neurowelt / keros-diffusion

Controlling SD XL diffusion inference

  • Public
  • 11 runs
Iterate in playground

Input

string
Shift + Return to add a new line

Input prompt

Default: "An astronaut riding a rainbow unicorn"

number
(minimum: 1, maximum: 3)

Low richness gives less detail and a misty look. High values make images stronger, rich, complex. Bugs expected for higher values

Default: 1.3

number
(minimum: 0.9, maximum: 1.1)

Quite delicate, can correct overexposure that guidance or richness create. Use 1.0 for no change.

Default: 1

number
(minimum: 0.7, maximum: 1.3)

Low values give high local contrast and sharper changes between objects better for illustrations, glitchcore etc. High values give smooth textures, better for photos.

Default: 1

number
(minimum: 0.25, maximum: 0.3)

For historical reasons, 0.3 is a starting point equivalent to normal SDXL. 0.25 will (depending on prompt) cancel background objects if they shouldn't be there, allows for single color images or high contrast pure white and black.

Default: 0.25

number
(minimum: 0, maximum: 0.5)

0.5 will have crispy sharp and contrasty images. 0.25 will be misty and delicate. Scaling is still non-linear, so 0.5 and 0.49 will have high difference compared 0.1 and 0.01 smaller. It's best to keep it as is.

Default: 0

number
(minimum: -0.1, maximum: 0.1)

Hardest to control as it changes how other parameters operate. Low value of -0.1 is great for txt2img as it helps to get more interesting results and should be used with high **Richness** (`param1` of 1.0 up to ~2.5). 0.1 is great for img2img as it keeps structure of image but allows large changes in texture and style, to be used with **Richness** of 0.4 up to 1.0.

Default: 0.1

string
Shift + Return to add a new line

Input Negative Prompt

Default: ""

file

Input image for img2img or inpaint mode

file

Input mask for inpaint mode. Black areas will be preserved, white areas will be inpainted.

integer

Width of output image

Default: 1024

integer

Height of output image

Default: 1024

integer
(minimum: 1, maximum: 4)

Number of images to output.

Default: 1

string

scheduler

Default: "Keros Euler"

integer
(minimum: 1, maximum: 500)

Number of denoising steps

Default: 50

number
(minimum: 1, maximum: 50)

Scale for classifier-free guidance

Default: 9

number
(minimum: 0, maximum: 1)

Prompt strength when using img2img / inpaint. 1.0 corresponds to full destruction of information in image

Default: 0.8

integer

Random seed. Leave blank to randomize the seed

string

Which refine style to use

Default: "no_refiner"

number
(minimum: 0, maximum: 1)

For expert_ensemble_refiner, the fraction of noise to use

Default: 0.8

integer

For base_image_refiner, the number of steps to refine, defaults to num_inference_steps

boolean

Applies a watermark to enable determining if an image is generated in downstream applications. If you have other provisions for generating or deploying images safely, you can use this to disable watermarking.

Default: true

number
(minimum: 0, maximum: 1)

LoRA additive scale. Only applicable on trained models.

Default: 0.6

boolean

This model’s safety checker can’t be disabled when running on the website. Learn more about platform safety on Replicate.

Disable safety checker for generated images. This feature is only available through the API. See [https://replicate.com/docs/how-does-replicate-work#safety](https://replicate.com/docs/how-does-replicate-work#safety)

Default: true

Output

output
Generated in

Run time and cost

This model runs on Nvidia L40S GPU hardware. We don't yet have enough runs of this model to provide performance information.

Readme

Our aim at Keros AI was to give users more control over the effect of Stable Diffusion XL inference. We achieved that by manipulating the noise at scheduler step level.

To increase the control over the results of diffusion inference we introduce the following parameters:

  • Richness (1.0-3.0): Low richness gives not much detail or even objects and misty look. High values make images stronger, rich, complex, but bugs might also happen.

  • Contrast (0.9-1.1): Quite delicate, can correct overexposure that guidance or richness create. Use 1.0 for no change.

  • Texture (0.7-1.3): Low values give high local contrast and sharper changes between objects better for illustrations, glitchcore etc. High values give smooth textures, better for photos.

  • Background (0.25-0.3): For historical reasons, 0.3 is a starting point equivalent to normal SDXL. 0.25 will (depending on prompt) cancel background objects if they shouldn’t be there, allows for single color images or high contrast pure white and black.

  • Focus (0.0-0.5): 0.5 will have crispy sharp and contrasty images. 0.25 will be misty and delicate. Scaling is still non-linear, so 0.5 and 0.49 will have high difference compared 0.1 and 0.01 smaller. It’s best to keep it as is.

  • Variance (-0.1-0.1): Hardest to control as it changes how other parameters operate. Low value of -0.1 is great for txt2img as it helps to get more interesting results and should be used with high Richness (param1 of 1.0 up to ~2.5). 0.1 is great for img2img as it keeps structure of image but allows large changes in texture and style, to be used with Richness of 0.4 up to 1.0.