Trained on ~20 Midjourney 5.1 generations of old/damaged/rusted/broken electronics (ipods/radios/computers/devices).
You'll likely need to play around w/ the weighting to get it to conform well with prompts. Trigger is "Topnotch Artstyle"
Dreambooth (Local repo by Joe Pena) was used for training. Supermerger extension for A1111 was used to extract LoRA from CKPT.
Description
FAQ
Comments (4)
Great job!
This one is awesome! :) Thanks for the upload!
So what happens if I want to use one of your other LORA models and their trigger words are topnotch?
Yea, I've only started playing w/ LoRA w/in the past month. Prior to that I was doing hundreds of Dreambooth checkpoint models. You can't use those in tandem (unless you merge models) so I stuck w/ the same trigger. Some are still worth extracting to LoRA imho. In the future I won't use the same trigger word. Apologies if you were trying to use a couple I've posting so far but they conflict.
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.