Loading
Have a ComfyUI workflow? Get a readable Workflow Doc in your browser.Make a workflow readable.Try it now ↗
Package profile
README
This repository contains custom nodes for ComfyUI.
image
Example workflow in the prompt_with_styles.json file and mikey_node_workflow.json
This node allows you to prompt with a style and loras. You can add a lora with or syntax for a weight of 0.8 as an example.
Related resources
40 connectionsSources
3 sourcesSource excerpts
1 excerptSource context: Repo bash-j/mikey_nodes
<lora:lora_name><lora:lora_name:0.8>You can add styles by using the <style:style_name> syntax.
Some of the styles built in are:
SAI-Enhance, SAI-Anime, SAI-Photographic, SAI-Digital art, SAI-Comic book, SAI-Fantasy art, SAI-Analog film, SAI-Neonpunk, SAI-Isometric, SAI-Lowpoly, SAI-Origami, SAI-Line art, SAI-Craft clay, SAI-Cinematic, SAI-3d-model, SAI-pixel art, SAI-Texture, photographic, deep-field, analog-film, cinematic, red, documentary, nature-photography, editorial, polaroid, 3d-model, low-poly, modeling-compound, diorama, miniatures, abstract, anime, comic-book, digital-art, fantasy-art, dark-fantasy, oil-painting, watercolor, charcoal, color-pencil, crayon, cross-stitch, felt, origami, scrapbook, isometric, line-art, vector-art, neon-punk, pixel-art, tile-texture, lithography, mosaic, woodblock-print, sticker, stained-glass, tattoo, statue, album-art, wes-anderson, vaporwave, clarendon, gingham, juno, lark, nonagon, kaleidoscope, haunting, glam, mecha, padam, collage, paper-mache, macrame, quilling, pottery, batik, felting, marquetry, wildstyle-graffiti, bubble-graffiti, stencil-graffiti, throw-up-graffiti, tag-graffiti
There is a V1 and V2 Prompt with Style node with different functionality, style selection instead of style syntax, no clip conditioning. Check them out if you like to break your workflow up into more nodes. image
Wildcards are supported using the __word__ syntax. Folder location is comfyui/wildcards You can select more than 1 line from the wildcard by using the syntax 2$$__wildcard__ for 2 lines in this example. You can also add a word to search for in the wildcard file e.g. __wildcard|word__
You can also make sure it selects the same line from the wildcard by using the __!wildcard__ syntax, not the ! before the filename. You can also use __+wildcard__ __-wildcard__ and __*wildcard__ to use the offset + 1 or - 1, or randomly select a different line.
You can add your own styles to a file named user_styles.json in the root ComfyUI folder. You have to create this file yourself.
The format needs to look like the following example:
{
"styles": {
"newspaper": {
"positive": "newspaper comic strip, panels, black and white",
"negative": "photograph, color"
}
}
}
You can also add your own custom ratios to a file named user_ratios.json in the root ComfyUI folder. You have to create this file yourself.
The format needs to look like the following example:
{
"ratios": {
"test": {"width": 1024, "height": 1024}
}
}
image image
Example workflow: mikey_node_workflow.json
This is a sampler node as a companion to the Prompt with Style node, to allow for very straightforward image generation with SDXL.
The node uses a base -> refiner -> upscale with model -> base to produce the final image. It has an inbuilt image complexity detection function to make sure there aren't too many steps run on the final base sampler which can scramble faces and other simple areas on a large image.
The Base Only version of the sampler uses a slightly different approach. It uses a base -> base -> upscale with model -> base to produce the final image. The other key d
Nodes in this pack
74 nodesVerified Sep 12
Verified Sep 12