Tags Down Below
It is HIGHLY recommended to use Autism Mix XL, preferably the Pony version it was trained on that, and the results are most distinct with that base model.
This is the former lead mascot behind GinjaNinjaOwO (YouTube), you might know them as ActuallyRea, or just Rea if you've seen them before.
I took on a few Bounties, and THIS was one of them. NGL I wasn't hopeful about it. Although along the way I got to research the artist, their style, and personality mainly. It was a lot of fun! Despite how unconventional theirs art is I'm so glad I could replicate it.
I recommend you check them out, if you like their work, and personality don't hesitate to donate to their Patreon, you'll get to see their new mascot, references, WIP sketches, wallpapers, and more. support artists, and have fun.
>> TAGS <<
Pony XL Quality Tags
score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, anime screenshotBase Body
1girl, solo, long hair, red hair, hair ornaments, freckles, glasses,Main Outfit
jacket, shirt, high-waist shorts, chokerDescription
FAQ
Comments (3)
Oh my God, thank you, thank you, thank you, thank you. I've been looking for one of these forever; it looks so good. I can't thank you enough. I've been trying to make my own, but I didn't know where to start. If you have any tips or knowledge, I'm more than happy to listen. Just message me if you want. Again, thank you so much.
Thank-you, I appreciate your enthusiasm. I think the best knowledge I can offer is to learn by example, and innovate through experience.
You can download the entire dataset I used to train this LoRA from the details panel, it has all the images I used, and any training prompts associated to each image, I focused on screenshots, trying to get a variety of poses as much as possible, and describe each image in exacting detail to best nail down her design it took roughly 3-4 hours to create the dataset, but the results speak to the effectiveness of the workflow.
Speaking of which I have also written an Article on dataset workflow, which should be helpful whenever you want to locally train a LoRA as well.
https://civitai.com/articles/4774/an-insight-on-dataset-workflow-for-sdxl
I hope this is helpful, and leads you to make much better LoRAs!
Hello, there is a chance for you to port this one to illustrous?
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.