Ganonmaster
75 lines
3.0 KiB
Markdown
75 lines
3.0 KiB
Markdown
# Open3DLab SlopScaling (OLSS) 5.0
|
|
|
|
This repository contains several components used to make Open3DLab's slopscaling April Fools prank work.
|
|
|
|
## Widgets
|
|
|
|
These widgets are the core of the "AI Chat experience". The first element is the `ai-assist-bar`, which is the main chat bar. The chat bar renders canned responses based on the "model" selected.
|
|
|
|
During the time the elements were live, queries were sent to the backend when the user confirmed. The elements in this repo have been modified to not send responses to the Open3DLab backend.
|
|
|
|
The other element is the `ai-assist-clippy`, which is an orb that renders a dialog with pre-defined options and fake responses. Let us know if you find the fake options and responses interesting or funny.
|
|
|
|
In viewing the source code, you may find that there's more messages and placeholders that you may not have seen at the time the april fools prank was live.
|
|
|
|
The `widgets` folder contains the standalone web component package for the two Lit custom elements used by the prank UI:
|
|
|
|
- `ai-assist-bar`
|
|
- `ai-assist-clippy`
|
|
|
|
The package is managed with `pnpm`.
|
|
|
|
### Run the widgets locally
|
|
|
|
From the repository root:
|
|
|
|
```sh
|
|
cd widgets
|
|
pnpm install
|
|
```
|
|
|
|
Once dependencies are installed, you can run the package checks:
|
|
|
|
```sh
|
|
pnpm check
|
|
```
|
|
|
|
### Preview the widgets in Storybook
|
|
|
|
The easiest way to inspect and tinker with the components is through Storybook:
|
|
|
|
```sh
|
|
cd widgets
|
|
pnpm storybook
|
|
```
|
|
|
|
This starts a local Storybook server at `http://localhost:6006/` where you can:
|
|
|
|
- preview both widgets in isolation
|
|
- switch between the supported `mode` values with Storybook controls
|
|
- interact with the components directly to test their loading, typing, and display behavior
|
|
|
|
To build a static Storybook bundle instead of running the dev server:
|
|
|
|
```sh
|
|
cd widgets
|
|
pnpm build-storybook
|
|
```
|
|
|
|
## ComfyUI Workflow
|
|
|
|
In the `workflow` folder, you will find the ComfyUI workflow that was used to mangle all the thumbnail images for the Open3DLab SlopScaling. This is an extremely simple, and extremely bad workflow that should under no circumstances be used for anything serious. However, you can theoreticaly run this on any device that has sufficient VRAM to run SDXL.
|
|
|
|
### Run it locally
|
|
|
|
To run it locally, you will need:
|
|
|
|
- Graphics hardware with sufficient VRAM
|
|
-  installed on said hardware
|
|
-  (can be installed through ComfyUI Manager)
|
|
- the  (specifically, Ragnarok_by_RunDiffusion)
|
|
|
|
Once you have all your dependencies installed, and the checkpoint in the right directory, open the Yassification nodegroup, and set the model on the `Load Checkpoint` node. Plug in an image in the `Load Image` node, and hit run.
|
|
|
|
Your inference times will vary. On a Macbook M1 Pro, inference time was roughly 30 seconds per image. On an RTX4090, it is 2-3 seconds. On an Intel Arc Pro B50, it was roughly 15 seconds. It was not tested on cloud instances.
|