Follow 10 curated learn from repos courses and learning paths, from beginner fundamentals to advanced practice.
A 100-day structured Python path: basics → web → data analysis
A famous 100-day Python learning path that takes you from absolute fundamentals to real applications: Python basics, OOP, web development (Flask/Django), data analysis with pandas, databases, automation, and practical mini-projects for every phase. One of the most-starred Python roadmaps on GitHub - ideal as your day-by-day plan.
18 lessons: LLMs, prompting, RAG, agents, fine-tuning
Microsoft's flagship practical introduction to building Generative AI applications in 18 lessons. Covers LLM fundamentals, prompt engineering, RAG, AI agents, fine-tuning and shipping GenAI apps - with code examples in Python/TypeScript and open-source lesson repos. The single best starting point before touching any framework.
Code an LLM from zero: tokens → attention → training
Build a GPT-style large language model step by step to truly understand what happens inside an LLM: tokenization, embeddings, attention mechanisms, the transformer architecture, pretraining loops, and fine-tuning. Written by Sebastian Raschka; pairs with his book of the same name. For engineers who want depth, not just API calls.
12-week / 26-lesson classical ML curriculum (Scikit-learn)
A structured 12-week, 26-lesson curriculum covering classical machine learning: regression, clustering, classification, NLP basics, time-series forecasting, and model evaluation using Scikit-learn. The right starting point if you want ML fundamentals before jumping into LLMs.
Practical recipes for shipping with OpenAI models
Official collection of practical examples and guides for building with OpenAI models: embeddings & semantic search, function calling, RAG pipelines, evaluation, fine-tuning recipes, and agent patterns. The place to go when you want to move from learning to actually building production features.
Original latent-diffusion text-to-image implementation
The original Stable Diffusion implementation and research code from CompVis (LMU Munich & Runway). For those interested in generative image models: latent diffusion architecture, conditioning mechanisms, samplers, and how text-to-image generation actually works under the hood. Research-grade - read alongside the paper.
Agentic patterns, tools, frameworks & multi-agent systems
A practical course for understanding and building AI agents: agentic design patterns, tool use, RAG inside agents, agent frameworks (AutoGen, Semantic Kernel, Azure AI Agent Service), planning, multi-agent systems, and running agents in production. Lessons include video walkthroughs and working code samples.
12-week / 24-lesson tour of all of AI: CV, NLP, DL, RL
A structured 12-week, 24-lesson introduction to the full AI landscape: neural networks, computer vision, NLP, deep learning frameworks (TensorFlow/PyTorch), genetic algorithms, reinforcement learning, and classical AI approaches - with quizzes and hands-on labs. Broad grounding before specializing.
Real-time RAG, enterprise search & live-data LLM pipelines
A repository focused on building practical LLM applications with real-time data: RAG pipelines that stay in sync with live sources, enterprise document search, vector search over streaming data, and monitoring-ready architectures. Ideal once you know the basics and want production-shaped LLM systems instead of notebooks.
Promptable image segmentation foundation model + dataset
Meta's foundation model for promptable image segmentation: point/click or box a region and SAM segments any object without task-specific training. Includes model weights, inference code, and the SA-1B dataset. Worth exploring if you're interested in computer vision and multimodal AI systems.
No ads · No tracking · No data collection, ever.
Even one share makes a difference 🧡
