SDXL LoRA trained with Civit-AI trainer for Pony V6/Pony merges, trained on the style of MLP artist Lunarmarshmallow.Lunarmarshmallow is primarily a MLP artist with a focus on detailed rumps and genitals. Their figures get decidedly squishy proportions. The LoRA can prompt for other things, but keep in mind it’ll do best with those kind of camera angles. See gallery for details.Starting with Eular A seems the better sampler, though YMMV.Uses Derpibooru tags. Works well with stock pony diffusionSample prompt&lt;lora:lunarmarshmallow_Style_v2_PonyXL:0.7&gt; by lunarmarshmallow, lunarmarshmallow, score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, source_furry, source_pony, source_cartoon, rating_explicit, anatomically correct, thick legs, wide hips,

Lunarmarshmallow Style [PonyXL]

Re-train with higher epochs and steps, much better style adherence even on basic pony diffusion model

Files with SHA256: 3d535b2d036adba2149783b082a1acb1dc35aebc871acb505c1a65da4e5f10d9