Okay, so I stumbled upon this “lenny herzog” thing and figured, why not give it a shot? I’m always up for trying something new, especially if it involves messing around with images. I mean, who doesn’t love a good face swap, right?
First, I scoured the internet for any info I could find about this lenny herzog. I saw that use stable diffusion to make it. Seemed easy enough. I’d never really messed with stable diffusion before. So I first downloaded stable diffusion on my pc.
Getting Started
- Installed stable diffusion: Because it’s the tool to make it!
I muddled my way through the setup – I’m no tech wizard, believe me. I basically just clicked whatever looked like it would get me to the next step. I followed some random online guides, and somehow, it actually worked!
Then I searched some models on the internet, I chose the “revAnimated_v122EOL” model and put it in the /models/Stable-diffusion folder in my stable diffusion webui folder.
I opened my stable diffusion webui, changed to the “revAnimated_v122EOL” model on the top-left corner.
Then I typed my prompt in the prompt area, like these:
lenny herzog, best quality, masterpiece, ultra high res, (photorealistic:1.4), raw photo, 1girl, solo, medium shot, realistic, hyperrealistic, photo – realistic, facing the camera, facing the viewer, looking straight ahead, piercing gaze,
I also typed some words in Negative prompt area:
(worst quality:1.4), (low quality:1.4), (monochrome:1.1), easynegative, badhandv4,
I also set the Sampling method to “DPM++ 2M Karras”, Sampling steps to 20, Width to 512, Height to 768, CFG scale to 7.
Finally, I hit that glorious “Generate” button and…BAM! after several attempts I get some pretty awesome images.
It was a bit of a bumpy ride, to be honest. Lots of trial and error. My first few attempts looked, uh, “interesting”. Lots of weird distortions and wonky features. It definitely took some practice to figure out how to tweak everything to get the results I wanted. I played around with settings like “sampling steps” and that CFG Scale thing. I still don’t fully understand them, but I jiggled them around until things started looking better.
I changed prompt words many times.
It wasn’t perfect, but hey, I learned something new! And I got a cool new image out of it. Might try messing with different settings later, see if I can make it even better. If you’re bored and have some time to kill, give it a shot! Just be prepared for a bit of a learning curve.
