In KSampler, cfg stands for classifier free guidance and controls how much the model follows the prompt, hight the cfg, closer is matches the prompt

first image show 4, 2nd show 7 for cfg

seed in KSampler, is the starting noise, changing it will change image regardless of prompt

order matters, things that come first in prompt usually go in front, notice , olde truck is first than cabin is first

TO BE HONEST, not sure about this

he explains diffusion first, but i dont' care too much, breezing through this, taking notes

He is is explaing the settings in KSampler in the the

• fixed seed : starts noise for diffusion
• steps : is time it performs image difusion,

  This model likes 35

He add primitive node and if you select increment, it increases value by 1 each run

Add (( )) parenthesis to get things in front in prompt

** Notice sun is more pronounced, it proably reasons that the sun can't be in front of ocean