'cuz finetuning Pony is a pain in the ass.
Probably a one-time release, updates are not planned at all.
Fine-tuned with 2 different custom LoCons, with approx. 30 hours of GPU training time and the performance is still not great enough for me, and still score tag dependent.
Use as you usually use Pony.
I hate this.
Description
I got some feedback from some of the creators here to rather fine-tune AutismMix.
Well, they're kind of correct this time. It's not as bad as base Pony.
Using 7.7k dataset trained on Pony and 19.8k dataset trained on Animagine XL 3.1 to fine-tune AutismMix Confetti
Details
Downloads
29
Platform
TensorArt
Platform Status
Available
Created
4/9/2024
Updated
9/8/2024
Deleted
-
Files
Available On (4 platforms)
Same model published on other platforms. May have additional downloads or version variants.