Preface
A small links guide to what we feel are good extensions to use, there is no tutorial attached (at this stage). If we've missed anything, or you have a reason why ANOTHER extension or tool is useful - OR you use another UI/ux system please let me know in the comments.
UI/UX Versions:
This is ONLY for Automatic1111, and SDNext which is mostly cross compatible.
(Though i've never used SDNext, i'm just aware that it's a fork off A1111)
Extensions List:
Downloading your models:
https://github.com/NoCrypt/batchlinks-addon
https://github.com/etherealxx/batchlinks-webui
https://github.com/butaixianran/Stable-Diffusion-Webui-Civitai-Helper.git (We haven't seen this working as of recently, it could just be us, but when it does work in tandem with the CivitAI plugin it's great)
(Paid resource): https://github.com/civitai/sd_civitai_extension - Must have plugin, you only have to sub at he base level to CivitAI to use this, and it's pretty dandy. It's sort of still in a beta stage, but it's really REALLY useful.
EXIF Data:
While discord DOESNT STRIP EXIF data anymore, a few older bots don't always see the data unless you have PNGinfo extensions like this as far as we're aware:
https://github.com/derrian-distro/sd_webui_stealth_pnginfo
Merging & Mixing:
https://github.com/hako-mikan/sd-webui-supermerger.git - OH BY GOD, i'm sure i'm going to be told that this is not CORRECT TO USE BLA BLA BLA. Opinions in the comments, but don't pass them as facts! This is a WONDERFUL tool, and I have YET to meet anyone who actually understands the TRUE MATH BEHIND IT. (I'm sure there are some) This is a complicated but fairly EASY to use tool - DOWNSIDE is in a recent update the sliders were introduced in a new spot and don't work even wtihout a theme - you have to keep clicking a blank spot and refreshing - but it seems to be ok if you just leave them AS IS, and just add your extra numbers down in the bottom section. IT's recent update is a littttle tricky to find stuff - but LoRa baking is still in there, and it's still WAY fun to use!
Note: That above plugin? Also allows you to merge WITHOUT saving it, so you can test the generations at the same time and THEN save it. (We're chicken shits.) - Also you can MERGE & GEN - JUST BE AWARE THIS CAN BE A RESOURCE INTENSIVE PLUGIN when you're running say on Vast or Runpod - you may end up finding yourself with a "QUIT" function on runpod in case it breaks something.
Model CONVERSION & Trimming:
https://github.com/Akegarasu/sd-webui-model-converter.git - THIS ONE entirely right now has been updated so you MAY never need to use the Toolkit again. (That's not to say don't use either or both, but i've left toolkit off this list.)
Model converter's NEW update NOW features similar conversion and trimming tools to Model Toolkit. Fix clip, force clip into INT64 position (I barely understand what that is). IT also now employs "FIX KNOWN JUNK DATA". It doesn't always trim it the same as Toolkit, so if you're looking for a PURE 1.99 toolkit trim? Use toolkit.
Either that or baking in Vae makes it more like 2.5GB?
I'm not even sure.
Last time we baked vae in for Earthnicity, it barely worked.
Creating your own Embeds to go with YOUR models:
https://github.com/klimaleksus/stable-diffusion-webui-embedding-merge - We have written a tutorial already for this, so we won't rewrite it in here. This lets you create Textual Inversions/Embeds at the click of the button. 75 MAX tokens (words) for a prompt, and the tool itself has actual regex and otherwise not just on github but AFTER installation has a full tutorial on how the math and other regex works. IT explains HOW the numbers are required and what does and doesn't work. Plus when you pick a filename, and you hit save and you've done something wrong? IT WONT save the file, giving you a chance to look where you went wrong.
Reccomended UI Theme for A1111:
https://github.com/canisminor1990/sd-webui-lobe-theme - Weirdly i dont know if it's an error on my end but sometimes with a new update this WONT show Textual Inversions in the Extra networks cards. THIS COULD BE A ME thing, but regarldess if you're already used to your activation tags, or have access when yo're renting a GPU to Jupyter's file list - easy enough to grab the file activation tags. Like i said this small blip may be because I MASS Upload all my TI's from my own computer because not all of them sync properly from Civit. (not civit's fault, I make too many).
Also that this is EXTREMELY nice on the eyes, and is similar enough to the likes of SDNext. It has CUSTOMIZABLE fonts, color themes, it's also got a toggle going between it's FORMER version "KITCHEN THEME". You can toggle QoL features, screen reader features DARK AND LIGHT mode!
If you're like me and you DONT want to fuss with CSS coding a whole new theme for base A1111- this is kinda where you might go. Look, it's great because everything's organized into an easier to see function.
Backing up to Huggingface:
https://civitai.com/models/99314/huggingface-backup-jupyter-python-and-colab
We'll be editing the TUTORIAL for this soon, as we have screenshots now for the Jupyter version of this.
Conclusion:
Nothing in this list is a FORCED FACT, as in - this is the list of content WE use on a weekly basis. Recently we were confused at Lobe's not loading our TI's so we didn't use it the one time, but we're not a programmer, and as we said? It's likely an error on our end.
Now, if you NEED tutorials on any of the above feel free to ask, and we'll try and get around to making either videos or even text based ones for here AND on our website.