Explore machine learning fundamentals in generative AI using OpenAI and Azure OpenAI, guiding learners from beginner to pro-level techniques.

Explore a real-world Netflix machine learning case, where machine learning informs catalog curation, content production, encoding, streaming, and acquisitions across 190 countries for 120 million members.

Explore core machine learning terms like algorithms, models, training, labels, and features, and see how data and statistics drive learning through examples such as Gmail filters and house price prediction.

Discover how machine learning imitates human learning by training algorithms on data to identify patterns, build models, and predict outcomes with supervised, unsupervised, and reinforcement learning, stressing data quality.

Explore supervised, unsupervised, and reinforcement learning, including classification, regression, and clustering, and understand labeled versus unlabeled data and the feedback-driven reward-punishment loop.

Explore supervised machine learning, where a labeled data set acts as the supervisor to train a model mapping input features to outputs, enabling classification, regression, and awareness of overfitting.

Classify data by assigning category labels to new observations using labeled training data. Learn binary and multiclass classification with feature selection, and consider overfitting and underfitting in spam detection.

Explore regression in supervised learning, linking dependent and independent variables to predict continuous outputs like house prices and stock prices; compare linear and multiple regression, and address overfitting and underfitting.

Explore unsupervised machine learning, analyzing unlabeled data to identify hidden patterns through clustering and association, with applications in market segmentation and recommender systems, plus anomaly detection.

Reinforcement learning trains an AI agent to maximize cumulative rewards by taking actions in an environment, observing state, and receiving rewards or penalties, balancing exploration and exploitation.

Discover how Jupyter notebooks offer an interactive coding environment with code cells and live execution for Python and other languages, plus rich text, data visualization, and easy sharing.

Install and explore Anaconda, a beginner-friendly open source data science platform powered by conda, and use the Navigator GUI to launch a Jupyter notebook and write Python code.

Explore iris data with sepal length, sepal width, petal length, and petal width across setosa, versicolor, virginica. Load with pandas, view shape, and inspect records by id and species.

Learn to build a logistic regression model on the iris data by splitting into X and y, training on the features with their labels, and generating predictions for new samples.

Explore generative AI fundamentals through a machine learning quiz, guiding beginners to pro skills with OpenAI and Azure OpenAI.

Explore deep learning within generative AI using OpenAI and Azure OpenAI to elevate your skills from beginner to pro.

Deep learning mimics the brain with multilayer artificial neural networks to process unstructured data, powered by GPUs and tools like TensorFlow, PyTorch, and Keras, enabling autonomous vehicles and NLP.

Explore how artificial neurons form deep neural networks with input, hidden, and output layers, using weights, biases, activation functions, and forward and backward propagation.

Discover deep learning models, including convolutional neural networks for image and video processing, recurrent neural networks and LSTMs for sequences, GANs for fakes, and transformer models powering GPT-3 and ChatGPT.

Discover how the transformer model powers generative AI and natural language processing with encoder-decoder architecture, self-attention, and parallel processing for fast text generation, translation, and question answering.

Explore how generative adversarial networks create deepfake videos, using a generator and discriminator to craft convincing images, audio, and video hoaxes, including face swaps.

Build a deep learning model on the Pima Indians diabetes dataset from Kaggle using Keras and TensorFlow. Train in a Jupyter notebook, evaluate accuracy, and discuss data quality and limitations.

test your deep learning knowledge in generative AI with a focused quiz aligned to OpenAI and Azure OpenAI concepts, helping you gauge readiness from beginner to pro.

Explore Generative AI from beginner to pro with OpenAI and Azure OpenAI, as introduced in the course on Generative AI.

Generative ai, a subset of deep learning powered by neural networks, creates text, images, music, and code from data-driven prompts, learning from high-quality data to generate outputs efficiently.

Compare predictive ai and generative ai: predictive ai analyzes historical data to forecast future outcomes, while generative ai creates new content such as text, images, music, and code.

Learn how large language models, or lm, process, understand, and generate language, with billions of parameters, transformer architecture, continuous learning, and customization options for chatbots, content, and translation.

