This is minor and just work can be done on the way to support LoRA, but hey, people can experiment. Added 'Text to Image' / 'Image to Image' label along 'Strength' in Settings, this will help people to understand what's the relationship between these two. * UI improvements: added tooltips for all buttons on the UI, which will help on discoverability. There are still limitations, namely: you can only use CoreML for 512x512 image size, less than 77 tokens (for both positive / negative prompts), and no ControlNet image generations. Now CoreML supports v2.x models, depth2img models, inpainting models and Instruct Pix2Pix model. * Expanded CoreML support: previously, CoreML only supports v1.x models. It is a versatile tool that can be used for a variety of purposes, including creating flowcharts, mind maps, UML. Program has a simple interface, is easy to use.

draw.io is a web-based diagramming application that enables users to create, edit, and share diagrams. For 'pose', besides the initial pose detected from given image, you have full editability, meaning that you can 'erase' a joint, move joint around, pinch to zoom (mouse wheel on desktop), and loading more poses from images. Representative Data in Graphics: Download Draw.io & Install on Windows & Mac. For 'scribble', both 'eraser' and 'paintbrush' can be used, and you can load / save them for outside apps to consume / edit. Depth / scribble / pose are now what we call 'Control Layers' you can switch to from bottom right 'Layers' button in the canvas. * ControlNet support: now we've added canny / depth / scribble / pose support of ControlNet inside the app. So we have to switch gear to support ControlNet obviously.

One particular thing is 'ControlNet' exploded in popularity. There are a lot of things happened in February. Hi, for people who don't know: This is the app currently under active development for Stable Diffusion in Apple ecosystem.

Dcraw for mac