDistentanglementЧтениеBeta-VAEЧтениеSolving Tasks Through Prompting

Module Wrap-Up

Module Wrap-UpЧтениеCongratulations!Чтение
Evidence Lower Bound (ELBO) Part 2Видео
Module 9- Assess Your Learning 2Задание
Module 9- Assess Your Learning 3Задание
Training Objective for Normalizing FlowsЧтение
Module 10- Assess Your Learning 3Задание
Alternating Optimization in GANsЧтение
ExamplesЧтение
Module 11- Assess Your Learning 2Задание
Comparison Between NSM and DSMЧтение
Tweedie FormulaЧтение
Overview of Sliced Score Matching (SSM)Чтение
Module 12- Assess Your Learning 3Задание
Role of the Score FunctionЧтение
Module 13- Assess Your Learning 3Задание
Задание
Чтение
Relationship Between Inception Score and KL DivergenceЧтение
Module 14- Assess Your Learning 5Задание
Чтение
Holistic Evaluation of Language Models (HELM)Чтение
Module 14- Assess Your Learning 7Задание