Learn how embeddings convert a word's features into vectors of real numbers, forming the backbone of generative AI and enabling context, similarity, and dimensionality reduction to 2D.

Explore how vector databases store embeddings—arrays of numbers representing data in high-dimensional space—and enable fast similarity searches using methods like Euclidean distance and cosine similarity.

Explore how embeddings convert real world data into numerical vectors, store them in a vector database for efficient similarity search, and power applications like Netflix recommendations.

Explore retriever augmented generation (rag): it retrieves internal data with a vector database using embeddings and semantic search, then generates contextually accurate responses with a language model.

Explore Lange chain, a modular framework that integrates external data sources with large language models, using embeddings and a vector database for retrieval augmented generation.

Langchain serves as an end-to-end framework for retrieval-augmented generation, integrating document retrieval, embeddings, vector databases, and enhanced prompts to produce LM-driven responses.

Learn prompt engineering and fine tuning to optimize generative AI outputs by crafting effective inputs, tailoring prompts, and fine-tuning models with task-specific data for better accuracy and adaptability.

Test your understanding of generative ai concepts with an OpenAI and Azure OpenAI quiz, aligned with beginner to pro learning.

Develop ai infrastructure fundamentals for generative ai with OpenAI and Azure OpenAI. Identify key components and integration points to support modern ai workloads.

Demystify the graphics processing unit, explaining how GPUs enable parallel processing for AI training, graphics rendering, and virtual reality, with integrated and dedicated options, and why Nvidia leads the field.

Watch a GPU vs CPU demonstration that contrasts sequential CPU actions with thousands of GPU cores running in parallel, using a Nvidia paintball Mona Lisa analogy.

Explore rdma cluster networks, enabling remote direct memory access to bypass the operating system and cpu bottlenecks, delivering low latency and high throughput for hpc, cloud computing, and big data.

Test your knowledge of generative AI using OpenAI and Azure OpenAI in this quiz, reinforcing key concepts from the course.

Explore OpenAI and Azure OpenAI APIs, harness ChatGPT capabilities, and build practical generative AI applications from beginner to pro.

OpenAI, a 2015 ai research lab, aims to develop friendly ai for humanity and powers ChatGPT via GPT iterations and Azure partnerships, focusing on ethics, safety, and token pricing.

Explore how ChatGPT, built on generative pre-trained transformer technology, enables an advanced conversational AI that generates text, answers questions, writes code, and can be integrated via APIs for custom applications.

Explore how to access ChatGPT, compare free 3.5 and GPT-4, and see real-time generative prompts—from stories and poems to Python code—within OpenAI and Azure OpenAI.

Demonstrate ChatGPT's rapid growth by showing it reached 100 million users in about two months, faster than Google Translate, Uber, or Telegram. Show OpenAI’s generative AI as a game changer.

Explore the heart of OpenAI: diverse models tailored to specific use cases, including GPT-4 turbo, GPT-4, and GPT-3.5, with capabilities like multimodal inputs, fine-tuning, and image generation with DALL-E.

Compare GPT-3 and GPT-4 by highlighting multimodal input, improved creativity, context handling, and accuracy, with GPT-4 surpassing GPT-3 in bar exam performance and reliability.

Learn about GPT-4o, OpenAI's omni multimodal model that accepts text, audio, image, and video as input and outputs text, audio, or image, with optimized, faster, and cheaper performance.

Compare GPT-4 and GPT-4o in a demo, showing GPT-4o delivers a one-page story twice as fast and at about 50% lower cost when given the same prompt.

Explore GPT-4 mini, a compact, edge-deployable model that rivals small to mid-size competitors with faster responses and lower costs, enabling local deployment and resource efficiency.

Explore how tokens form the backbone of OpenAI models, how tokenization breaks text into tokens, and how token counts affect processing and pricing.

Explore OpenAI pricing and token-based costs, including input and output charges, model differences, API vs. chat pricing, plus budgeting and billing tips.

Learn prerequisites for making OpenAI API calls, including setting up billing, generating project API keys, creating an openai.env file with your API key, and authenticating requests.

