!!! Recently some of my images generated with the previous version were blocked for TOS violation. Keep the characters "clean", and above all, dressed !!!
V2
Begin your prompt with: The image showcases a card from the "GARBAGE PAIL KIDS" series. The card features a character named 'CHAR NAME'. 'CHAR NAME' is and start describing character, background, objects etc.
Should work well at 1
V2 should be better at prompt adeherence, hands and text.
I'm sorry about the early access but this is the only way to avoid the immediate automatic download made by bots that then smear my LoRas all over their shitty sites (yes, I'm pissed).
Wait for 3 days if you think it's not worth spending the buzz.
V1
weigth: from 0.4 to 0.8
If you like what I do donate some Buzz or buy me a coffee here
If you want to post an image made with this LoRa on civitAi, please, do not cross-post but create a new post on this page with the ( + add Post) button, it will help me gain buzz without the need to rely on tipping.
Please, DO NOT REPOST ON OTHER SITES without my permission.
Description
FAQ
Comments (21)
500MB?? sigh... [downloads]
Yeah, 64 dim and alpha
@Clumsy_Trainer it's a great Lora
you have 'the touch' I implore you to make some Schnell loras.. filter by schnell and its a wasteland, which is too bad cause the licensing is better
@Narz atm my Pc is broken so, after training a LoRa I'm not able to test. Every LoRa I publish is usually thoroughly tested for best performance/prompt adehernce and quality. Also the vast majority of my LoRa networks are merge of caption and captionless. I promise that as soon as my Pc is back I'll train a Schnell LoRa. If you have suggestion for a particular LoRa just let me know.
@Clumsy_Trainer sorry to hear that! but it'd be awesome. recommendations? really just filling in gaps that dev is leaving for schnell.. there's only a couple realism loras, barely any lighting/sharpness/detailing tool loras.. pretty much anything is welcome! i figure there will be a renaissance for schnell once people realize they can't sell many of these pretty dev pics heh
since you do a lot of styles, i guess a good comic style is in order for schnell. some good artists mashed up for nice B&W line work / full color graphic novel art? there's my suggestion ;)
This is amazing. That's it. That's the review.
Hey, you're too kind, thank you!
This is crazy good. My 80s inner child sends you the biggest of thanks!!
thanks! glad you liked it!
ok OK.. I'll stop for now 😅
GOOD JOB! 🙌 💯
I spent an entire night and the day after just coming up with silly names. It's addicting :D
Once I saw your instructions about keeping the strength below 0.8 the text cleared right up. I can't believe how well this LoRA behaves. The style is everything I remember from being a kid and collecting these. Great work!
this is so good it might as well have been made by the original artist. i am curious to see how it responds to prompts!
thanks! It depends on the prompts but flux it's pretty solid as a base model and if you don't crank up the weight beyond 0.8 it should interpret your prompt pretty well. If you generate images with this model, please, share and let me know
Oh my god!!!!! This is much greater than any of the NSFW loras! THANKSSSSSSSS!!
yeah, no tiddies can still be fun 😅
Stunning. This LORA shows the advantage of training with massive amounts of new model data- hence the size. The tiny LORAs merely re-enable styles already present in Flux (Flux-Dev was nerfed by preventing a ton of style activation via prompting - the tiny LORAs merely reverse this)
Great job! But I can't seem to replace ‘Garbage Pail Kids’ with ‘Les crados’, which is the name in the French version. Do you have any tips?
OMG! This is wonderful!
hey, thanks! Hope you'll have some fun with this