If you’d like to show some love, feel free to react w/ emoji, post, rate, follow, or simply drop some feedback, I'd appreciate it. 🫡🫰 Feel free to also report any issues.
The Anima model should be better at maintaining the style. Should also have way less eye issues.
Addressing common DMs from Illustrious model: If your results look off, disable additional LoRAs. They actually distort the core style, even the ones that claim to add 'detail, aesthetics, shading', etc. Different checkpoints may do this as well. This also applies to all my models.
Anima Base v1.0
Images trained and generated on Anima - base-v1.0. Generation details, prompts, and model info are listed with most images: CFG: 5, Steps: 20-30, Sampler: DPMPP 2M Simple, and SeedVR2. Feel free to experiment!
Trigger Word: @mxldx
Illustrious trigger word is on their detail card: mxldx. Illustrious model may need adetailer.
Description
FAQ
Comments (11)
Can you do Masoq style?
Any idea how tf Maldo generates 256 images per day?
Pretty sure it’s all automated, using categories set up as (character) (action/location) (angle) on repeat. Maldo uses the same setups every time, only swapping the character to mass-generate similar images. Seems common whenever I look at a dataset when someone suggest for me to make a style of an AI image creator.
@Ty_Lee Do you have any idea how to do it? I've been trying for days on ComfyUI, but I can't get the images to appear consistently and without deformations
@newselitecraft802 How are you writing your prompts, if you keep getting deformations it might be related to prompting and a few other things.
@Ty_Lee I mean deformations that I ca't generate multiple photos in batches like he does, sometimes the pose isnt right, sometimes the angle is static, there are always flaws, he must have some trick.
@newselitecraft802 When I said automated, I meant just the general image creation, they will still have to create a larger batch then check images to then cut it down to 200~ images that are decent enough. If you get a lot of deformations (other than 4-3 fingers) though, this could be because of prompts or other things.
@Ty_Lee I understand Ive been looking for a workflow that does something similar and I cant find it. Sadly, am not good at ComfyUI.
@newselitecraft802 Ohhh, like the prompting... Yeah, I think I’ve got a few examples posted a few times that do that, but I never really looked into it since I don’t actually need 😭.
Only one I remember is comfyui prompt cycler, might have to prefix with a node the character if you dont want to change it manually each time.
@Ty_Lee Can I talk to you in DM?
@Fhoon What i do is when i have a workflow i have fully setup, i either batch it with a setup that reads 50 prompts from a txt file for example, but then there is no dynamic lora config ect, just prompt. Or i use chatgpt to generate the python version of the workflow and then you can parameterize everything, a bit more performant too