Trained Lora with Pony Diffusion

Note: strength should be between 0.4 - 0.6 for better results

images were generated with DuchaitenJourney Model and Realistic Vision V6.0 B1(soon)

Recommended configuration

Model: CyberealisticPony

LORA Strength: 0.8- 1

Embedding: Fast Negative

VAE: vae-ft-mse-840000-ema-pruned

Sampling Method: DPM++ 2M Karras

Steps: 20-25

CFG: 9-10

Clip Skip: 1

Upscale: 4x-UltraSharp

UST: 39

RNI: 0.5

This configuration gives as results looks like face but only in this model, working in a new version for other models