Description
scenery:1, nightime, landscape, snow, (christmas), (town streets),
(applejack, youthful feral pony applejack),
sleigh (applejack is pulling sleigh, sleigh filled with presents, Christmas applejack harnessed to sleigh),
<lora:stirrups:0.5> (reins:1), reins:(leading to sleigh), (bite gag:0.4), mouth detail, closed mouth smile,
looking up, closed eyes, joyful horse, walking on street,
source_pony, source_cartoon, accurate feral,
masterpiece, ultra high resolution,
high angle, three-quarter view,
Negative prompt: mutation, deformed, 3d, censorship, jpeg artifacts:1.6, saddle, applejack breasts, additional characters, clone, clones, duplicates
Steps: 20, Sampler: DPM 3M SDE, Schedule type: Karras, CFG scale: 2, Seed: 3555421829, Size: 1024x768, Model hash: 6abedc4dfc, Model: cinema_scope, Clip skip: 2, Lora hashes: "stirrups: eb1c01497a90", Version: neo```
Source
not provided yet
It’s understandable. Generating a perfect image is very hard. This is why many people generate a reasonably good image and then use inpainting to correct problematic areas.
Edited
None of what I said was actually intended as negative, although the last paragraph would not be totally wrong to interpret that way depending on the prompter’s intent.
Getting an image generated of a pony to pulling a sleigh isn’t easy. Lots of trial and error.