Ollama - Mastering Generative AI and Beyond

Define generative AI and review its market examples and usage, then compare machine learning with artificial intelligence and explore key AI models such as GPT, Dall-E, and Codex.

Establish a basic understanding of generative AI, its content generation across images, text, music, and code. Explore tools like Dall-E, ChatGPT, Copilot, Tabnine, and code-generation models.

Explore the basic terms of generative AI, including artificial intelligence, machine learning, supervised and unsupervised learning, reinforcement learning, algorithms, and models, with real-world examples.

Use cake-baking analogy to explain the machine learning model: data collection, apply an algorithm, train, refine, finalize, and deploy a model for text generation, image generation, or video generation.

Understand how tokens and embeddings enable AI models to process prompts, with tokenization (word, subword, and character) converting text into embeddings fed to neural networks.

Master prompt engineering for DevOps and DevSecOps, using zero-shot, one-shot, few-shot, and chain-of-thought strategies. Explore input design, output guidance, model interaction, and practical prompts for Terraform and Kubernetes.

Explore how ChatGPT works by examining unsupervised, supervised, and reinforcement learning from human feedback, the transformer architecture, and the generative pre-trained transformer model powering GPT.

Explore how Dall-E generates images from text using a transformer neural network and the GPT-3 model, with unsupervised learning and supervised fine-tuning on internet text-image pairs.

Explore the models, algorithms, and training processes behind GitHub Copilot, including the Codex model, transformer architecture, unsupervised and supervised learning, and public code fueling code generation.

Learn how Ollama runs ai models on cpu without strong gpus using cpu-optimized models, quantization, and pruning, enabling text generation, live data processing, and chatbots.

Install Ollama on Windows to run local language models, assess model sizes and required RAM, download and pull models, and run them offline from the command prompt.

Perform a hands-on install of Ulama on a Linux AWS EC2 instance using a curl command. Run the session in CPU-only mode with the Ulama API at localhost:1143.

Build a Spring Boot REST API to send prompts to the Ulama Mistral model using a chat client, exposing a GET endpoint to interact with Ulama and fetch live data.

Build and run a Spring Boot rest api with the Gradle wrapper, call the mistral ulama model via a llama rest controller, and test with httpie for the formatted response.

Explore a case study that fetches and ingests live weather data into Ulama models via a Spring Boot weather query app, using weather api.com and a GitHub repo walkthrough.

Clone a spring boot weather API project to fetch live data and feed it to Ulama models, via a city weather endpoint and a weather data service.

Follow a code walkthrough that fetches live weather data via an API, ingests it into an AI model (Mistral via Llama), and validates results across sources.

Install Docker Desktop on Windows, enable Virtual Machine Platform and WSL 2, log in to Docker Hub, and prepare to run OpenWebUI with Ollama.

Learn to run Open Web UI, a self-hosted offline interface for Ulama on your local system via Docker. Upload code files, switch between models, and view streaming results without internet.