Install the OpenAI library, load your OpenAI API key from an environment file, and make a chat completions API call in a Jupyter notebook to test prompts.

See how embeddings convert words into vectors and reflect context, using cat, kitten, and dog, then reduce dimensions with a small OpenAI text-embedding model.

learn to generate images with the OpenAI DALL·E 3 API by loading API keys, calling client.images.generate, and crafting prompts to produce unique 1024 by 1024 images.

Experience a quick demo of converting speech to text with OpenAI's whisper model, processing mp3 and mp4 audio files into accurate transcriptions and grammar-aware text.

OpenAI O1 models are new large language models trained with reinforcement learning to perform complex reasoning, featuring internal chain-of-thought and reasoning tokens, plus a faster, cheaper O1 mini.

Compare GPT-4o and OpenAI O1 using the OpenAI playground to test speed, accuracy, and reasoning style, and see how O1 trades speed for crisp, concise results.

Take a quick quiz to assess your understanding of generative AI concepts in the beginner to pro course on OpenAI and Azure OpenAI.

Explore Azure OpenAI to harness generative AI within the OpenAI ecosystem, bridging OpenAI and Azure OpenAI capabilities for practical projects.

Explore Azure OpenAI intro, covering what the service is, Microsoft history, region-based model availability, access limits, setup, studio, and completions and image creation playgrounds with web app deployment and quiz.

Discover how Azure OpenAI combines Microsoft Azure and OpenAI to bring GPT-3.5, Codex, and DALL·E to enterprises, with security, compliance, and easy customization through a graphical user interface and API.

Trace the Azure OpenAI history—from the 2019 Microsoft OpenAI partnership and Musk's withdrawal to a 2021 enterprise launch, featuring Codex, DALL-E, and GPT-3.

Explore how Azure OpenAI regions determine model availability, and learn to check region-specific support for GPT-4, GPT-3.5, embeddings, Dall-E, Whisper, and TTS.

Navigate quotas and limits for Azure OpenAI by region, including tokens per minute, six requests per minute per 1000 tokens, up to 30 resources and five fine-tuned deployments.

Explore Azure OpenAI pricing by understanding context, input, and output tokens billed per 1000 tokens, compare regions and currencies, and use the pricing calculator to estimate costs.

Demonstrates setting up the Azure OpenAI service via portal.azure.com, requesting access, creating a resource group, naming the endpoint, choosing pricing, and deploying embeddings and chat via Azure OpenAI Studio.

Learn how Azure OpenAI Studio provides a platform to build, train, and manage models with playgrounds for assistant, chat, completions, and Dall-E, plus bring-your-own data, fine-tuning, and security and compliance.

Navigate the Azure OpenAI Studio to explore chat and completions playgrounds, create deployments for base models, manage data files and quotas, and configure content filtering for security.

Create an Azure OpenAI chat deployment, selecting GPT-3.5 turbo 16k and enabling auto update. Set the deployment name, quotas, and dynamic quota to ensure reliable chats in the playground.

Explore the charts playground in three parts—setup, chat, and configuration—covering prompts, templates, rag data, system messages, and context management with code and JSON views.

The demo shows moving from dev and uat to production by deploying an OpenAI model to a web app via Azure deployment utility, with Azure AD authentication and endpoint launch.

Explore generating images with the dall-e playground in azure openai, learn about deployments, and preview and download prompts that produce a black dog and a white cat.

Explore the completions playground to craft prompts for Azure OpenAI and OpenAI, practice summarization and classification, and learn deployment steps and model compatibility to avoid errors.

Explore the completions playground to deploy your model and select deployments. Run real-world use cases like summarizing financial reports, generating quizzes, creating chatbots, and explaining SQL queries.

Take this quiz to reinforce core generative ai concepts using OpenAI and Azure OpenAI, solidifying beginner to pro skills with hands-on practice.

Explore Azure AI Foundry within the Generative AI course, leveraging OpenAI and Azure OpenAI to understand platform capabilities.

