Please post the first picture directly here in the Model Gallery. This is the best way to support creators :)
Hello ♥
for whatever reason you want to show me appreciation, you can: ❤️Ko-Fi❤️
_________________
This was my first images back in 2022 with Midjourney with the prompt:
green spongebob squarepants with tons of oreo
Imagine, this was my first ever generated image with AI and this was the first “prompt” I had in my head 🤣
Now, with this Lora and ForRealXL this is the result
This Lora was a test with a new captioning method that delivers much better results than expected.
So you can try it out too.
This Lora was trained with 34 Midjourney V6.1 images. (Version 0.1)
Settings
0.3 - 1.0
Trigger (v0.2+)
aidmaMJ6.1Trigger (v0.1)
midjourneyv6.1Hope you get a lot of beautiful results ♥
Description
v0.2
start with lower Strength
same dataset
trained with rapid
its the "light" version of the regular v0.2 Version
Try to mix both Versions!
FAQ
Comments (16)
With the awesome shit you put out on the regular, you GOTTA make a checkpoint. It would be one of the best checkpoints for Flux bar none
Thank you so much for your kind words.
Maybe one day 🥶
Let me first say this a a well done LoRA that definitely has its uses to bring back the sdxl style with FLUX capabilities under the hood.
So while I appreciate this existing it gives me a bit of a "last gen graphics" kind of feel. To me it makes images look more cartoonish and more obviously ai generated, which might be something you want from your images, but I prefer the base flux graphics and more realistic loras over sdxl.
Feel free to share your opinions about bringing back old styles and the classic "ai look" <3
that's the point of a style lora; you don't need to leave a comment that you prefer something else; this is subjective stuff, it can go without saying
The whole point of this LoRA is to generate that "MJ A.I. Look" 😅
@volatilevisage2 I agree that it's subjective stuff, for sure. But I think the point still stands that it makes images look more cartoonish and (at least for me) more obviously ai generated and that is a valuable information that has a point to be posted here. It's a good Lora and it has its uses, but it also has its drawbacks.
Maybe I should have asked it in a way that encourages a discussion about why that might be and how to deal with it
amazinggg...
good
How did you do the flux animation? Port into SDXL?
no this is just with Luma Dream maschine :)
Hi, is the trigger word necessary to obtain a result? No one developed a keyword database downloader yet so trigger words feels like passwords whereas when you lose the trigger word, you might also lose the purposed lora effect...
If I were to make a standard for checkpoint and lora naming, I would name them with name first, then a list of each trigger word, then, of course, finished by the file extension.
I keep an excel file that lists name, version, civit ID, strength, triggers and filenames. Otherwise I'd be lost.
for my lora you need the trigger to activate the lora. there are lora without trigger too.
i try to improve my trigger that you can remeber the name, but its hard, can understand.
@Ai_Express this is smart.
There is a solution if you generate your images locally. If you already run ComfyUI or another local image generator with a simpler interface, try SwarmUI. LORAs can be fed directly with metadata like image, trigger phrase, creator description etc. via the LORA's Civitai link. It runs very stable (C#). ComfyUI runs in the background and can be used directly or via a simplified interface. Of course it depends on your personal preference, but it is worth a try.
@CheesecakeJohnson, @Ai_Express, @AIDigitalMediaAgency, @o639
There is a LoRa Metadata Viewer available online, just drag and drop your safetensors file and you can see it's trigger words and tags. Enjoy:
https://xypher7.github.io/lora-metadata-viewer/
Looks like you can download it on CivitAI too:
https://civitai.com/models/249721
In FORGE you can simply use styles
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.