Generative AI
Aug 26, 202619 min read
Granite 4.2 vs Gemma 4 vs Qwen 3.8 Benchmark
IBM dropped the Mamba hybrid in Granite 4.2. We measured what that costs against Gemma 4, Qwen 3.8 and Ornith 1.5 on one RTX 5090.
granite 4.2gemma 4qwen 3.8
Code-first AI for production
Build Production-ReadyAI Agents
Learn to build and deploy machine learning, RAG, and AI agent systems through practical, step-by-step projects.
200K+
Learners
IIT KGPIIT Kharagpur
Alumnus
4.8★
Instructor rating
Guided Roadmaps
Structured Learning Pathways
Follow a sequenced roadmap instead of piecing together disconnected tutorials.
Roadmap Path
Your Path to AI Agent Mastery
A structured, progressive roadmap for developers seeking to master AI agents, covering Python foundations, LangChain, LangGraph, and private multi-agent RAG.
AI AgentsLangChainLangGraph
Foundation Journey
Complete Data Science Journey
A comprehensive curriculum taking you from zero programming knowledge to professional data manipulation, mathematical visualization, and core ML model building.
PandasFeature EngineeringScikit-Learn
Latest from the library
Practical Technical Tutorials
Start with recent code-first guides, or filter the library by the skill you are building.
Generative AI
Aug 25, 202615 min read
Ornith 1.5 9B and 35B-A3B Benchmark on RTX 5090
We benchmarked Ornith 1.5 9B and 35B-A3B against Gemma 4-26B-A4B and Qwen 3.8 27B on one RTX 5090, and found the two mixtures are nearly identical on hardware and separated entirely by training.
ornith 1.5mixture of expertslocal llm
Generative AI
Aug 25, 202616 min read
Qwen 3.8 27B Uncensored vs Base Abliteration Teardown
We diffed all 866 GGUF tensors of an uncensored Qwen 3.8 27B against the original and found 131 matrices edited along one shared direction, with no measurable cost to quality.
qwen 3.8abliterationuncensored llm
Meet your instructor
I left a $100K/year job to teach AI full time.
I'm Laxmi Kant Tiwari, IIT Kharagpur alumnus, founder with a successful startup exit, and an engineer with 10+ years across industry and academia. Everything here is taught the way real production systems are built.
Read my storyFree Video Tutorials
Latest on YouTube
Full video walkthroughs, free: new tutorials every week.
Qwen 3.8 27B vs Qwen 3.6 vs Qwen 3.5: What Actually Changed
Qwen 3.8 27B vs Muse Glimmer vs Gemma 4 - Tested Locally on RTX5090 and Ollama
Qwen 3.8 27B Speed Settings Explained: MTP, KV Cache and Flash Attention
In-Depth Courses
Featured Professional Courses
Go deeper with complete projects, private repositories, and certificates.
Master Langchain v1 and Ollama - Chatbot, RAG and AI Agents
Master Langchain v1, Local LLM Projects, Ollama, DeepSeek, LLAMA 3.2, Complete Integration Guide.
- Setup & Integration: Professional Ollama and Langchain configuration
- Custom Chatbots: Memory, history, and advanced features with Streamlit
Master LangGraph v1 and Ollama - Build Gen AI Agents
Agentic RAG and Chatbot, AI Agent, DeepSeek, LLAMA 3.2 Agent, FAISS Vector Database.
- Memory-Enabled Chatbots: Dynamic conversations with persistent memory
- Database Integration: Direct MySQL query execution with LLMs
MCP Mastery: Build AI Apps with Claude, LangChain and Ollama
Build MCP servers & clients with Python, Streamlit, ChromaDB, LangChain, LangGraph agents, and Ollama integrations.
- MCP Architecture: Client, server, and transport layers
- Claude Desktop Integration: Direct MCP server connections
Browse the Entire Udemy Directory
Explore all available curriculum packages with our worldwide student ratings and pre-applied coupons.
4.8+ Avg Rating200,000+ Enrolled Students
What Students Say About Us
“The instructor explained clearly and its examples were easy to implement. I recommend it for introducing yourself into the latest advancements in transformers.”
“This is a nice hands-on video. The contents enhance the knowledge of fine tuning the model. I have also found that the 'theory' parts are very worth to learn.”
“The lecture is well-organized and delivered clearly, and easy to learn.”