(v1.5 UPDATE): Fixed the Network Rank and Network Dimensions to be 128 now.
Hello there! I've created a LoRA based on 17 movie poster images from the 1980s and 2 epochs!
I've compiled multiple movie posters of E.T., Star Wars, Back to the Future, etc.
I absolutely love the grainy realistic oil painting/aesthetic that 80s movie posters (usually) had and I wanted to see if I can replicate that through a LoRA.
Recommended LoRA weight: ~0.9-1.0
Recommended sizes: either 512x768 or 768x1024 if you want a similar dimension to the actual movie posters
Recommended checkpoints to use: Deliberate and Dreamshaper (but you can experiment with others if you'd like)
Make sure you add the trigger word "80s movie poster" in your prompt
You can also add the keywords "on the top", "on the middle", and "on the bottom" if you want to place certain images in a particular area as well.
Enjoy!
Description
FAQ
Comments (5)
An interesting idea.
But bad performance.
I recommend that you use the training parameters Network Rank (Dimension) - 128
Network Alpha - 128.
I won't rate or post any images.
I really liked the idea, but I would have worked very seriously on the model before posting it.
I sincerely wish you good luck)
Thanks for the feedback. I'll see if I can fix that in version 2 for 'ya.
Would love to see a version 2 as this is a really cool idea!
This summer's biggest blockbuster
"What Even ARE Fingers?"
An ALL AI generated movie event extravaganza!
"Parosstlo amnd Lgikee" is my favorite 80s movie.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.