Process Explorer for AI training. Real-time PyQt6 graphs, pattern detection, threshold overlays. Supports Ostris, & eventually Kohya & HuggingFace. Native desktop PyQt app with resource monitoring. Alpha release.

The AI-Avtaar pipeline, built on Streamlit, automates creating custom AI characters through three stages: Train (builds a LoRA model from user images using Kohya_ss), Generate (creates realistic images via Automatic1111), and Try-On (applies virtual clothing using CatVTON).

Skills that make AI coding assistants more useful — neurodivergent support, LoRA training guidance, and more. Built for Claude Code, works with any LLM agent that reads markdown.

🏆 Mejor TFM Deep Learning 2025 y Mejor Proyecto Académico MIOTI 2024-2025 - Sistema de generación de outfits Fortnite con IA: Fine-tuning SDXL + LoRAs especializados. - Incluye API REST, interfaz web y guía completa de despliegue en Vast.ai. - Stack: Python | FastAPI | Stable Diffusion XL | KOHYA_ss | ComfyUI | OpenAI

Train Stable Diffusion XL (SDXL) LoRA models locally using Kohya SS

Enhance LLM agents with practical skills for neurodivergent support, focus, and communication to improve AI coding assistance and interaction quality.

A Generative AI Dojo for LoRA/Dreambooth/TI training using kohya_ss/sd-scripts with a WebUI layer on top

一個極簡、即用即棄的 AI 繪圖與訓練環境，在嘗試各種新套件或訓練模型時，提供無心理負擔的純淨體驗

LoRA Training Notebook for Stable Diffusion XD (Google Colab) A Google Colab workflow for training LoRA (Low-Rank Adaptation) models on Stable Diffusion 1.5 using kohya-ss sd-scripts . ( Older Version 0.1.0 )