Posted for civitai.com and tensor.art only
Each of these LoRAs take me dozens of hours and GPU usage out of pocket. Every review is appreciated and encourages me to do more of these.
-MoC
Description
new method reduces size and increases quality (32 rank)
training res: 1248px
native bucketed resolutions: 960x1472, 1216x1216
Note strength range (as I slightly overcooked this version)
FAQ
Comments (6)
Your models are very impressive! Are you interested in paid work in exchange of some models? Feel free to DM me.
Thank you so much for the work you do!! No one has done anything on Jane Wilde, did you think of adding her?
I wouldnt mind, it just depends on if I can optimise the whole dataset gathering, cleaning, optimising process. That really is the drag to the entire process. Mostly because I'm picky about source material quality...
Very impressive. Can I learn from you? Pay ~~~
What would you like to know? The LoRA game is always constantly evolving. DM me.
@ManOfCulture I'd be interested to see your training set. I've been working on a few, but they're either inflexible, or undertrained.