Greetings! I've created a single layer trained LoRA model for the Corinthian based on TheLastBen's aitoolkit configuration, thus the INSANE file size. As a result, the LoRA gives you less overbaked results compared to v1.0. You'll need to reinforced some key features of the Corinthian like his sunglasses and his age, you'll notice that in my workflow prompt. I recommend using 0.9 strength. Enjoy!
Welcome to my second LoRA, based on Boyd Holbrook's brilliant performance as "The Corinthian" from the Netflix's "The Sandman". I've learned a lot about flux training. I used Kohya ss fluxgym tool for this LoRA with default settings. I highly recommend you use this LoRA around .8 strength or you'll get artifacts. I'm hoping to remake my Blake Chillenhaal LoRA with what I learned.
Strength set to .8
If you're getting a bunch of results with no sunglasses, just make sure you include "his signature sunglasses" in the prompt
This is probably the best version of the Corinthian I can get with Flux. V1 was way over baked, but this model has a better balance between likeness and prompt flexibility. Please note the trigger word has changed since there's a possibility that people might want "Corinthian Columns" in there prompt.
1.転載モデルは学習・共有目的のみで使用し、著作権は原作者に帰属します
2.モデルの認証は公式チャンネルでご連絡ください。クリエイターの権利保護に努めています クリックして認証
