Hervy

Disclaimer: I'm a bit new to LoRA creation, so I don't expect this to blow anybody's socks off. However I think it looks pretty good.

Versions

Here's some outlined details for each new version released on this LoRA page.

This version was actually submitted to me by someone who wanted to improve the LoRA some more.

This version was trained with kohya tooling on civit's training platform, with a bigger dataset.

This is my first iteration of the LoRA. Nothing crazy interesting of note. Trained using OneTrainer.

Here's a link to my dataset for this version.