AI

Hands-on guides to LLMs, agents, prompt engineering, and the AI tools Botmonster runs every day for real work, not demos.

SDXL 2.0 LoRA: 50-300 MB adapters on 12 GB VRAM

SDXL 2.0 LoRA: 50-300 MB adapters on 12 GB VRAM

Fine-tune Stable Diffusion XL 2.0 with LoRA adapters: dataset prep with vision-language models, training with kohya_ss, custom styles on 12GB VRAM.

Underground vault library with glowing holographic books arranged in vector space and a robot librarian retrieving relevant volumes

Setup a private local RAG knowledge base

Build a private RAG with Qdrant, BGE-M3 embeddings, and Ollama. Index your documents locally and answer questions with zero data leaving your machine.

Building multi-step AI agents with LangGraph

Building multi-step AI agents with LangGraph

Build production-grade AI agents with LangGraph's graph architecture. Add self-correction, memory management, and multi-agent patterns to Python apps.

High-end gaming desktop with illuminated NVIDIA GPU visible through a glass side panel, surrounded by floating holographic neural network diagrams and data streams

Run Llama 4 Scout locally: 24GB VRAM, GGUF, real speeds

Run Llama 4 Scout locally with Unsloth dynamic GGUF. Real VRAM by GPU, which quant fits 24GB, Ollama vs llama.cpp, and honest tokens per second.