where is version 2 and 3? well they failed and were sent to silicon heaven to live with all the little calculators.
I started from scratch, used more source images, again edited to remove text...i missed a couple... and some had backgrounds blurred out to try and show the ai that i wanted the ships not the planets (plan to do one for planets during the week)
Very happy with the results.
Happy accident: I tried the prompt "a kitchen window looking out to a field with a DaveSpaceFour landed" the idea would be a normal kitchen window looking out at a field with a spaceship landed in it... the results came out not as expected but rather awesome.
It also works well as a style... for example "A livingroom by DaveSpaceFour" ... though this was an accident and not by design lol.
I remain amazed by Stable diffusion.
I made the embed with the 768 SD 2.0 at 768x768. I generated all example images with sd 2.1768 ema pruned with DPM++ SDE 60 to 100 steps.
This is version 4 of the embed...
where is version 2 and 3? well they failed and were sent to silicon heaven to live with all the little calculators.
I started from scratch, used more source images, again edited to remove text...i missed a couple... and some had backgrounds blurred out to try and show the ai that i wanted the ships not the planets (plan to do one for planets during the week)
Very happy with the results.
Happy accident: I tried the prompt "a kitchen window looking out to a field with a DaveSpaceFour landed" the idea would be a normal kitchen window looking out at a field with a spaceship landed in it... the results came out not as expected but rather awesome.
It also works well as a style... for example "A livingroom by DaveSpaceFour" ... though this was an accident and not by design lol.
I remain amazed by Stable diffusion.
I made the embed with the 768 SD 2.0 at 768x768. I generated all example images with sd 2.1768 ema pruned with DPM++ SDE 60 to 100 steps.