Looking at articles associated with this tag.
An implementation guide for automating image dataset creation using Qwen 3.5 in a local environment. Ensure privacy and operate for free using a local VLM. This post includes a Python script for bulk-generating image captions. I’ve compiled practical tips ranging from Base64 encoding and controlling "Thinking" models to prompt engineering for excluding specific training targets.
Take full control of Ollama from Python! Learn how to implement your own API client using only the requests module. This guide covers essential customization techniques—from managing conversation history to prevent context overflow to manual VRAM offloading—all while minimizing dependencies and achieving total prompt transparency.
[AI Image Generation • LoRA Training] Boost your dataset creation efficiency! This guide explains an automated cropping method using the latest object detection model, YOLOv26. I'll walk you through how to automate the process to make high-quality training images easily—preventing subject cutoff and optimizing aspect ratios without the manual grind.
Using the removal of FaceFusion's NSFW filter and its counter-tampering measures as a reference, we will dive deep into the mechanisms of tamper-proofing and bypass logic that every engineer should know—including hash value check implementation in Python and security strategies for Git distribution.