<< Click to Display Table of Contents >> Navigation: 3. Script Language > AI - Artificial Intelligence Commands > SD - Stable Diffusion AI > SD - Shared Commands > Stable Diffusion - Shared Commands |
MiniRobotLanguage (MRL)
SDL.Set_Steps
Set the number of steps for the Stable Diffusion (SD) process for SDL. and for SDO.
"Beauty of Nations". Left side "native Indian", right side "African". Created with the SDL.-Command and Stable Diffusion Local.
Intention
The SDL.Set Steps command configures the number of iterations or cycles the Stable Diffusion (SD) process will undergo. By adjusting the number of steps, users can influence the depth and precision of the diffusion process, directly affecting the quality and detail of the resultant output.
In the context of the Stable Diffusion process, "steps" refer to the number of iterations the algorithm undergoes to transform, blend, or diffuse the input data into the final output. Each step can be seen as a refinement phase, where the algorithm further processes the data, adding details, enhancing features, or smoothing out inconsistencies.
The number of steps plays a pivotal role in determining the quality of the output:
•Depth of Process: More steps mean the algorithm has more opportunities to refine and process the data, leading to richer and more detailed outputs.
•Quality & Detail: As the number of steps increases, the output tends to have more clarity, depth, and detail. This can lead to sharper images, more intricate patterns, or smoother gradients.
•Sampler Dependency: The impact of steps on quality is also influenced by the chosen sampler. Some samplers might produce optimal results with fewer steps, while others might require more iterations to achieve their best output.
•Computational Time: While more steps can lead to higher quality, they also introduce more computational demands. This means the process will require more processing time to complete.
'This example configures the SD process to undergo 100 steps.
SDL.Set Steps|100
'In this instance, the number of steps for the SD process defaults to 70.
SDL.Set Steps
While increasing the number of steps can lead to enhanced quality and detail in the Stable Diffusion process, it's crucial to balance this with the associated computational costs and the characteristics of the chosen sampler.
Syntax
SDL.Set_Steps[|P1]
Parameter Explanation
P1 - (optional) <Number of Steps>: Optional. The desired number of steps for the SD process. This should be a number between 10 and 150. If omitted, the default value is set to 70.
Example
'***********************************
'
'***********************************
$$PRO=wooden Geodesic Dome,honeycomb, natural, a rainbow
$$NEG=ugly,comic,unrealistic,fat,unhealty,malformed faces
$$FIL=?exeloc\Sample_?.png
SDL.SetSteps|150
FOR.$$LM8|1|25
RND.1|21|$$SEM
SDL.Set_Sampler|$$SEM
POP.$$SAM
SDL.Set Model Free|dreamlike-photoreal-2.0.safetensors
SDL.Set Extra Parameter|"restore_faces": true
VAN.$$TIM=#dtime#
SDL.gtf|$$PRO|$$FIL|$$NEG|4
VAN.$$TIM=#dsince#|i
DBP. Set $$SAM Sampler. In $$TIM Seconds.
NEX.
DMP.6
MBX.!
ENR.
Remarks
-
Limitations:
-
See also:
• ! Smart Package Robot 's Parallel Robot Operations