Been fighting with A1111, think I figured out a problem, A1111 had ALL my LoRa's wanting to enable or other keywords overwriting things. Think I fixed it finally, I was wondering why I was getting such odd results than the prompts from other users using the same model. (All models and lora's I use usually are available on civit ai website.

Twilight Sparkle,solo,((derpibooru_p_95)), sitting, pony, open mouth, ((safe)),
\<lora:twilightSparkleFlat\_v10:1\>
Negative prompt: (derpibooru_p_low), sfm, extra legs,bad anatomy compression artifacts,(salvador dali:0.5),CCTV surveillance ,ugly,poor quality,deformed (distorted:0.9),signature, 2girls, ||penis,erection||,

Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 705497202, Size: 640x384, Model hash: f9ad6f8833, Model: ponyDiffusion_v4, Denoising strength: 0.7, Clip skip: 2, Hires upscale: 2, Hires upscaler: net_g_latest pony, Hashes: {"model": "f9ad6f8833"}