Agree the purpose of the repo

[darth-veitcher]

Discussed in #1

^{Originally posted by darth-veitcher February 25, 2023}
As discussed in AUTOMATIC1111/#8029 there's a gap currently for stable diffusion users to effectively manage models at a header level, their associated experiment outputs, and reproducibility.

This repo is to attempt an approach for resolving this with a model workbench type approach which, at a high level, would encompass the ability to:

1. display, search categorise, obtain, and record model related metadata. Header level type stuff and will require extending base model class in A1111 as well as providing tight integration with certain repositories like HF and Civit out the box; and
2. store configuration and replicability for our experiments with said models. Things such as settings used, base model, outputs, prompts etc. We should be able to persist this, search, and share with other users in a format that can be easily imported.

From a UI perspective it feels it needs its own tab(s) for dedicated detailed admin/workings but then snippets then extruded into the other tabs for quality of life.

Day 1 approach might be to just see if we can recreate some kind of sidebar that scrolls your outputs and displays them in an easy-to-reuse manner when you're in sections of the UI such as text2img or img2img.