Mecha Musume + Gundam + Mecha Slider LoRA
Making models can be expensive. Do you like what I do? Consider supporting me on Patreon 🅿️ or feel free to buy me a coffee ☕
I've been trying to make a mecha/gundam LoRA for a while now, but between me being picky and other creators making AWESOME ones, I never posted any.
I'm finally ready to share this one, even if today is one of the worst days of my life. I'll look back at this model and remember it.
I've tested it on all my models and on Counterfeit, and works an all of them and with all my negative embeddings in interesting ways. Regarding the weight, you'll get a middle ground change at weight 1, but this really works as a slider (as you can see from the final 2 comparisons). Adjust the weight from 0 to 2 for the desired effect.
Can make Gundam, Mecha, Mecha Musume, Bodysuits of different colors.
Have fun and consider leaving hearts and reactions.
How to use LoRA's in auto1111:
Update webui (use
git pulllike here or redownload it)Copy the file to
stable-diffusion-webui/models/loraSelect your LoRA like in this video
Make sure to change the weight (by default it's
:1)
Description
FAQ
Comments (29)
Adding a post with more examples that didn't fit the first 20.
Sorry to hear you're having a hard time, hope it gets better soon. Hang in there. And since apparently posting help you relax somewhat, keep in mind community appreciate your work :)
I'll check out how this works later.
Yeah I know. This community has been great so far. A happy bubble
Sick concept as usual. Hope you're better soon.
Time heals everything
Take a break if you need to dude, you're one of my favorite creators on this site, wouldn't wanna lose you
If anything I'll do even more stuff. Need to keep me distracted
Works extremely well! I don't know how you manage to do it, but you nail a lot of obscure concepts.
Not to add to your probably gigantic work load, but do you think a Godzilla lora would work? I'm not sure how well the AI would understand the concept, and I've not seen anyone try a non-Waifu Godzilla one so it may be too janky for the Ai to figure out
I've failed this so many times lol. I'll see what I can do with Godzilla
Love this Lora model, it's dope like all of your models, Can u give me some information how can i train clothes, i haven't found a single resource that teach training clothes, or if you can give me a video to follow, that would be amazing, thanks
I think @EDG made a guide on that.
Yea just saw it, but it only says about captioning, i know captioning, but the nothing about training is mentioned such as repo or parameters or steps and epoch, I don't know anything about it, wish he added that too would've been very helpful
@Astro9 he probably made some other more general guide about training. However dataset creation and captioning is 98% of what matters. You can literally shover a good dataset into a pre made lora script such as Lina's and you'll get a freat lora out of it. Trust me
@Lykon Thank you so much for helping🙏🙏, where can i see Lina's and other premade script
@Lykon Thank you so much
First of all, I'm sorry to hear that. I'm not gonna pry--and no one should, but whatever it is, it must be bad enough and personal enough for you to say it like that, and on here, your Communty, of all places. I truly hope you're okay. Gambatte.
On a lighter note, I originally wanted to celebrate this release (but it'd be in poor taste right now), because even though, as you said, there are mecha LoRAs out there--and they ARE great, their design or aesthetics aren't EXACTLY what I had in my mind (at least for most of 'em). But this model of yours, I think it "solidly" nails it, again. Maybe I'll explain it better in a later post or review. For now, thank you for another creation, author-san.
My notification panel is full of releases from you, and I have nothing but gratitude for that. But if you're working non-stop, and you're going through something right now, you may want to ease up on the pedal a little, for your own sake of course. I'm sure everyone in the Community would agree to missing a little of your extreme hard work, in exchange for your physical, mental and emotional well-being. Stay cool @Lykon-sensei, cheers and God Bless. ❤
Wow thank you so much
I can't tell you entirely, how much i appreciate you. Thanks to you, I can generate good mecha.
Please, Please make Lora of Gundam Aerial(The witch of Mercury).
Please.
Planning to do some stuff from the Witch from Mercury already. Stay tuned
I am surprised how well the hands are turning out for me. Still kinda gacha, but it is a big improvement.
2 Weeks can make a huge adjustment in our perspectives. I hope it has helped yours as well. I hope you are safe and happy. Thank you so much for your work on this Lora.
Wow looks amazing. How did you avoid face being trained ?
This one is AWESOME! Exactly what I've been looking for! Thanks and thanks again for posting this one. Looking forward to your new creation too!
This is some impressive work man! How do you train and setup a Lora so that the weight determines the style? Can you bake in a multi action frame like a boxer making a punch in 3 stages? Would love to understand the process. Thank you!
I didn't quite get what you mean. Anyway you just need to guide the learning process with proper data
It's turned out really amazing, thanks, Would you be able to share training parameters please
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.