MindSpore online courses: Step into LLM

Cerno is a local-first research platform that leverages agentic AI to break down complex queries into verifiable, multi-step workflows. Switch seamlessly between cloud LLMs and self-hosted models, track every reasoning step, and optimize cost and tokens—all while keeping your data on your machine.

TreeThinkerAgent is a lightweight orchestration layer that turns any LLM into an autonomous multi-step reasoning agent. It supports multi-step planning, tool execution, and final synthesis while exposing the entire reasoning process as a tree you can explore.

Step into the world of LLMs with this practical guide that takes you from the fundamentals to deploying advanced applications using LLMOps best practices.

🚀 Next.js Contextify - AI-Ready Context Generator for Next.js Projects Transform your Next.js codebase into optimized context files for AI/LLM analysis. Features step-by-step UI, 12 professional prompt templates, intelligent file prioritization, and smart directory selection.

This project showcases the power of LangGraph, a library for building stateful, multi-step applications using LLMs. By combining LangGraph with tools like RAG, web search capabilities, and specialized knowledge bases, AI assistant can navigate complex information landscape and synthesize findings into coherent, well-structured reports.

a modular Obsidian plugin (or equivalent JS-based extension) that converts raw clippings into clean, structured atomic notes through an LLM-assisted extraction step.

TreeThinkerAgent is a lightweight orchestration layer that turns any LLM into an autonomous multi-step reasoning agent. It supports multi-step planning, tool execution, and final synthesis while exposing the entire reasoning process as a tree you can explore.

🔮 LLM Work Samples – A peek into my adventures with Large Language Models, exploring cutting-edge AI and natural language processing. Step into the future of text! 🚀

A modular, step-by-step implementation of the Transformer architecture (GPT-style) from scratch using PyTorch. This repository breaks down the complexity of LLMs into digestible modules, from BPE tokenization to the final training loop

LLM-powered autonomous barista robot using UR5 arm, ROS2 Humble, MoveIt2, and Ignition Gazebo. Natural language orders decomposed into multi-step robot actions with simulated grasping

This repository is a hands-on, step-by-step journey into Natural Language Processing (NLP) — starting from fundamental text preprocessing techniques all the way to building and fine-tuning Transformer-based models and real-world LLM applications.

A DSA Visualizer that transforms Data Structures & Algorithms into interactive, animated steps using natural language queries. Just type something like “Visualize Bubble Sort” or “Explain Dijkstra visually”, and get a step-by-step animated explanation powered by Groq LLM + Framer Motion.

“Build an AI-powered automated book reader that processes printed and handwritten text using OCR, translates content into selected languages, explains concepts (especially math) step-by-step using LLM, and reads aloud with TTS, with interactive UI and accessibility support.”

📄➡️📊 Convert PDFs into AI-generated presentation decks using a fully serverless architecture. ⚡ AWS Lambda + Step Functions orchestration 🧠 Multimodal LLM (Qwen) for slide generation ☁️ S3 + DynamoDB for scalable processing

Starter pack for building LLM-powered agents and RAG workflows in the browser with the no-code Langflow builder—complete with step-by-step memory, multi-tool, and multi-agent demos. langflow.org datastax.com Seamlessly links to the LangChain ecosystem (LangChain, LangSmith, LangGraph) so you can scale prototypes into production

AI Multi-Agent Blog Writer converts technical specifications into accurate, engaging blogs using four LLM agents — Researcher, Writer, Fact-Checker, and Style-Polisher. Built with Next.js, TailwindCSS, Supabase, and OpenAI GPT-4, it logs every step for transparency and runs securely on Vercel

Smart Planning: LLM splits complex instructions into executable steps. Desktop & Web Automation: Supports applications, folders, and browsers. UI Awareness: Screenshots + OmniParser element detection for accuracy. Step Execution: Logs actions in real time for monitoring and debugging. Extensible: Add new apps, commands, or workflows easily.

This project showcases the power of LangGraph, a library for building stateful, multi-step applications using LLMs. By combining LangGraph with tools like RAG, web search capabilities, and specialized knowledge bases, AI assistant can navigate complex information landscape and synthesize findings into coherent, well-structured reports.

An engineering-grade, LLM-driven pipeline that transforms high-level Epics into complete test assets — Features, User Stories, Test Plans, Test Cases, and Playwright automation. Designed with step-wise generation, human-in-the-loop review, resumable execution, batching to avoid truncation, and versioned artifacts for traceability and evaluation.

Uniq is an LLM-powered browser automation agent that uses Gemini 3 Flash to decompose user tasks into multi-step plans, then executes them via Playwright with a cognitive loop of observe, decide, and act on real web pages. It handles everything from simple navigation to deep research tasks

You describe a function. An LLM hallucinates it. The BEAM compiles and runs it live in your runtime with no restart, no compile step, no test suite, no PR reviews, no deploy. The source also lands in your codebase for further use. Straight from user input into production. Probably fine?

This project demonstrates the use of Large Language Models (LLMs) for autonomous network security. By fine-tuning Microsoft's phi-2 on network traffic logs, I transform raw numerical data into a "language" that the AI can reason about. This is a foundational step toward 6G Native AI, where networks are self-healing and self-securing.

大模型

Steps into LLM

docs for https://github.com/mindspore-courses/step_into_llm

Stepping into LLM Models and its versatile functioning

My journey into AI & LLM development, guided step-by-step by ChatGPT 🤖🔥

Blueprint generator for OpenClaw - breaks complex tasks into step-by-step build plans using local LLM agents

A straight foward step-by-step to set up and train a LLM and turn it into an assistant