Azure AI Foundry hosts OpenAI and other models (Meta llama, deep seek) in a single model catalog, enabling safe, customizable, governed AI development on a trusted platform.

Azure AI Foundry architecture exposes a model catalog and core AI services, including the Azure OpenAI service, with Copilot Studio and the SDK, plus observability for evaluations, customization, and governance.

Explore hubs and projects in Azure AI Foundry, enabling governance, asset reuse, and collaborative sharing of models, datasets, and prompts across teams.

Navigate the Azure AI Foundry portal to access Azure OpenAI models, switch to the model catalog, and manage deployments, quotas, and playgrounds.

Compare Azure OpenAI service with and without a project to show direct model access versus an end-to-end gen AI platform with cross-provider models, enterprise features, and a unified SDK.

Create an Azure Foundry project and hub through AI dot Azure.com, connect to an Azure OpenAI service, review resources in portal.azure.com, and launch Studio for models and compute.

Compare model benchmarks in the foundry across Azure OpenAI, OpenAI, and other third-party models using quality, cost, latency, and throughput metrics to guide model selection.

Deploy a model in Azure Foundry from the model catalog, configure a global standard REST deployment, obtain the endpoint and API key, and run prompts in the playground.

Explore how to make API calls with Azure OpenAI to harness generative AI capabilities. Gain hands-on API integration skills for OpenAI and Azure OpenAI.

Compare OpenAI and Azure OpenAI API calls, then guide end-to-end setup—endpoint URL, API key, API version, creating an Azure OpenAI service, and a chart deployment to invoke the model.

Compare OpenAI and Azure OpenAI API calls, highlighting differences in deployments, API version, and endpoint usage; learn to pass API keys via environment variables and reference deployment names for completions.

Create a new Azure OpenAI service instance in east US or reuse an existing one, configure subscription, resource group, and name, then deploy to move from creating to succeeded.

Discover how to obtain the Azure endpoint URL and API keys for your OpenAI service, secure them with environment variables or Azure Key Vault, and use dual keys.

Create and populate an azureopenai.env file to store the azure openai endpoint and api key, enabling secure authentication for api calls in a jupyter notebook workflow.

Identify api_version, azure endpoint url, and api key; use preview or ga release and test upgrades. Implement completions with client dot completions dot create via a deployment, not a model.

Create a gpt-3.5-turbo deployment in Azure OpenAI Studio via the chat playground, naming it for your code. Use default 0301, standard tokens, and content filter, and begin making api calls.

Make your first Azure OpenAI API calls by loading the environment file with your API key and endpoint, installing the OpenAI package, and calling chat or completions in a notebook.

Explore BYOD in the realm of generative AI, leveraging OpenAI and Azure OpenAI to empower beginners toward professional mastery.

Explore retrieval augmented generation with rag using azure ai search to securely query your own private data, learn vector search, and set up BYOD with embeddings and charts deployments.

Explore Rag with Azure AI Search to query private data at scale, and learn about indexing, AI enrichment, inverted indexes, semantic search, and Azure ecosystem integration through practical demos.

Explore how Azure AI search uses embeddings and vector databases to perform approximate nearest neighbor and relevance-ranked similarity searches from multi-modal data sources.

Demonstrate how rag with Azure AI search retrieves ACAS documents on dispute resolution, training, and employee rights using semantic search and a ChatGPT deployment.

Set up the prereqs for rag with azure ai search by creating a storage account, configuring embedding and chat deployments, and provisioning azure ai search service with ingestion and indexing.

Create a chats deployment using GPT-3.5 turbo 16k within your private data space, ensuring data stays local, then configure a rag chat model deployment.

Create an Azure AI search resource in portal.azure.com, select east us and a free tier, and monitor indexes, data sources, quota, latency, and queries per second for retrieval augmented generation.

Explore how to upload documents to a chat deployment with Azure OpenAI, enabling ingestion, pre-processing, and indexing to power accurate search queries using hybrid and vector search.

Query your own data with retrieval augmented generation, using private documents and semantic, vector search to produce reference-backed answers. Explore bullet-point rules for paying final salary and other employer questions.