Zalecane podpowiedzi

Concept art, depth of field, sharp focus, colorful

Zalecane negatywne podpowiedzi

Asian, doll, sketch, painting, 8bit, pixel art, pixelart, b&w, blac and white, plastic, text, monochrome, watermark, canvas frame, ugly, closed eyes, elongated body, elongated neck, blurry, out of frame, disfigured, deformed, fused fingers, stump fingers, extra hands, extra feets, extra fingers, extra arms, extra legs, poorly drawn hands, mutated fingers, mutation, mutated, mutated hands, (poorly drawn face:1.4), normal quality:1.4), crossed eyes, (worst quality, low quality

Zalecane parametry

samplers

DPM++ 2M Karras

steps

42 - 65

cfg

4.5 - 8

clip skip

2

resolution

720x1280

vae

included VAE, xlVAEC_c9.safetensors (235745af8d)

Zalecane parametry wysokiej rozdzielczości

upscaler

4x-UltraSharp, 8x-ESRGAN

upscale

1.5

steps

10

denoising strength

0.07 - 0.4

Wskazówki

Leave the negative prompt field blank to also achieve good quality of the generated image.

Works great with various LoRa and embeddings.

Tamarin_XL
----------------------------------------------------------
+ You will be pleasantly surprised by the excellent results.
----------------------------------------------------------
+ To create Tamarin_XL model, only the best, in my opinion, models were used.
+ The refiner is unnecessary, and VAE is included.
+ All in one. Best model ever for everyone!
+ If you leave the negative field blank, you can also achieve good quality of the generated image.
-----------------------------------------------------------
+ Works great with various LoRa and embeddings.
-----------------------------------------------------------
+ Good luck to everyone!

Współtwórca

Poprzedni
epiCPhotoGasm - V4 - One 4 All
Następny
CyberRealistic - V4.1 "Back to Basics"

Szczegóły modelu

Typ modelu

Checkpoint

Model bazowy

SDXL 1.0

Wersja modelu

v1.0

Hash modelu

2d1805f294

Twórca

Dyskusja

Proszę się log in, aby dodać komentarz.